❌

Normal view

Today β€” 18 December 2025Tech

CISA Flags Critical ASUS Live Update Flaw After Evidence of Active Exploitation

The U.S. Cybersecurity and Infrastructure Security Agency (CISA) on Wednesday added a critical flaw impacting ASUS Live Update to its Known Exploited Vulnerabilities (KEV) catalog, citing evidence of active exploitation. The vulnerability, tracked as CVE-2025-59374 (CVSS score: 9.3), has been described as an "embedded malicious code vulnerability" introduced by means of a supply chain compromise

Cisco Warns of Active Attacks Exploiting Unpatched 0-Day in AsyncOS Email Security Appliances

Cisco has alerted users to a maximum-severity zero-day flaw in Cisco AsyncOS software that has been actively exploited by a China-nexus advanced persistent threat (APT) actor codenamed UAT-9686 in attacks targeting Cisco Secure Email Gateway and Cisco Secure Email and Web Manager. The networking equipment major said it became aware of the intrusion campaign on December 10, 2025, and that it

Forza Motorsport Effectively Sunset After Prior Layoffs

18 December 2025 at 09:24
Forza Motorsport, in its most recent iteration, launched in October 2023, but despite the relatively recent release date, it seems as though the game's developer, Turn 10 Studios, is already starting to sunset the multiplayer racing sim, although the game will remain playable for the foreseeable future. Turn10 said in a recent announcement that it would no longer be adding any in-game contentβ€”cars, tracks, or featuresβ€”to Forza Motorsport, nor will the studio provide regular bug fixes or updates. However, the studio says that it will still offer online support for multiplayer gameplay, although it's unclear how long this will last.

In addition to the continued server maintenance, Turn 10 will also host a number of special events and competitions throughout 2026, seemingly in an effort to keep players around and as a way for players to earn past special event rewards, cars, and other content that was previously limited. Turn 10 says it is ceasing further development of Forza Motorsport in order to "shift its focus" to the upcoming Forza Horizon 6 game that was announced earlier this year. This news also comes after Turn 10 Studios lost about half of its staff in July 2025 when Xbox Games laid off a sizeable number of staff across multiple subsidiaries.

Half-Life 3 May Launch Alongside Steam Machine in Early 2026

18 December 2025 at 08:10
Fans of the Half-Life game series have long been grasping at straws for any news about a potential Half-Life 3 release date, even over two decades after the launch of its would-be predecessor. Although there have been leaks and rumors about Half-Life 3 here and there throughout the years, none of those have yet panned out. The latest claim was that Half-Life 3 would be announced alongside the upcoming Valve Steam Machine, but that announcement has since come and gone without a peep, leading to speculation that the third installation in the game series would instead launch alongside the Steam Machine when the living-room gaming PC actually launches commercially in early 2026. Now, Mike Straw, a game journalist with a reputation for reporting on leaks from insider sources, has claimed in a recent appearance on the Insider Gaming Weekly podcast on YouTube that "everybody I've talked to are still adamant that this is a game that will be a launch title with the Steam Machine."

He also mentions that there are concerns that Valve has not finalized the price of the Steam Machine, largely due to the volatility and sky-high prices of DRAM. His comments also suggest that Valve has been feeding false launch datesβ€”all of which have since passed with no announcements from Valveβ€”to a number of games journalists to determine leak sources and potentially even obfuscate the actual launch date of Half-Life 3. Valve has not committed to a specific launch date for the Steam Machine, nor has it given any indication of pricing other than that it will launch in "early 2026," indicating Q1, and that it will be priced more in line with a similar DIY gaming PC.

Splitgate: Arena Reloaded Emerges as Reworked Splitgate 2 After Disastrous May Launch

18 December 2025 at 07:22
The launch of Splitgate 2 has been a bit of a mess, with the game initially launching to criticism from fans of the franchise earlier this year. In response to the negative reception, 1047 Games, the studio behind the fast-paced FPS pulled the game from Steam Early Access, where it was available at the time and announced that it would be going back to the drawing board and re-releasing Splitgate 2 in December this year. Now, 1047 Games has delivered on that promise, shipping Splitgate: Arena Reloaded as the full-fledged 1.0 release of Splitgate 2. The game studio has published a long list of changes that have been implemented for the full release, but the gist of it is that Splitgate: Arena Reloaded aims to keep the identity and some of the unique gameplay that made the original Splitgate special while also modernizing the game, "add the features people consider standard today," and tighten the gameplay loop to focus on the arena shooter mechanics more than the original Splitgate 2.

Much like the original game in the franchise, Splitgate: Arena Reloaded is a free-to-play arena shooter that spices up the gunplay with portals for sneaky tactical moves and increased mobility. It claims much of the same vertical mobility and fast-paced gunplay as the previous Splitgate game, and it features 20 arenas, 18 primary weapons, 9 secondary weapons, and almost dizzying number of game modes for multiplayer PvP and co-op gameplay. The biggest changes to Splitgate: Arena Reloaded appear to be the removal of factions and abilities to focus on the portal mechanics and gunplay and clean up the gameplay loop. Players can also now deploy both the entry and exit portals on a wall to block other players from using it, and the Arena Reloaded also features EMP grenades to deactivate portals from an area and control the space. The developer also says that the game has seen performance optimizations, specifically focusing on frame rate consistency on low-end PCs and last-gen consoles. At the same time, Spligate: Arena Reloaded has received a visual, UI, and audio overhaul to make the game feel more like a classic arena shooter.

Google’s Gemini 3 Flash: A New Standard for Efficient AI Performance

18 December 2025 at 05:45

The post Google’s Gemini 3 Flash: A New Standard for Efficient AI Performance appeared first on StartupHub.ai.

The recent unveiling of Google’s Gemini 3 Flash by Matthew Berman marks a pivotal moment in the generative AI landscape, signaling a clear shift towards models that prioritize not just raw intelligence but also unparalleled efficiency and cost-effectiveness. Berman, a prominent AI commentator from Forward Future AI, meticulously detailed how this new iteration of Gemini […]

The post Google’s Gemini 3 Flash: A New Standard for Efficient AI Performance appeared first on StartupHub.ai.

Claude and Anthropic Emerge as New AI Champion, Consumer Trends Shift Rapidly

18 December 2025 at 05:16

The post Claude and Anthropic Emerge as New AI Champion, Consumer Trends Shift Rapidly appeared first on StartupHub.ai.

The landscape of consumer preference in generative AI, video streaming, and wireless carriers is undergoing a seismic shift, with established leaders facing unexpected challengers. This dynamic was vividly illustrated in a recent interview on Mad Money, where HundredX Founder and CEO Rob Pace joined host Jim Cramer to dissect critical consumer trends. HundredX, a privately […]

The post Claude and Anthropic Emerge as New AI Champion, Consumer Trends Shift Rapidly appeared first on StartupHub.ai.

Google fixed month-long delay with page indexing report

18 December 2025 at 05:23

Google Search ConsoleΒ appears to have fixed the month-long delay with the page indexing report just about an hour ago. The report is now showing data as early as a few days ago, which is the normal timeframe for when this report is updated.

Plus, emails about indexing issues have started going out from Search Console to site owners again.

Page indexing report.Β It shows which pages Google can find and index on your site, along with any problems. You can also submit fixes there and see whether Google confirms they worked. Site owners and SEOs were stuck, they were unable to verify their β€œfixes” and unable to see if new pages were being indexed and if old pages were having issues being indexed.

Fixed. Here is a screenshot of the report showing December 14th, a much more recent date than the November 21st date that many were stuck on:

Google also fixed the performance reports delay just yesterday. So all the major reports should now be running normally, that is until they break again – which is not that uncommon.

Why we care. Many of you were unable to do full reporting for your SEO clients and stakeholders over the past month. Now you can get recent data both for page indexing and performance reports.

So you should be able to catch up on your reporting before you go into the holiday season.

BeatSync PRO – Auto-sync music and visuals to render pro videos in minutes


BeatSync PRO is a Windows desktop app that turns your music into beat-synced videos in minutes. It detects every hit with Β±5 ms precision, analyzes clip content with AI, and auto-matches visuals to your track. Render locally on your NVIDIA GPU and export standard MP4s for social, shows, or VJ sets. Plans include curated AI clip libraries, royalty-free audio, and a full commercial license. Import your own footage, reuse analyzed clips without extra credits, and keep quality intact with non-destructive exports.

View startup

Teaching Sand to Think: The AI Infrastructure Moonshot

18 December 2025 at 04:45

The post Teaching Sand to Think: The AI Infrastructure Moonshot appeared first on StartupHub.ai.

β€œTeaching sand to think” serves as the profound metaphor for the ambition underpinning the current era of artificial intelligence. It encapsulates the monumental task of transforming inert matter into conscious, capable systems, an endeavor that the a16z Original video, β€œTwo Futures | Runtime 2025,” posits as the biggest infrastructure supercycle in history. This visionary production, […]

The post Teaching Sand to Think: The AI Infrastructure Moonshot appeared first on StartupHub.ai.

Metal: Hellsinger Studio Co-Founder Preps for Studio Revival After October Closure

18 December 2025 at 03:19
The Outsiders, the game studio behind rhythm boomer-shooter, Metal: Hellsinger, became one of the victims of the volatile gaming industry when Funcom shuttered the studio back in October 2025. The studio shutdown was despite some estimates putting Metal: Hellsinger sales at nearly 2 million units since its launch in 2022, and it was only a small part of the large-scale layoffs that Funcom implemented shortly after the release of Dune: Awakening. Now, The Outsiders co-founder, David Goldfarb, has said in an interview with PC Gamer that he and his co-founder are planning for a revival of the studioβ€”which he calls "the 2.0 version of The Outsiders."

Goldfarb blames a variety of factors for the studio's closure, including everything from Covid hiring to lack of funds for investment, and even Metal: Hellsinger's sales, which he says were good, but not quite the exceptional sales figures that are needed to be successful in the current climate. So far, the revitalization of The Outsiders is still in its infancy, and Goldfarb says that even he is not sure about where the new studio will go, although he says that "we're working on something, and we're talking to people, and then if things work out, we'll be in a position to go ahead and make that, and maybe get some of our colleagues back." It's worth noting that the indie game scene is still very much alive and well, and games like Hollow Knight: Silksong and Terraria, among many others, are examples of how indie studios can succeed on creative vision and careful risk-taking, despite the current gaming climate.

Windows Server 2025 Gets Native NVMe SSD Support After 12 Years

18 December 2025 at 02:31
Microsoft has introduced native NVMe SSD support in its Windows Server 2025 build, now available as an opt-in feature for users. After 12 years of NVMe's existence and years of support in the Linux kernel and Linux-based operating systems, native support has been added to the Windows Server stack. Previously, Microsoft converted NVMe drive commands into SCSI commands, which resulted in processing latency and overhead, slowing down read/write speeds in typically high-performing storage configurations. With native NVMe support, the entire stack and I/O processing have been redesigned to achieve optimal SSD performance.

System administrators don't need data to recognize the significant performance improvements that native NVMe support will bring. Windows Server 2025's native NVMe support fundamentally transforms storage performance by enabling direct multi-queue access to modern hardware. It delivers up to 3.3 million IOPS on PCIe Gen 5 SSDs and exceeds 10 million IOPS on HBAs, while reducing latency through streamlined, lock-free I/O paths. This leap in efficiency is essential because traditional SCSI-based processing, originally designed for rotational disks with a single-queue model limited to 32 commands, cannot fully utilize flash storage.

Google’s AI Mode Personal Context Features β€œStill To Come”

18 December 2025 at 02:29

Google SVP Nick Fox says AI Mode’s personal context features, including opt-in Gmail connections teased at I/O, are still in internal testing.

The post Google’s AI Mode Personal Context Features β€œStill To Come” appeared first on Search Engine Journal.

The Future of AI: Arm’s Vision for Distributed Intelligence

18 December 2025 at 02:18

The post The Future of AI: Arm’s Vision for Distributed Intelligence appeared first on StartupHub.ai.

Arm's 2026 predictions highlight a future of AI characterized by distributed intelligence, modular silicon, and pervasive on-device capabilities.

The post The Future of AI: Arm’s Vision for Distributed Intelligence appeared first on StartupHub.ai.

Silver’s Ascent: AI and Data Centers Drive Critical Metal to Record Highs

18 December 2025 at 02:15

The post Silver’s Ascent: AI and Data Centers Drive Critical Metal to Record Highs appeared first on StartupHub.ai.

As artificial intelligence and data centers rapidly expand, their insatiable demand for critical materials is creating unexpected market dynamics. Michael Steinmann, CEO of Pan American Silver, illuminated this burgeoning trend in a recent CNBC β€˜Closing Bell Overtime’ interview with Sara Eisen and Jon Fortt, discussing the forces propelling silver to record highs and the structural […]

The post Silver’s Ascent: AI and Data Centers Drive Critical Metal to Record Highs appeared first on StartupHub.ai.

Agentic AI Design: Beyond Screens, Into Systems

18 December 2025 at 01:17

The post Agentic AI Design: Beyond Screens, Into Systems appeared first on StartupHub.ai.

Agentic AI design fundamentally redefines the role of designers, moving them from interface creators to architects of AI's understanding and behavior.

The post Agentic AI Design: Beyond Screens, Into Systems appeared first on StartupHub.ai.

AI Sector Faces Reality Check: Arrogance, Debt, and Discerning Capital

18 December 2025 at 01:15

The post AI Sector Faces Reality Check: Arrogance, Debt, and Discerning Capital appeared first on StartupHub.ai.

The era of broad-brush enthusiasm for artificial intelligence in the tech sector appears to be waning, giving way to a more discerning market where fundamental strength, rather than sheer hype, is becoming the ultimate arbiter of value. This was the overarching sentiment during a recent discussion on CNBC’s β€œClosing Bell,” where Adam Parker of Trivariate […]

The post AI Sector Faces Reality Check: Arrogance, Debt, and Discerning Capital appeared first on StartupHub.ai.

China sidesteps US ban with EUV chipmaking breakthrough

18 December 2025 at 00:41

Chinese scientists build a working EUV prototype, striking a major blow against America’s semiconductor leadership The US has placed heavy restrictions on tech exports to China, hoping to slow down the nation’s progress and limit the growth of its tech sector. Specifically, the US wants to prevent China from making cutting-edge semiconductors, keeping its chips […]

The post China sidesteps US ban with EUV chipmaking breakthrough appeared first on OC3D.

(PR) Micron Technology, Inc. Reports Results for the First Quarter of Fiscal 2026

18 December 2025 at 02:03
Micron Technology, Inc. (Nasdaq: MU) today announced results for its first quarter of fiscal 2026, which ended November 27, 2025.

Fiscal Q1 2026 highlights
  • Revenue of $13.64 billion versus $11.32 billion for the prior quarter and $8.71 billion for the same period last year
  • GAAP net income of $5.24 billion, or $4.60 per diluted share
  • Non-GAAP net income of $5.48 billion, or $4.78 per diluted share
  • Operating cash flow of $8.41 billion versus $5.73 billion for the prior quarter and $3.24 billion for the same period last year

IdeaLift – The fastest way to go from idea β†’ shipped


IdeaLift turns everyday team conversations into shipped product work. With a simple emoji reaction in Slack or Discord, IdeaLift captures the idea, uses AI to clean it up, categorize it, check for duplicates, and create a structured issue in GitHub or Linearβ€”no copy-pasting, no context switching, and no lost ideas buried in chat threads.

IdeaLift works wherever ideas happen: Slack, Discord, Microsoft Teams, meetings, the web, and across 5,000+ apps via Zapier. Teams use it to capture feature requests, bugs, and decisions in real time, then move seamlessly from raw conversation to clear, actionable workβ€”faster, cleaner, and with full context intact.

View startup

Sales of chip production equipment to reach $156 billion by 2027 β€” China, Taiwan, and Korea lead intense demand

Driven by demand for AI and HPC accelerators, sales of chip production equipment are projected to increase through 2027. Asian countries are expected to lead the pack, according to SEMI.

AI-Driven Kernels: Accelerating PyTorch with Agentic Optimization

18 December 2025 at 00:47

The post AI-Driven Kernels: Accelerating PyTorch with Agentic Optimization appeared first on StartupHub.ai.

In the relentless pursuit of computational efficiency, the fine art of low-level kernel optimization has long remained the exclusive domain of a select few, a bottleneck in the age of rapidly evolving AI models. Natalie Serrino, Co-founder of Gimlet Labs, recently illuminated this critical challenge and her company’s novel approach at the AIE Code Summit. […]

The post AI-Driven Kernels: Accelerating PyTorch with Agentic Optimization appeared first on StartupHub.ai.

UCSD Lab Advances Low-Latency LLM Serving with DGX B200

18 December 2025 at 00:17

The post UCSD Lab Advances Low-Latency LLM Serving with DGX B200 appeared first on StartupHub.ai.

UC San Diego's Hao AI Lab is pushing the frontier of low-latency LLM serving by leveraging NVIDIA's DGX B200 system and pioneering disaggregated inference.

The post UCSD Lab Advances Low-Latency LLM Serving with DGX B200 appeared first on StartupHub.ai.

Hut 8’s Strategic Pivot: Fueling AI with Financial Fortitude

18 December 2025 at 00:15

The post Hut 8’s Strategic Pivot: Fueling AI with Financial Fortitude appeared first on StartupHub.ai.

β€œCredit counterparty was extremely important,” stated Asher Genoot, CEO of Hut 8, on CNBC’s β€œPower Lunch,” underscoring a pivotal shift in the digital infrastructure landscape. Genoot, speaking with the CNBC team, provided commentary on Hut 8’s recent deal with Fluidstack, backed by Google, and its implications for data center demand and energy infrastructure. His insights […]

The post Hut 8’s Strategic Pivot: Fueling AI with Financial Fortitude appeared first on StartupHub.ai.

Amazon in talks to invest $10 billion or more in OpenAI

17 December 2025 at 23:45

The post Amazon in talks to invest $10 billion or more in OpenAI appeared first on StartupHub.ai.

The escalating arms race in artificial intelligence demands not just groundbreaking algorithms but the foundational infrastructure to train and deploy them. In a recent CNBC segment, reporter MacKenzie Sigalos, in conversation with anchor Kelly Evans, unpacked the strategic implications of Amazon’s potential $10 billion-plus investment in OpenAI, a deal poised to reshape the competitive landscape […]

The post Amazon in talks to invest $10 billion or more in OpenAI appeared first on StartupHub.ai.

NVIDIA’s OpenUSD Boosts Robotaxi AI Safety

17 December 2025 at 23:19

The post NVIDIA’s OpenUSD Boosts Robotaxi AI Safety appeared first on StartupHub.ai.

NVIDIA's integrated approach with OpenUSD, Omniverse, and Halos is setting new standards for robotaxi AI safety through advanced simulation and certification.

The post NVIDIA’s OpenUSD Boosts Robotaxi AI Safety appeared first on StartupHub.ai.

Code World Model: Meta’s Leap Beyond Code Syntax to Computational Reasoning

17 December 2025 at 23:15

The post Code World Model: Meta’s Leap Beyond Code Syntax to Computational Reasoning appeared first on StartupHub.ai.

The future of AI-driven software engineering hinges not merely on generating code, but on truly understanding its computational dynamics. This profound shift was at the heart of Jacob Kahn’s presentation on the Code World Model (CWM) at the AI Engineer Code Summit. Kahn, a Research Scientist at FAIR Meta, introduced CWM as a novel world-model […]

The post Code World Model: Meta’s Leap Beyond Code Syntax to Computational Reasoning appeared first on StartupHub.ai.

Googlebot dominates web crawling in 2025 as AI bots surge: Report

18 December 2025 at 00:42
AI search crawlers, user agents, and bots

Googlebot once again generated more traffic than any other crawler in 2025, according to a new Cloudflare report. It outpaced every search and AI bot as Google continued crawling the web for search indexing and AI training.

By the numbers. Googlebot accounted for more than 25% of all Verified Bot traffic observed by Cloudflare.

  • Googlebot alone generated 4.5% of all HTML request traffic – more than all other AI bots combined (4.2%).
  • AI β€œuser action” crawling surged more than 15x year over year, showing a sharp rise in bots that simulate human behavior.
  • Googlebot’s crawl volume dwarfed every other AI crawler, including OpenAI, Anthropic, and Meta.

AI crawling surges. AI crawlers were the most frequently fully disallowed user agents in robots.txt files.

  • Anthropic showed the highest crawl-to-refer ratio among major AI and search platforms, meaning it crawled far more content than it sent back as traffic. The ratio peaked near ~500,000:1 early in the year, then settled between ~25,000:1 and ~100,000:1 after May. For comparison:
    • OpenAI spiked to ~3,700:1 in March.
    • Perplexity was the lowest among major AI platforms. It started below 100:1, briefly jumped above 700:1 in late March during a PerplexityBot crawl spike, then stayed mostly below 400:1 and under 200:1 from September onward.

Search platforms looked very different:

  • Microsoft hovered between ~50:1 and ~70:1 with a weekly cycle.
  • Google rose from just over ~3:1 to ~30:1 by April, fell back to ~3:1 by mid-July, then gradually increased again.
  • DuckDuckGo stayed below 1:1 for the first three quarters, then jumped to ~1.5:1 in mid-October and remained elevated.

Google still monopolizes search. Traditional search dominance barely changed.

  • Google remained the top search engine by a wide margin, delivering nearly 90% of search engine referral traffic.
  • Bing (3.1%), Yandex (2.0%), Baidu (1.4%), and DuckDuckGo (1.2%) rounded out the top five.
  • Cloudflare saw minimal movement during the year.
    • Google stayed dominant throughout.
    • Yandex slipped from 2.5% in May to 1.5% in July.
    • Baidu rose from 0.9% in April to 1.6% in June.

The report. The 2025 Cloudflare Radar Year in Review: The rise of AI, post-quantum, and record-breaking DDoS attacks

SonicWall Fixes Actively Exploited CVE-2025-40602 in SMA 100 Appliances

SonicWall has rolled out fixes to address a security flaw in Secure Mobile Access (SMA) 100 series appliances that it said has been actively exploited in the wild. The vulnerability, tracked as CVE-2025-40602 (CVSS score: 6.6), concerns a case of local privilege escalation that arises as a result of insufficient authorization in the appliance management console (AMC). It affects the following

Kimwolf Botnet Hijacks 1.8 Million Android TVs, Launches Large-Scale DDoS Attacks

A new distributed denial-of-service (DDoS) botnet known as Kimwolf has enlisted a massive army of no less than 1.8 million infected devices comprising Android-based TVs, set-top boxes, and tablets, and may be associated with another botnet known as AISURU, according to findings from QiAnXin XLab. "Kimwolf is a botnet compiled using the NDK [Native Development Kit]," the company said in a report

(PR) Texas Instruments Begins Production at Its 300 mm Semiconductor Fab in Sherman, Texas

17 December 2025 at 23:15
Texas Instruments (TI) today announced the start of production at its newest semiconductor fab in Sherman, TX, just three and a half years after breaking ground. Leaders from TI were joined by local and state elected officials to celebrate the opening of this state-of-the-art 300 mm semiconductor fab in North Texas.

The new facility, called SM1, will ramp according to customer demand, ultimately producing tens of millions of chips daily that go into nearly every electronic deviceβ€”from smartphones, automotive systems, and life-saving medical devices to industrial robots, smart home appliances, and data centers.

AMD Ryzen 7 9850X3D Listed at $553, Slightly Above Ryzen 7 9800X3D

17 December 2025 at 22:22
AMD's upcoming Ryzen 7 9850X3D processor is slowly hitting the retailer space, and the pricing has been leaked by @momomo_us. At $553, the new "Granite Ridge" Ryzen 9000 X3D SKU is priced roughly $70 more than the Ryzen 7 9800X3D at its launch. The Ryzen 7 9850X3D CPU features 8 cores and 16 threads, 96 MB of L3 cache, and a boost clock up to 5.6 GHz, with a 120 W power envelope. This represents a 400 MHz increase over the Ryzen 7 9800X3D model, which has the same cache and core configuration. At Switzerland retailer Orderflow.ch, the CPU is listed for 473.55 CHF, which is 20% higher than the Ryzen 7 9800X3D.

Officially, AMD's Ryzen 9000 series supports DDR5 memory speeds up to 5600 MT/s, but this model could push that number to 9,800 MT/s if earlier reports are accurate. AMD is likely using higher-binned dies it has saved over the past few months to offer a better alternative to Intel's upcoming "Arrow Lake Refresh," scheduled for early 2026, and "Nova Lake," scheduled for late 2026. As Intel has confirmed it will ship with native support for DDR5 memory speeds reaching 7200 MT/s on CUDIMM modules, AMD is fighting with a faster CPU clock and supposedly better memory speeds to offer good competition in the consumer DIY PC space.

Merriam-Webster's word of the year is "slop"

17 December 2025 at 22:36

"Slop" is Merriam-Webster's 2025 word of the year, and the company is quick to clarify that human editors did all the selecting. According to Encyclopedia Britannica's subsidiary, capturing the elusive, low-quality content flooding screens in recent years in a single four-letter word once again demonstrates the English language's ability to...

Read Entire Article

Yesterday β€” 17 December 2025Tech

Chandler’s AI Data Center Rejection Signals Shifting Local Tech Landscape

17 December 2025 at 22:46

The post Chandler’s AI Data Center Rejection Signals Shifting Local Tech Landscape appeared first on StartupHub.ai.

The unanimous decision by the Chandler City Council to reject a proposed AI data center, a move openly endorsed by Mayor Kevin Hartke, underscores a growing friction between the rapid expansion of artificial intelligence infrastructure and the specific economic and environmental priorities of local communities. This wasn’t merely a political skirmish; it was a calculated […]

The post Chandler’s AI Data Center Rejection Signals Shifting Local Tech Landscape appeared first on StartupHub.ai.

Anthropic Wins TTFT, But OpenAI Dominates LLM Benchmarks

17 December 2025 at 22:25

The post Anthropic Wins TTFT, But OpenAI Dominates LLM Benchmarks appeared first on StartupHub.ai.

New LLM benchmarks reveal a critical trade-off: Anthropic models deliver instant responsiveness, while OpenAI maintains a commanding lead in raw generation throughput.

The post Anthropic Wins TTFT, But OpenAI Dominates LLM Benchmarks appeared first on StartupHub.ai.

Binti and Claude: Accelerating Foster Care Licensing Through Applied AI

17 December 2025 at 22:16

The post Binti and Claude: Accelerating Foster Care Licensing Through Applied AI appeared first on StartupHub.ai.

Artificial intelligence is not merely a tool for efficiency; it is a catalyst for re-humanizing critical public services, freeing professionals from administrative burdens to focus on their profound impact. This principle stands at the core of Binti’s mission, as illuminated by a recent video showcasing their integration of Anthropic’s Claude AI to revolutionize foster family […]

The post Binti and Claude: Accelerating Foster Care Licensing Through Applied AI appeared first on StartupHub.ai.

Amazon’s OpenAI Bet Signals Intensifying AI Chip Wars

17 December 2025 at 21:45

The post Amazon’s OpenAI Bet Signals Intensifying AI Chip Wars appeared first on StartupHub.ai.

The proposed multi-billion dollar investment from Amazon into OpenAI, coupled with OpenAI’s increased utilization of Amazon’s proprietary Trainium chips, marks a pivotal moment in the escalating battle for AI infrastructure dominance. This isn’t merely a financial transaction; it represents a strategic realignment that underscores the critical importance of specialized silicon and diversified compute resources in […]

The post Amazon’s OpenAI Bet Signals Intensifying AI Chip Wars appeared first on StartupHub.ai.

Proxies for prompts: Emulate how your audience may be looking for you

17 December 2025 at 22:35
Proxies For Prompts – Featured image

In this new era of generative AI technology, searchers have begun to swap keywords with prompts. Shorter and long-tail queries are being replaced by more conversational prompts, which tend to be longer and more in-depth. These days, searchers are expecting more complete answers than a paginated list of results.

Until we get an AI-specific equivalent of Google Search Console or Bing Webmaster Tools, we can’t really see for certain what or how our audience is behaving on AI search platforms as they look for our content, brands or products.

However, we can still look for proxies to emulate how this journey works. Here are multiple ways to use other data points as proxies to find prompts used by your audience. You can then use your AI Tracking Tool of choice to track how these prompts are performing.

People Also Ask

Hidden in plain sight, you can use a very popular SERP feature to move from keywords to prompts/questions. People Also Ask (PAA) was introduced in 2014 and suggests multiple related questions to your query. You can easily go from a keyword to a list of questions. When clicking on any PAA result, the list expands and gives you more terms.

Go query by query to find relevant PAAs, or you can use AlsoAsked to extract the exact questions at scale. PAAs are long questions that attempt to answer the next questions asked in the search journey, so they’re a step closer to prompts written in AI Search platforms.

Userbots

Userbots like ChatGPT-User and Perplexity‑User are powerful ways to see how your pages are being used in AI Search. It doesn’t give you the prompts they use, but it’ll help you assess which pages are being cited without trying to guess by tracking prompts that may or may not be relevant at all.

These bots ping the URLs on your website when they’re used to formulate an answer to a user. The process is called RAG (Retrieval-Augmented Generation).

Your customers search everywhere. Make sure your brand shows up.

The SEO toolkit you know, plus the AI visibility data you need.

Start Free Trial
Get started with
Semrush One Logo

To paraphrase Mike King, one of the most trusted sources in SEO & AI, RAG is a mechanism by which a language model can be β€œgrounded” in facts or learn from existing content to produce a more relevant output with a lower likelihood of hallucination.

Translation: Your page was used as an answer and, in some shape or form, your content helped a user. This may give you clues about what type of content is most used by your audience on these platforms, even if the answer hasn’t turned into a click.

Historically, log files have been difficult for SEOs to get, despite every website having them in its servers (yet another reason why SEOs should have access to them!).

You could use a combination of pages with userbot visits, search for their main keywords (as seen on Semrush or GSC), and then see which PAAs Google displayed.Β 

Long queries on GSC/BWT

Despite not giving a breakdown of AI Mode or AI Overview queries on GSC, smart SEOs are finding proxies that can be used to find queries that resemble the behavior we expect in these platforms. One of them is Ziggy Shtrosberg, who came up with a huge regex you can copy and paste to your GSC.

His guidelines are to:

  1. Filter by Search Appearance: Desktop
  2. Add a page filter with your root domain (e.g., β€œhttps://www[dot]example[dot]com/”) and add this massive regex under the Query filter:

^(generate|create|write|make|build|design|develop|use|produce|help|assist|guide|show|teach|explain|tell|list|summarize|analyze|compare|give me|you have|you can|where|review|research|find|draft|compose|extract|process|convert|transform|plan|strategy|approach|method|framework|structure|overview|summary|breakdown|rundown|digest|perspectives|viewpoints|opinions|approaches|angles|pros and cons|advantages and disadvantages|benefits and drawbacks|assuming|suppose|imagine|consider|step by step|procedure|workflow|act as|adapt|prepare|advise|appraise|instruct|prompt|amend|change|advocate|aid|assess|criticise|modify|examine|your|assign|appoint|delegate|nominate|improve|expand|calculate|classify|rank|challenge|check|categorize|order|tag|scan|study|conduct|contradict|update|copy|paste|please|can you|could you|would you|help me|i need|i want|i'm looking for|im looking for|how do i|how can i|what's the|whats the|walk me through|break down|pretend you're|pretend youre|you are a|as a|from the perspective of|in the style of|format this as|write this in|make it|rewrite|i'm trying to|im trying to|i'm struggling with|im struggling with|i have a problem|i'm working on|im working on|what's better|whats better|which|pros and cons of|recommend|suggest|show me how|guide me through|what are the steps|how do i start|whats the process|take me through|outline the procedure|brainstorm|come up with|think of|invent|what if|lets explore|let's explore|help me think|i'm a beginner|im a beginner|as someone who|given that i|in my situation|for my project|i'm currently|im currently|my goal is|depending on|based on|taking into account|considering|given the constraints|with the limitation|improve this|make this better|optimize|refine|polish|enhance|revise|teach me|i want to learn|i don't understand|i dont understand|can you clarify|what does this mean|eli5|i'm confused about|im confused about|also|additionally|furthermore|by the way|who's|whos|find|more|next|also|another|thanks|thank you|please)( [^" "]*){9,}$

Take this strategy with a pinch of salt, as some of these queries might be generated by LLM trackers.Β 

For instance, I found a pattern of prompts starting with β€œevaluate,” which have a high number of impressions by zero clicks (not a small number of clicks, exactly zero clicks). If longer prompts have a high number of impressions and no clicks, beware that it might not be humans using these prompts.

Perplexity follow-up questions

One of the main AI Search platforms, Perplexity has a feature called β€œRelated” where it displays up to five follow-up prompts. While the initial prompt is still yours and may not be how others are prompting, the related follow-ups are still a good indicator of how humans promptβ€”or at least how the platform expects humans would.

These answers are country-specific, so run your research locally.

Semrush AI Visibility Tool

Considering we don’t have the search volume metric per single keyword and that prompts are a lot more unique than keywords, it’s not realistic to track every single prompt relevant to our companies. A way to mitigate this is to combine these prompts into topics and use AI to summarize what they mean.

The new Semrush AI Visibility Tool has a feature called β€œPrompt Research” that matches your keyword to a topic and gives you a list of prompts alongside brands mentioned, intent, and sources.

Currently, the tool allows you to filter results between the US and the UK, including the full AI Response and a list of brands and URLs.Β 

Even though I typed a single keyword (β€œused cars”), it picked the closest available topic (β€œUsed Car Sales and Dealerships”) and returned me all prompts, brand mentions, and source domains.

You might decide not to track single prompts, which can grow fast and become overwhelming to measure. Rather, use the Semrush prompt database for optimization and measure the results by looking at the whole topic performance.

Without grounding, your chances are low

Keep in mind that not every prompt requires RAG, meaning that if the answer is already on the AI Search platform training data, no pages will appear as sources. For some brands, just getting a mention is fine. If, say, someone is looking for a museum or restaurant to visit, the mention might be enough to convince them to reach the destination and convert offline (e.g., buy a ticket or a meal).

In most cases, however, SEOs are still looking for traffic, so the prompt must list pages in their answers to give you a chance to be visible. Ironically, while the results you get from ChatGPT are one answer instead of a SERP, the LLM is actually doing searches for you in the background.Β 

Luckily, you can find:

  • The searches ChatGPT is doing in the background
  • The probability of this search requiring a RAG

You can find these by looking for β€œqueries,” β€œsearch_queries,” and β€œsearch_prob” inside Chrome Dev Tools (Inspect > Network > Conversation > Response).

Or, to simplify, you can add this script as a bookmark on Chrome and click on it after prompting a question on ChatGPT. This is an improved version of Ziggy Shtrosberg’s script.

While these searches look more like traditional searchers as opposed to prompts, your strategy may be to optimize for them and win on AI search as a secondary benefit.

These are the searches ChatGPT did in the background for the prompt β€œResearch five all-inclusive hotels to visit in Antalya with my family. Please give me a price estimate for a five day stay.”

When it comes to search_prob (also on the script above), it’s the probability that an answer requires grounding (RAG). This answer ranges between 0 (low) and 1 (high). Every answer is unique (even if you and I search for the same prompt, we’ll have different answers), so this can act as a proxy for the opportunity of pages being listed as a source.

As with every new technology, things change fast. How people use AI tools and which tools are being used are constantly changing. New models (like ChatGPT5) change how RAG is used, and the increase in adoption across different industries also affects what prompts you should track, so you must also evolve and reevaluate what and how to track AI searches.

See the complete picture of your search visibility.

Track, optimize, and win in Google and AI search from one platform.

Start Free Trial
Get started with
Semrush One Logo

LinkedIn opens up top-of-feed Reserved Ads to all managed advertisers

17 December 2025 at 22:13
LinkedIn Ads retargeting: How to reach prospects at every funnel stage

LinkedIn is making Reserved Ads generally available to all managed accounts, giving marketers the ability to lock in the first ad slot in the feed for premium visibility.

What’s new. Reserved Ads let advertisers secure top-of-feed placement at a fixed rate, providing predictable delivery, consistent reach, and greater share of voice. Early results show the format drives up to 75% higher dwell time, 88% higher view-through rates, and delivers 99% of forecasted impressions, according to LinkedIn.

How it works. Reserved Ads appear in the most visible ad slot on LinkedIn’s feed and support most Sponsored Content formats, including Video, Single Image, Carousel, Document, Thought Leader, and Event Ads. Advertisers work with their LinkedIn account representative to reserve inventory and pricing.

Why we care. LinkedIn Reserved Ads give you guaranteed top-of-feed placement, increasing visibility, attention, and engagement for campaigns. This premium positioning helps cut through the typical noise in B2B feeds, improving recall and early-funnel impact.

Additionally, the predictable delivery and fixed pricing allow marketers to plan campaigns with more certainty while building higher-quality retargeting audiences for future conversions.

The big picture. LinkedIn is positioning Reserved Ads as a bridge between brand and demand. By anchoring awareness campaigns at the top of the feed, marketers can build higher-quality retargeting pools β€” with LinkedIn reporting up to a 101% lift in mid-funnel engagement when audiences are warmed with Reserved Ads ahead of time.

The bottom line. By turning premium feed placement into a reservable product, LinkedIn is giving B2B marketers a more predictable way to buy attention β€” and convert it into downstream demand.

Google scraps unified pricing rules in Ad Manager after antitrust pressure

17 December 2025 at 21:38
Google Ad Manager

Google has removed its long-standing unified pricing rules in Google Ad Manager, once again allowing publishers to set different price floors for Google demand versus other programmatic buyers.

What changed. Publishers can now set bidder-specific floor prices in Ad Manager. For example, one buyer can be required to bid at least $5 while others compete at a lower $2 floor. Google has also rebranded β€œunified pricing rules” as simply β€œpricing rules.”

The backstory. Before 2019, publishers often set higher floors for Google to counterbalance its data advantage. That flexibility disappeared when Google mandated uniform pricing across exchanges β€” a move later scrutinized by regulators in both the U.S. and Europe.

Why we care. Bidder-specific pricing rules change how auctions clear and how competitive different demand sources are inside Google Ad Manager. As publishers regain the ability to set higher floors for certain buyers, advertisers may see shifts in win rates, CPMs, and available inventory depending on their buying setup. Over time, this could reshape pricing dynamics and push advertisers to reassess bidding strategies and diversification across exchanges.

Regulatory pressure: The rollback follows major antitrust actions against Google’s ad tech business. In the U.S., Google was found guilty of anti-competitive behavior, prompting proposed remedies that included ending unified pricing. In Europe, the European Commission fined Google €2.95 billionΒ ($3.45 billion)Β and ordered the company to end self-preferencing practices across the ad tech supply chain.

What Google says: Google said the change will make it easier for publishers and advertisers to use competing ad tech providers while minimizing disruption. The company framed the update as part of broader near-term product changes across display, video, and app ads.

Industry reaction. Jason Kint, CEO of Digital Content Next, called the move a meaningful β€” if limited β€” win for publishers, noting that unified pricing often lowered yield and that this change offers immediate, tangible relief. He also suggested the update may be designed to show regulatory compliance and head off stronger remedies, including potential divestitures.

The bottom line. After more than six years, publishers are regaining pricing control inside Google Ad Manager β€” a shift driven less by product strategy and more by mounting antitrust pressure on Google’s ad tech empire.

Google Search adds read more links to search result snippets

17 December 2025 at 21:17

Google recently rolled out β€œread more” links in Google search results, which appear at the end of the snippet’s description. When you click on the read more link, you are anchored down to a specific portion of the web page that you clicked on.

Not all search result snippets include these read more links, but many do.

What it looks like. Here is a screenshot of this in action, but you can probably replicate it for most of your queries now:

Google was testing this, or variations of this, Β back in JulyΒ and now it seems to have been rolled out.

Why we care. These read more links do add an additional eye-catching link to the search result snippets. Hopefully, this leads to encouraging more clicks to websites and no less.

More clicks to websites is a good thing, so hopefully this feature will last.

Google adds animation and image editing tools to Merchant Center’s Product Studio

17 December 2025 at 21:04
Google Shopping Ads - Google Ads

Google has expanded Product Studio inside Merchant Center, rolling out three new creative features that go beyond its original image generation tool.

What’s new. In addition to image generation, Product Studio now lets merchants animate static product images into short videos using suggested text prompts, a move aimed squarely at short-form ads and social-style creative.

Google has also added one-click background removal to help isolate products and create cleaner, more consistent Shopping visuals.

The third update increases image resolution, allowing advertisers to upscale older or lower-quality assets to meet modern visual standards.

Why we care. Product imagery plays a major role in Shopping performance, but creating and refreshing assets is often slow and resource-heavy. These updates give merchants more ways to produce high-quality visuals quickly β€” without leaving Merchant Center or relying on design teams.

The big picture. Google continues to embed AI-powered creative tools directly into commerce workflows. By housing animation, editing, and enhancement inside Merchant Center, Google is lowering the barrier to frequent creative testing β€” a key lever for Shopping and Performance Max campaigns.

What to watch. These tools could significantly speed up asset iteration for advertisers with limited creative resources, especially as Google pushes more video-forward and visually rich ad formats across Search, Shopping, and YouTube.

First seen. This update was spotted by Senior PPC Specialist – VojtΔ›ch Audy

Argos launches a final pre-Christmas sale β€” save up to 50% on tech, appliances, toys, and more

Argos has launched one last sale before Christmas, featuring savings of up to 50% across tech, appliances, furniture, toys, and more – I've rounded up the 11 best deals right here.

Urban VPN Proxy is the latest free VPN spying on users – here's how to stay safe

The free Urban VPN Proxy Chrome extension has recently made headlines after being caught silently stealing users' AI chats. Yet, it's not an isolated incident but rather a stark reminder to use only reputable VPN apps.

When Windows 11? Microsoft boosts Windows Server with Native NVMe support

17 December 2025 at 21:50

Microsoft ushers in a new era of Windows Server storage performance with Native NVMe support Microsoft has officially added Native NVMe support to Windows Server 2025, an opt-in feature that can deliver boosted storage performance when using modern NVMe SSD storage solutions. Until now, Microsoft’s storage stack considered all storage devices as SCSI (Small Computer […]

The post When Windows 11? Microsoft boosts Windows Server with Native NVMe support appeared first on OC3D.

Chinese EUV Lithography Machine Prototype Reportedly Undergoing Testing

17 December 2025 at 21:52
According to sources close to Reuters, China has developed a working prototype of an EUV machine, which is currently undergoing testing. This domestic achievement involved Chinese companies successfully reverse-engineering ASML's EUV lithography scanners and creating a functional version using second-hand components. The new EUV prototype reportedly takes up an entire factory floor, comparable in size to modern High-NA EUV machines from ASML. Chinese companies are said to have obtained parts from older ASML machines on secondary markets. The Chinese government initially set a goal of producing working chips from the prototype by 2028.

Leading this initiative is the Chinese technology giant Huawei, which is working to establish a domestic AI supply chain to bypass foreign tech restrictions. In Guanlan, China, Huawei established a comprehensive facility for manufacturing semiconductors using 7 nm technology for its custom processors. Dissatisfied with SMIC's limited output capacity, Huawei has assumed control of the entire silicon production process, from sourcing materials and chemicals to wafer fabrication equipment and chip design. The company is making an unprecedented effort to develop every component of the AI supply chain domestically, from wafer fabrication equipment to model building, and now EUV scanners for more advanced nodes.

Rapidus Explores Glass Panel-Level Packaging to Rival TSMC

17 December 2025 at 21:21
It seems that the Japanese foundry Rapidus is gearing up for strong competition with TSMC. The company has developed a panel-level packaging (PLP) prototype using a glass interposer, with plans to move towards mass production by 2028. At this week's SEMICON Japan 2025 in Tokyo, the prototype will be showcased. It replaces conventional round silicon wafers with large square glass panels, intended for use in high-performance AI packages that combine multi-chip GPUs and more than a dozen of HBM dies. This approach could reduce material waste and enable larger, denser multi-chip assemblies necessary for next-generation AI accelerators, without breaking the traditional laws of 300 mm wafer.

The company is developing 600x600 mm glass panels, which offer a significantly larger usable area than 300 mm silicon wafers and enable the production of more interposers from a single sheet. Glass substrates provide advantages in electrical performance and flatness compared to organic package materials, making them appealing for dense interconnects. However, glass also introduces new manufacturing challenges, such as fragility and potential warpage as panel sizes increase. Intel has notably discontinued its in-house development of glass substrates and has chosen to license the technology to others. The company can thus monetize its innovations, though large-scale Intel-based technology production is not anticipated until after 2030.

AMD's rumored Ryzen 7 9850X3D spotted at Swiss and U.S. retailers β€” listings reveal preliminary price of ~$550-600, significantly higher than 9800X3D's $449 MSRP

The Ryzen 7 9850X3D has been spotted at two retailers with slightly different prices, suggesting it would cost between $550 to $600, which is quickly coming up with CES 2026 right around the corner. These early listings are likely inaccurate and should be taken with a huge grain of salt, especially considering that the 9800X3D launched at a $479 MSRP.

Google unveils β€˜Gemini 3 Flash’ AI model focused on speed and cost

17 December 2025 at 21:15

The post Google unveils β€˜Gemini 3 Flash’ AI model focused on speed and cost appeared first on StartupHub.ai.

Google has recently announced Gemini 3 Flash, a new AI model designed for speed and cost-effectiveness. Deirdre Bosa reported on this development for CNBC, highlighting its strategic importance in the rapidly advancing AI landscape. The Gemini 3 Flash model aims to provide a more accessible and efficient AI solution, distinguishing itself from more resource-intensive models. […]

The post Google unveils β€˜Gemini 3 Flash’ AI model focused on speed and cost appeared first on StartupHub.ai.

Blue Owl Rejects Oracle’s Michigan Data Center Deal Over Unfavorable Economics

17 December 2025 at 20:45

The post Blue Owl Rejects Oracle’s Michigan Data Center Deal Over Unfavorable Economics appeared first on StartupHub.ai.

β€œThe FT story is incorrect. Our development partner, Related Digital, selected the best equity partner from a competitive group of options, which in this instance was not Blue Owl. Final negotiations for their equity deal are moving forward.” This statement from Oracle, provided to CNBC, directly refutes earlier reports that Blue Owl had decided against […]

The post Blue Owl Rejects Oracle’s Michigan Data Center Deal Over Unfavorable Economics appeared first on StartupHub.ai.

Reimagining Public Safety: Technology, Culture, and the Fight Against Crime

17 December 2025 at 20:18

The post Reimagining Public Safety: Technology, Culture, and the Fight Against Crime appeared first on StartupHub.ai.

America faces a startling reality: the chance of a murder being solved is barely a coin flip, with national clearance rates hovering around 47%. This chilling statistic, highlighted by Ben Horowitz, cofounder of a16z, underscores a profound societal failure in crime enforcement, leading to what he terms β€œlost generations.” It’s a crisis demanding more than […]

The post Reimagining Public Safety: Technology, Culture, and the Fight Against Crime appeared first on StartupHub.ai.

Google rolls out Gemini 3 Flash to AI Mode in Search globally

17 December 2025 at 20:40

Google today began rolling out Gemini 3 Flash as the default model powering AI Mode in Search worldwide. The upgrade brings faster performance and stronger reasoning to AI-generated search responses, Google said.

Why we care. With AI Mode, Google continues to transition toward an AI-first search approach. More queries could be answered directly in AI Mode, reducing reliance on traditional organic listings. Improved reasoning allows AI Mode to handle comparison and planning tasks, multi-intent searches, and research-style queries.

What’s changing. Gemini 3 Flash now powers AI Mode in Search globally.

  • It replaces earlier Flash-class models previously used in AI Mode.
  • AI Mode responses now use Gemini 3-level reasoning with lower latency.

Google is also expanding access to Gemini 3 Pro in Search in the U.S.

  • Users can now select β€œThinking with 3 Pro” in the AI Mode model menu for more in-depth help on complex questions, including dynamic visual layouts and interactive tools generated on the fly.

What AI Mode does. According to Google, AI Mode:

  • Breaks complex queries into multiple parts.
  • Pulls real-time information and links from across the web.
  • Presents answers in structured, visually organized formats.
  • Handles multi-step tasks (e.g., trip planning, learning complex topics).

What Google is saying. In a blog post, Tulsee Doshi, senior director, product management, wrote:

Building on the reasoning capabilities of Gemini 3 Pro, AI Mode with Gemini 3 Flash is more powerful at parsing the nuances of your question. It considers each aspect of your query to serve thoughtful, comprehensive responses that are visually digestible β€” pulling real-time local information and helpful links from across the web. The result effectively combines research with immediate action: you get an intelligently organized breakdown alongside specific recommendations β€” at the speed of Search.

This shines when tackling complex goals with multiple considerations like trying to plan a last-minute trip or learning complex educational concepts quickly.

Image generation expands in AI Mode. Google also announced expanded access to Nano Banana Pro, its Gemini 3 Pro–powered image generation and editing model, in Search.

  • More U.S. users can now create and edit images directly in AI Mode by selecting β€œThinking with 3 Pro” and then β€œCreate Images Pro.”
  • Users can add visual explainers, diagrams, and infographics alongside AI-generated answers.

Google’s Danny Sullivan: SEO for AI is still SEO

17 December 2025 at 20:23
Google AI search

Google Search’s Danny Sullivan and John Mueller pushed back again on the idea that brands need a separate AI SEO strategy during the latest Search Off the Record episode.

Sullivan’s point is simple: the acronyms keep changing (GEO, AEO, etc.), but the advice doesn’t: Write for humans, not for ranking systems, whether those systems are traditional search or LLM-powered experiences.

Why we care. As AI search grows, a lot of publishers and SEOs are feeling pressured to try something new. Google’s take: chasing AI tricks can actually backfire and distract you from making content people actually like.

Google says the north star hasn’t moved. Sullivan said Google aims to reward content made for people, not for search algorithms or for LLMs. If you’re already doing that, he said, you’re β€œahead” as formats continue to shift.

  • If you optimize narrowly for a specific AI system, you risk permanent catch-up as those systems evolve.
  • Modern CMS platforms handle much of the old β€œmake your site crawlable” work by default, Mueller added.

Original, authentic, multimodal. Sullivan argued that AI features speed up a reality publishers have faced for years: commodity content is easy to replace. His examples:

  • Pages that padded a simple fact like β€œWhat time is the Super Bowl?” into a long post eventually lost to direct answers.
  • Sites built on predictable, repeatable answers (e.g., word game solutions) are vulnerable when that information is given directly.

What Google wants creators to do:

  • Prioritize original value. Bring perspective, expertise, reporting, firsthand experience, or a voice that only you can provide.
  • Lean into authenticity. Not β€œmanufactured authentic,” but work grounded in real experience.
  • Go multimodal. Sullivan joked that he hates the term, but the point stands. Mix text with images and video, because users search across formats and often prefer video for how-to answers.

Structured data still matters. They also said structured data helps, but it isn’t decisive. Sullivan said it’s not β€œstructured data and you win AI.” It simply supports how systems understand and present content, just as it already does across Search features.

Focus on quality clicks. Google is seeing that traffic from AI formats can arrive more engaged, such as spending more time on-site. His hypothesis is that AI results create better contextual awareness. Users click when they are more confident that the result matches their intent.

  • Google’s advice: define and track outcomes that matter to your business, not just raw traffic.
  • Clicks alone don’t tell the full story anymore – especially as AI Overviews and conversational results guide users before they ever visit a site.
  • Focus on quality clicks and quality conversions over volume (and be clear on what a conversion actually is).
  • Sullivan noted that everyone defines β€œconversion” differently, which makes it hard for Google to surface that kind of value inside Search Console.

About query fan-out. They explained why β€œI rank in blue links but not in AI Overviews” is a flawed comparison:

  • AI features may run multiple related searches behind the scenes. Mueller described it as doing β€œa whole bunch of searches for you” and then synthesizing the results.
  • That means visibility in AI results may not map one-to-one with the exact query a user typed.

Clients still want β€œthe new thing.” Sullivan acknowledged the real-world challenge: Clients still demand β€œAI optimization” as a separate service.

  • He suggested reframing is to present the β€œsame old stuff” as the durable, long-term strategy.
  • Position β€œAI SEO” as monitoring and adapting, not rebuilding everything into a second content system.
  • Sullivan said Generative Engine Optimization (GEO) isn’t separate from SEO – it’s a subset of it. SEO has always been about understanding how people look for information and how systems surface it.
  • Optimizing for AI answers is conceptually no different from optimizing for local results, voice search, or other formats. The fundamentals still apply.

What to do now, according to Google. Based on the conversation, Google’s β€œSEO checklist” looks something like this:

  • Create human-first, satisfying content.
  • Offer original reporting, unique expertise, firsthand experience, and a strong voice.
  • Add images or video when they genuinely improve understanding.
  • Use structured data where appropriate.
  • Optimize for engagement and conversions, not just clicks.

The podcast. Thoughts on SEO & SEO for AI, part 1

Dig deeper:

πŸ’Ύ

AI search is not changing SEO fundamentals. Google's Danny Sullivan says to keep doing things that will make you successful in the long term.

APT28 Targets Ukrainian UKR-net Users in Long-Running Credential Phishing Campaign

The Russian state-sponsored threat actor known as APT28 has been attributed to what has been described as a "sustained" credential-harvesting campaign targeting users of UKR[.]net, a webmail and news service popular in Ukraine. The activity, observed by Recorded Future's Insikt Group between June 2024 and April 2025, builds upon prior findings from the cybersecurity company in May 2024 that

New ForumTroll Phishing Attacks Target Russian Scholars Using Fake eLibrary Emails

The threat actor linked to Operation ForumTroll has been attributed to a fresh set of phishing attacks targeting individuals within Russia, according to Kaspersky. The Russian cybersecurity vendor said it detected the new activity in October 2025. The origins of the threat actor are presently unknown. "While the spring cyberattacks focused on organizations, the fall campaign honed in on

Creating apps like Signal or WhatsApp could be 'hostile activity,' claims UK watchdog

An independent review of the newly implemented National Security Act suggests that developers of encrypted messaging apps may be considered hostile actors. This, as lawmakers' pressure against encryption keeps growing.

Google Updates JavaScript SEO Docs With Canonical Advice

17 December 2025 at 18:56

Google updated its JavaScript SEO docs with new guidance on canonical URLs for JavaScript-rendered pages. Keep canonicals consistent before and after rendering.

The post Google Updates JavaScript SEO Docs With Canonical Advice appeared first on Search Engine Journal.

Cable-free gaming power! Sapphire RX 9070 XT Phantom Link GPU unveiled

17 December 2025 at 18:41

Sapphire’s Radeon RX 9070 XT Phantom Link has been unboxed in China Sapphire has created a new RDNA 4 flagship, the Radeon RX 9070 XT Nitro+ Phantom Link, which has now been unboxed in China. This GPU is an enhanced version of Sapphire’s existing Nitro+ model, offering users both 12V-2Γ—6 and GC-HPWR power inputs. This […]

The post Cable-free gaming power! Sapphire RX 9070 XT Phantom Link GPU unveiled appeared first on OC3D.

(PR) Hunted Cow Games Announces Fallen Sword II for 2026

17 December 2025 at 18:49
Hunted Cow Games, the Scottish indie studio behind one of the longest-running browser MMORPGs in gaming history, today announced Fallen Sword II, a cross-platform RPG inspired by their cult classic that has captivated millions of players since 2006. The game is set for release in Q3 2026 on PC and mobile platforms, and is available to wishlist now on Steam.

A Legacy Forged in Scotland
When Hunted Cow launched Fallen Sword from their tiny Elgin office in December 2006, the team could never have predicted its meteoric rise, gaining over 250,000 players in just three months. By the end of its first year, over one million adventurers had signed up to explore the world of Erildath.

Senators probe whether AI data centers are driving up electricity costs

17 December 2025 at 19:02

In formal letters sent to seven tech firms – Amazon, Google, Meta, Microsoft, Equinix, Digital Realty, and CoreWeave – Senators Elizabeth Warren of Massachusetts, Chris Van Hollen of Maryland, and Richard Blumenthal of Connecticut raised concerns that data center growth has coincided with dramatic increases in local utility rates. They...

Read Entire Article

AI Engage: Train AI Search Engines – Get Cited in Google AI Search, ChatGPT, Perplexity & Copilot


AI Engage introduces a new way to win in AI Search. Instead of optimizing for keywords, it systematically educates AI search engines about your brand by prompting models to fetch and analyze your real content. Automated campaigns engage Google AI Search, ChatGPT Search, Perplexity, and Microsoft Copilot using realistic user queries.

Campaigns run in six languages across 150 million geo targeted IPs, with analytics to track visibility and performance by market.

View startup

Rapidus explores panel-level packaging on glass substrates for next-generation processors β€” aggressive plan would help it leapfrog rivals

Rapidus plans to outline its early-stage work on panel-level packaging using 600 Γ— 600 mm glass substrates at SEMICON Japan, highlighting an aggressive plan to leapfrog rivals by combining glass-core substrates and PLP for future AI and HPC chiplet packages.

Engineer turns E-ink tablet into computer monitor in Linux β€” perfect secondary reading screen to reduce eye strain over the network

17 December 2025 at 18:00
A software engineer and E-ink enthusiast recently set up a remote E-ink secondary display, upcycling an old E-ink tablet and setting up an awesome Linux DIY project for all those with the need to read on bespoke screen hardware.

ARC-AGI: The True Measure of Machine Intelligence Beyond Brute Force

17 December 2025 at 19:16

The post ARC-AGI: The True Measure of Machine Intelligence Beyond Brute Force appeared first on StartupHub.ai.

β€œIntelligence is measured by the efficiency of skill acquisition on unknown tasks.” This foundational insight, articulated by FranΓ§ois Chollet, creator of Keras and the Abstract and Reasoning Corpus for Artificial General Intelligence (ARC-AGI), underpins a critical shift in how the AI community evaluates progress. In a recent interview at NeurIPS 2025, Y Combinator General Partner […]

The post ARC-AGI: The True Measure of Machine Intelligence Beyond Brute Force appeared first on StartupHub.ai.

Wall Street’s New Hires are Grok, Claude, and GPT

17 December 2025 at 18:14

The post Wall Street’s New Hires are Grok, Claude, and GPT appeared first on StartupHub.ai.

The next wave in finance is foundation models that can actually invest, not just talk about markets. As of today, agentic finance is about foundation models quietly becoming the core infrastructure for how capital decisions are researched, prepared, and executed. These systems are shifting the paradigm from who can hire the most analysts to who […]

The post Wall Street’s New Hires are Grok, Claude, and GPT appeared first on StartupHub.ai.

Google clarifies canonicalization with JavaScript

17 December 2025 at 19:12

Google updated its JavaScript SEO best practices document, for the second time this week, this time to clarify canonicalization best practices for JavaScript. In short, Google said β€œsetting the canonical URL to the same URL as in the original HTML or if that isn’t possible, to leave the canonical URL out of the original HTML.”

What Google added. Google added a new section over here and it reads:

β€œThe rel=”canonical” link tag helps Google find the canonical version of a page. You can use JavaScript to set the canonical URL, but keep in mind that you shouldn’t use JavaScript to change the canonical URL to something else than the URL you specified as the canonical URL in the original HTML. The best way to set the canonical URL is to use HTML, but if you have to use JavaScript, make sure that you always set the canonical URL to the same value as the original HTML. If you can’t set the canonical URL in the HTML, then you can use JavaScript to set the canonical URL and leave it out of the original HTML.”

Google on noindex. Google also warned about using JavaScript for noindex tags earlier this week. Google said β€œyou do want the page indexed, don’t use a noindex tag in the original page code.”

Why we care. So if you use JavaScript for setting a canonical link, make sure to also check in Google Search Console’s URL Inspection tool if it is being picked up.

Review these updated best practices if you use JavaScript on your site, especially for canonical links.

How to use Google’s Channel Performance report for PMax campaigns

17 December 2025 at 18:00
How to use Google’s Channel Performance report for PMax campaigns

For years, PPC advertisers have considered Performance Max (and Smart Shopping before it) to be a black box, even a black hole.

While its powerful automation drives convincing results, the lack of transparency into channel performance has been a persistent frustration.Β 

Now, Google is beginning to provide some answers.Β 

The rollout of the new Channel Performance report marks a significant step toward the transparency advertisers have been demanding.Β 

This guide explains what the report is, highlights its strengths and weaknesses, and shows you how to use it.

What is the Channel Performance report – and why is it a big deal?

The Channel Performance report is essentially a pre-built network report (we can discuss the semantics of channel versus network another day), which can be found under Campaigns > Insights and Reports > Channel Performance (beta).

It offers tabular network data and an interactive flow diagram from impressions down through conversions.Β 

The Channel Performance report only works for Performance Max campaigns. However, credible clues suggest that this report may support additional campaign types in the future.

This is important because, while Performance Max is (in)famously a β€œchannel soup,” all campaign types are capable of serving across different ad networks within Google’s grasp, and many of them do so by default.

Previously, untangling this mix to see which channels were actually performing was a task left to manual reports or, in the case of PMax, third-party scripts based on guesswork.

The Channel Performance report is Google’s native solution.Β 

A tour of the Channel Performance report

The report is composed of two main elements:Β 

  • An account-level view that offers a compact summary of each campaign’s channel data (plus some hidden features).
  • A campaign-level view that offers a neat but, in my opinion, deeply flawed Sankey diagram, and another data table, more detailed than at the account level.Β 

Furthermore, there are various customization options, which can be saved as preferred views, and multiple export options.

1. The account-level overview: Channel data in the palm of your hand

The account view is a newer addition to the Channel Performance report, and in some ways my favorite view.Β 

Previously, when you accessed this report, you’d land on a blank page prompting you to select an individual Performance Max campaign.Β 

Now, this handy table is the first thing you’ll see.

SEL_pmax-channel-performance-report_asset-6

It has a series of rows for each campaign, nested rows for each channel, and columns for the performance metrics.Β 

One thing I love is that each nested row has the channel icon next to it.Β 

Tabular data can sometimes make my eyes cross, but this simple visual aid makes the data much easier to skim.

By default, the campaign rows are sorted alphabetically, and you’ll likely want to sort by something more practical, like impressions, costs, revenue, etc.

After that, you can really leap down the page easily, comparing the distribution of your key campaigns.

But that’s the obvious part.

My top tip for this view is that you can change your segment, and among the options, two really stand out for me:Β 

  • Ads using product data.
  • Ad event type (under Segment > Conversions).
SEL_pmax-channel-performance-report_asset-4

The first allows you to see the volume and performance of β€œads using product data” (feed-based ads) versus β€œads not using product data” (asset-based ads).

Yes, that’s right, finally a simple comparison of feed ads and asset ads. Besides network performance, this has been one of the most contentious and least transparent areas in PMax, prompting numerous advertisers to run so-called β€œfeed-only” PMax campaigns.

Now you can easily see what’s going on with this performance facet across all your PMax campaigns, plus an account-level summary row at the bottom.Β 

Whether you like or dislike what you’re seeing, you can head over to your asset-group-level and asset-level reporting to dig deeper.Β 

Be cautious when judging the performance of asset-based ads. They should not be held to the same efficiency standards.

The second segment, ad event type, might sound non-descript, but it’s really important.

It lets you easily understand the volume and performance of your click-through versus view-through conversions.Β 

This has been (yet another) divisive topic in PMax:Β 

  • Do view-based conversions belong mixed together with standard conversions?Β 
  • Does this inflate performance?Β 

Now you can answer these questions per campaign and also at the account view in the summary row.

But what if you want even more detail?Β 

What if, for example, you want to learn your feed versus asset share in, say, YouTube specifically?Β 

That’s not possible at the account level, but it certainly is at the campaign level.

Just click on any campaign and it will load a new page drilling down to the next reporting level.Β 

2. The campaign-level view: Data visualization and detailed analysis

The first thing you’ll notice on this page is the large Sankey diagram.Β 

It’s visually striking and has become a signature of the Channel Performance report.

That said, we need to set it aside for now. Scroll down to the data table below, which is similar to the one you just saw.

The campaign data table: A deeper dive

While the Sankey diagram gives a high-level view, the table below is where real analysis happens.Β 

It’s more reliable for decision-making because it shows the raw numbers without visual distortion.

The table breaks performance down by channel and ad type – the feed-based versus asset-based split we discussed earlier.Β 

For each segment, you can review multiple metrics by default, but my top tip is to go to Columns > Conversions.

There, you can select Conv. value / Cost (a.k.a. ROAS) and Cost / Conv. (a.k.a. CPA).Β 

These are hidden by default, but you can indeed see them, and I don’t think I have to tell you why they are interesting to know.

SEL_pmax-channel-performance-report_asset-1

Crucially, the table also includes an export function, plus scheduling options, allowing you to pull the raw data for deeper analysis in a spreadsheet.

The Sankey diagram: Visualizing the flow

As noted earlier, this visualization – officially called the Channels-to-Goals chart – is visually striking, but it has limitations.Β 

Before addressing those issues, let’s clarify its purpose and what it can tell us.

The Sankey diagram presents a visual breakdown of performance across the channels within your PMax campaign.Β 

SEL_pmax-channel-performance-report_asset-5

It maps the customer journey within your campaign – how users move from seeing an ad (impressions) to clicking or engaging with it (interactions), and, ultimately, to converting (results or conversions).

This is great. For the first time, advertisers can see the flow of core funnel metrics right in Google Ads, all segmented by the specific channel driving the traffic.Β 

This allows you to understand how PMax allocates your budget and which parts of its vast inventory are actually working for you.

Decoding the channels

People often look at the Sankey and get stuck. β€œWhere’s my Shopping data?” is probably the single biggest example of this.Β 

As we’ve discussed, a key feature of the report is how it segments ads into feed-based and asset-based ads.

When we combine that dimension with the network or β€œchannel” dimension, we can translate the labels into more familiar terms:

  • Search
    • Ads using product data: These are your Shopping ads.
    • Other ads: This represents your Dynamic Search Ads (DSA) and Responsive Search Ads (RSA) traffic.
  • Display
    • Ads using product data: These are Dynamic Product Ads, which in my assessment is likely a lot of Dynamic Remarketing and some Dynamic Prospecting.
    • Other ads: These are your standard Responsive Display ads.
SEL_pmax-channel-performance-report_asset-3

These are my interpretations of the data, which might not be perfect.Β 

It would be extremely helpful if Google offered more detailed documentation on what’s included.

For example, feed-based YouTube ads can comprise a variety of formats and placements, some of which, such as β€œGMC Image Shorts,” are not documented anywhere.

Google’s guidance is quite vague.

Get the newsletter search marketers rely on.


The limitations of the native report

While a welcome addition, the report has some shortcomings.

The misleading Sankey diagramΒ 

The visual proportions of the diagram are not based on volume, which makes it extremely misleading at a glance.Β 

A channel that appears to drive significant traffic may actually account for only a tiny share of your impressions.

In the example below, the asset-based Search ads segment appears to have a couple hundred thousand impressions, but in reality only has 4,500 impressions.Β 

This makes the chart almost useless for quick, accurate analysis, which is the entire point of data visualization.

SEL_pmax-channel-performance-report_asset-3

The lack of ratios in the data tableΒ 

The data table provides useful raw data, but it lacks key calculated metrics needed for analysis, such as conversion rate and cost per click.

To see the full picture, you must export the data and do your own calculations.

This feels, to be honest, a bit petty of Google.Β 

They could easily add these columns, but it seems they would prefer not to. Grab your calculator.

How to make the most of the report

Despite its limitations, you can still extract valuable insights into which channels deliver what.

The key is to focus on asset quality and traffic quality, because direct channel control is limited.

Analyze placement data for quality controlΒ 

While the report doesn’t let you directly control channel mix, it helps you monitor traffic quality.Β 

Use the placement reports to see exactly where your Display and YouTube ads are showing.

  • Export this data into Google Sheets. Note that, frustratingly, it only contains impression data.
  • Use built-in functions like =GOOGLETRANSLATE() to understand foreign-language placements and the integrated =AI() function to help categorize domains and videos for brand safety.
  • Exclude low-quality or irrelevant placements or content at the account level, prioritizing bad placements that are higher in volume.

Build your own Sheets-based reporting or try scripts

Google has confirmed that API access and MCC-level reporting are coming to the Channel Performance report.Β I also expect this data to be supported in the Report Editor.Β 

In the meantime, you can export the report as a .csv or send it directly to Google Sheets.

With a smart setup, these exports enable you to calculate custom metrics, build charts, apply heatmaps, and reshape the data as needed.

To help the community, I helped build a script that enhances Google’s report in several practical ways:

  • Adds key metrics like conversion rate, CTR, CPC, CPM, and more.
  • Applies clear, common-sense labels such as β€œShopping” and β€œResponsive Display.”
  • Includes charts with proportional visuals for more accurate interpretation.
  • Cleans and parses columns to remove friction.

The script works for individual PMax campaigns, not the account-level view. I’m waiting for Google’s feature set and scripting options to stabilize before expanding the script.

What’s next for PMax reporting?

We know Search Partner data is coming, along with API access, MCC-level reporting, and likely support for additional campaign types such as Demand Gen.

It’s encouraging to see Google share this level of detail, and there’s reason to believe this momentum will continue.Β 

The Channel Performance report already addresses one of the most persistent criticisms of Performance Max – that it operates as a black box.Β 

Three years ago, it would have been hard to imagine Google responding to advertiser feedback at this scale, particularly on transparency.

Still, better visibility doesn’t automatically translate into better decisions.Β 

Interpreting this data correctly takes time, context, and careful analysis – and that work remains firmly in the hands of advertisers.

SEL_pmax-channel-performance-report_asset-7

Gamers report huge gains with Monster Hunter Wilds PC update

17 December 2025 at 17:03

Monster Hunter Wilds newest PC update is a godsend to 8GB GPU users Monster Hunter Wilds has just received β€œTitle Update 4”, the first of a trio of planned PC optimisation updates for the game. Following the release of this update, PC gamers have reported significant improvements to the game. This includes lowered VRAM usage […]

The post Gamers report huge gains with Monster Hunter Wilds PC update appeared first on OC3D.

Don't delete your Xbox games to make room for the 2026 lineup β€” Seagate's cards now come with a Christmas discount that fixes the AAA storage crunch

17 December 2025 at 17:59
Christmas has come early for Xbox players as retailers have acquired discount deals on the Seagate Storage Expansion Card for Xbox Series X|S to upgrade their storage space for 2026's upcoming Xbox titles.

AMD Unveils Radeon RX 9060 XT Low-Power Graphics Card

17 December 2025 at 18:21
AMD has quietly introduced a new RDNA 4 SKU targeting lower overall GPU power consumption. According to the updated website listing, AMD has released the Radeon RX 9060 XT Low-Power GPU SKU, designed to fit within a reduced TDP envelope. With a TDP of 140 W, AMD recommends a minimum power supply of 450 W. The card features 32 Compute Units (CUs), 2,048 Stream Processors (SPs), 32 ray tracing accelerators, 64 AI accelerators, 128 texture units, and 64 ROPs, all on a die with 29.7 billion transistors. This is 20 W lower than the standard Radeon RX 9060 XT SKU, which also has 32 CUs and 2,048 SPs but operates at a 160 W TDP.

Memory and I/O specifications remain consistent for this class. The RX 9060 XT LP comes with up to 16 GB of GDDR6 on a 128-bit memory interface, supported by 32 MB of AMD Infinity Cache, and offers an advertised memory speed of up to 20 Gbps for a peak bandwidth of up to 320 GB per second. Display outputs include DisplayPort 2.1a and HDMI 2.1b. The card supports modern codecs and formats, including AV1 encode and decode, H.264 and H.265 encode and decode, and 4K HDMI support.

(PR) MicroProse and Geoff Crammond Reunite to Revive a Racing Legend

17 December 2025 at 18:10
MicroProse is excited to announce its renewed partnership with legendary game developer Geoff Crammond, creator of some of the most influential racing simulation games of the 1990s and early 2000s. Together, we are bringing back his groundbreaking racing game engine in a re-release of the original game technology, enhanced with updates, new tools, and Steam Workshop support, through the upcoming Geoff Crammond Racing series: GCR1, GCR2, GCR3, and GCR4, scheduled to release in 2026.

These titles are being reintroduced under new names to reflect our independent direction and to ensure full separation from any official motorsport licensing.
While the names are new, the core simulation engine, physics systems, and pioneering gameplay features developed by Geoff Crammond remain at the heart of each release.

NVIDIA Plans to Reduce RTX 50 Production by Up to 40% in Early 2026

17 December 2025 at 16:49
Reports from Asian supply chain sources suggest that NVIDIA might cut back on making GeForce RTX 50 series GPUs in the first six months of 2026, Videocardz reports. According to Board Channels, overall supply during H1 2026 could be down by roughly 30-40% compared to the same period in 2025, although the information remains unconfirmed. The report points out that the issue is not entirely related to GPU VRAM alone, but also to tight availability across GDDR6, GDDR7, and other memory components, including motherboard-related memory such as DDR5/DDR4. Board Channels also claims NVIDIA may adjust allocation strategies for add-in board (AIC) partners in Mainland China to better align supply with changing DIY market conditions.

Benchlife appears to support parts of the rumor, though without specifying a percentage reduction. Instead, it reports that two models (best value models in the Blackwell RTX 50 lineup), the GeForce RTX 5070 Ti and RTX 5060 Ti 16 GB are likely to see the earliest supply adjustments. While both Board Channels and Benchlife have long and good track records, the reports are based on regional sources and may not reflect global supply conditions. We will have to wait for NVIDIA or its partners to provide more information. However, it's worth noting that just a month ago, reports from Korean and Taiwanese outlets said AMD and NVIDIA were already considering GPU output cuts, while ASUS and other vendors were slowing motherboard plans due to memory shortages. In the end, this could end up being a broader, global production adjustment by NVIDIA rather than a region specific move.

Intel installs industry's first commercial High-NA EUV lithography tool β€” ASML Twinscan EXE:5200B sets the stage for 14A

Intel has installed and qualified ASML's TWINSCAN EXE:5200B, the first High-NA EUV lithography tool designed for commercial production, reiterating Intel's plans to use High-NA EUV patterning for 14A process technology and onwards.

How conversational AI is changing the economics of paid search

17 December 2025 at 17:00
How Microsoft Copilot turns conversations into richer searches – and higher-ROI ads

Microsoft Copilot is transforming search advertising by turning everyday conversations into intent-rich signals advertisers can act on.Β 

ROAS increases 13-fold when users engage with Copilot before performing a search, according to Microsoft.

Drawing from billions of first-party audience insights across Microsoft’s consumer ecosystem – including Bing, Edge, Xbox, LinkedIn, and Activision – Copilot identifies high-value audiences using deterministic data built from search intent, web activity, and profile information.Β 

This allows advertisers to reduce wasted impressions and stretch budgets further.

The mechanics of intent-rich searchΒ 

The core proposition of conversational search is that users provide significantly more context to a chatbot than to a traditional search bar.Β 

Instead of a fragmented keyword, users are increasingly asking detailed questions.

When a user submits a complex query – such as asking for specific product comparisons or local recommendations – the AI triggers multiple backend searches across reviews, specifications, and availability to construct an answer.Β 

For the advertising industry, this behavior change offers a potential goldmine of data.Β 

By interpreting these longer queries, platforms can identify β€œhigh-intent” buyers more accurately, turning a single conversation into multiple, precise ad opportunities.

Applying conversational intent to a real-world campaign

To understand how these metrics translate into strategy, consider a recent test I conducted for a well-known California-based university tasked with recruiting high school seniors for their hands-on engineering and architecture STEM programs.

The challenge

The university historically relied on broad keywords like β€œbest engineering schools.”

This resulted in high competition and wasted spend on students looking for art programs or out-of-state options they couldn’t afford.

The conversational approach

Using Copilot’s intent signals, the campaign shifts.

A prospective student might ask Copilot:

  • β€œFind me a university with a strong robotics program, under $30,000 tuition, located on the West Coast.”

The results

Applying Microsoft’s reported benchmarks to this scenario reveals significant efficiency gains:

  • Slashed waste: The university realizes a 32% reduction in wasted impressions because ads are not shown to students whose conversational context indicates irrelevant intent.
  • Budget efficiency: By targeting intent rather than broad volume, the campaign drives a 48% decrease in cost per acquisition (CPA) compared to search alone.
  • Higher engagement: Because the ad appears as a helpful solution to a specific question, engagement lifts by 153%.

Action plan: Transitioning to intent-based advertisingΒ 

For advertisers seeking to replicate these results, the shift necessitates more than simply enabling a new setting.Β 

It requires a strategic overhaul of how campaigns are structured to capture β€œconversational” demand.

Phase 1: Foundation and data (The signal layer)

Audit service offerings and solution data

Ensure your site’s structured data is rich with details on specific methodologies and industry specializations.Β 

AI assistants rely on this semantic depth to answer prospective queries about β€œcompetency, case studies, and communication options.”

Prioritize first-party data

Integrate customer data to train the model.Β 

Microsoft’s ecosystem leverages data points from LinkedIn to Xbox to refine targeting.

Advertisers must supply their own truth data to match this precision.

Phase 2: Campaign structure (The capture layer)

Embrace long-tail queries

Move away from strict exact-match keywords.

The UI overhaul of Copilot encourages users to ask β€œlonger, more detailed questions,” meaning broad match modifiers are necessary to capture these natural language phrases.

Optimize for answers, not just clicks

Structure landing page content to answer specific questions.Β 

Since Copilot acts as a β€œcompanion” guiding users through tasks, your ad content must align with helping them make a decision, not just selling a product.

Phase 3: Cross-channel integration (The scale layer)

Implement cross-device strategy

With 90% of Gen Z adults in the U.S. using the web while watching TV, campaigns must run across multiple platforms, including mobile, PC, and console, to capture their split focus.

Bridge the authenticity gap

For younger demographics, leverage integrations like Snapchat’s My AI.Β 

This places ads within β€œconversational flows” rather than interrupting them, a key factor in engaging Gen Z.

Get the newsletter search marketers rely on.


The Gen Z challenge: Authenticity vs. algorithmsΒ 

Bridging the gap with Gen Z remains a hurdle for most ad platforms, which often struggle with perceptions of inauthenticity.Β 

To address this, the industry is seeing a trend toward utilizing behavioral data from unlikely sources.Β 

By layering in data from gaming ecosystems like Activision, advertisers can target based on real behaviors – from play styles to in-game purchases – ensuring campaigns feel relevant rather than generic.

To legitimize whether Copilot is effectively targeting Gen Z – or just efficiently automating ad delivery – we must look beyond corporate claims.Β 

Microsoft’s strategy relies on a β€œclosed loop” of gaming data, social integration, and conversational signals.

Does this actually work for a generation that is famously resistant to traditional advertising?Β 

The answer lies in the tension between utility and authenticity.

The behavioral match

Microsoft’s claim that Copilot β€œcracks the code” is mechanically sound because it aligns with how Gen Z actually searches.

The shift from keywords to conversation

Data shows that Gen Z users write the longest search queries (avg. 5.83 words) and are the most likely to use complete sentences.

They treat search engines like companions, asking β€œWhat is the best…” rather than typing β€œbest shoes NYC.”

Legitimacy verdict: High. Copilot isn’t trying to force a behavior change. It is capitalizing on one that already exists.

By decoding these long, conversational queries, Microsoft captures intent often missed by a keyword approach alone.

β€˜Gaming data’ targetingΒ 

Using Activision data to target users based on β€œplay styles” is a strong differentiator for Microsoft.

The reality: 90% of Gen Z second-screens (uses a phone while watching/playing on another screen). Traditional demographics (e.g., β€œMale, 18-24”) are failing because they are too broad.

The legitimacy test: Targeting a user because they play Overwatch (identifying them as team-oriented and strategic) vs. Call of Duty (identifying them as reactive and fast-paced) allows for psychographic targeting that feels β€œrelevant” rather than β€œintrusive.”

The risk is that there is a fine line between β€œrelevant” and β€œstalker-ish.” 

While Microsoft’s targeting is effective, 76% of Gen Z actively avoid ads, and privacy concerns are their top barrier to trusting AI platforms.Β 

That said, the success of this strategy hinges on the ads feeling native to the experience, not like data extraction.

The authenticity paradox

This is the weak point in the strategy. Microsoft claims Copilot helps bridge the β€œauthenticity gap,” but Gen Z is inherently skeptical of AI-generated content.

The conflict: Studies show that Gen Z can easily identify AI-generated ads and often labels them as β€œannoying” or β€œboring” compared to human-created content.

The Snap integration: Embedding Copilot ads into Snapchat’s β€œMy AI” is a double-edged sword. While it places ads in a trusted social space, it risks polluting a private sanctuary.Β 

If β€œMy AI” starts feeling like a corporate shill, users may abandon the feature entirely.

Legitimacy is mixed. The placement is correct (Snapchat, Games), but the content is at risk.Β 

If advertisers use Copilot to auto-generate generic ad copy, it will fail. Success requires using the AI for targeting but keeping the creative 100% human.

The verdict: Is Microsoft effectively targeting Gen Z?

  • Technically: Yes, they have successfully built a mousetrap that catches Gen Z where they live (gaming, social, conversational search).
  • Culturally: To be determined. The efficiency is there (lower CPA, higher ROAS), but β€œlegitimizing” the strategy long-term requires overcoming the β€œuncanny valley” of AI trust.

Dig deeper: How Gen Z is redefining discovery on TikTok, Pinterest, and beyond

A new economic realityΒ 

The narrative from platforms like Microsoft Copilot is that AI-driven targeting creates a β€œclosed loop” where better engagement drives cost savings.Β 

As conversational AI reshapes how consumers interact with the web, advertising platforms are racing to translate natural language questions into actionable intent.Β 

Microsoft’s Copilot serves as a prime case study of this shift, demonstrating how emerging assistants generate richer, multi-step queries that potentially reshape search economics from a volume game to one of precision.

For advertisers, this signals a fundamental transition: moving away from the broad β€œspray and pray” tactics of keyword volume toward a model where conversational signals drive ROAS.

Dig deeper: The future of remarketing? Microsoft bets on impressions, not clicks

4 marketing problems AI can actually solve right now by Artlist.io

17 December 2025 at 16:00

Why this matters now

Marketing budgets in 2025 have stayed the same, yet expectations keep rising. CMOs report budgets stuck at roughly 7.7% of company revenue, which means teams are expected to do more with the same dollars. In that context, the most practical use of AI is not a moonshot, but a set of clear fixes to everyday bottlenecks that slow teams down and drive costs up.

This article breaks down four problems that marketers face right now and how AI is already solving them. The difference today is that Artlist AI, including image, video and voice generators, turns AI from a novelty into a reliable production system. When you use AI to streamline your workflow instead of chasing hype, you ship more creative, stay on brand and make decisions based on real performance data.Β 

1) Rising video costs and shrinking timelines

The problem: Video is still one of the most effective formats in a marketer’s toolkit, but teams feel the squeeze. Shorter formats dominate social feeds, content calendars never stopΒ  and production bottlenecks turn into budget overruns. Teams need more output in less time.Β 

What’s working: Marketers still see strong returns from video. Wyzowl’s 2024 study reports 90% of marketers say video delivers a good ROI, with 30–60 seconds rated the most effective length, perfect for social placements and paid tests. That supports a strategy shift that marketers need to ship more short pieces, produced in cycles measured in days instead of weeks.

How AI helps:

This is exactly where Artlist AI leads. It helps teams to finish videos in hours not weeks, giving you more room to test, refine, and scale video output without sacrificing quality

  • Script to screen speed. Artlist AI storyboards, image generation and AI video generation help teams audition more concepts in less time, then move proven ideas into full production.
  • On-the-fly variations. Once a master edit is locked, AI tools can generate multiple aspect ratios and quick alternates for A/B testing without re-editing from scratch.
  • Voiceover without the studio. High-quality Artlist AI voiceover makes late-stage copy changes effortless, eliminating the need for booth bookings or talent scheduling. And brands can easily keep tone and pacing consistent.

Klarna recently publicly quantified its savings: about $10 million annually tied to AI in marketing, including a $6 million reduction in image production costs and much faster iteration cycles. While every team’s baseline differs, the directional takeaway is strong and indicates that small time wins across dozens of workstreams add up to real money.Β 

2) Inconsistent brand voice across markets and channels

The problem: Global campaigns require many voices, languages and platform variations. Human recording sessions can create drift in tone and pacing, and late edits become expensive.

What’s working: Studio-grade text to speech models and voice cloning technology now produce narration that is indistinguishable from a human voice, even when using headphones. This makes versioning practical at scale while keeping quality consistent across dozens of outputs.

How AI helps:

Artlist’s AI voiceover gives you one brand voice you can trust, every time, across every market.Β 

  • Stable tone across languages. A single brand voice can be replicated across scripts and regions, then fine-tuned for pacing, warmth, and energy. This is made easy with Artlist’s voice cloning models.
  • Rapid revisions. Late copy changes are possible at sensible price points, covered by legal licenses and mean details such as new promotion dates can be updated in minutes, not weeks, with less stress on marketers.
  • Accessible variants. AI voice and caption pipelines support business localization efforts and accessible workflows without extra studio time. Global campaigns often stall when localized content is linguistically inconsistent. Translation alone doesn’t guarantee cultural fit, and recreating dozens of regional versions drains budgets. AI translation and dubbing tools are closing the gap between literal accuracy, cultural fluency, and still retain the emotion marketers aim for.Β 

airBaltic, the national airline of Latvia, uses Artlist’s AI voiceover to speed production and experiment with tone and pacing, reporting that work that used to take many hours now moves much faster, with tighter control over fit and finish before publication. For a team managing constant route and fare updates, shaving hours off every revision adds meaningful capacity.

3) Creative testing at the speed of social

The problem: Marketers know more than most how feeds change daily. What worked last quarter may stall today. Marketers need more creative swings, which means more thumbnails, cuts, and captions, all without blowing the budget.

What’s working: Recent data points to one clear advantage: brands that test creative variations more frequently outperform those that don’t. A 2024 Nielsen study found that campaigns using three or more creative versions improved ad recall by up to 32%, while those refreshing assets monthly saw 17% higher click-through rates than static campaigns.Β 

AI tools now make A/B testing much easier. Whether the changes are big or small, they are much less taxing, and keeping up with the increased cadence is possible by producing and refining short-form assets in hours instead of days. AI tools like video generators allow marketers to generate alternate visuals, swap voiceovers or localize content without requiring new studio sessions.

In 2023, Coca-Cola invited consumers to produce artwork and short videos using AI trained on its licensed brand assets. Within the first week, participants generated over 100,000 original pieces, driving more than 30% higher digital engagement that quarter. Internally, the company’s marketing team analyzed those submissions to understand which visuals and tones drew the strongest responses. That feedback reshaped future campaign planning, trimming production time and improving message precision.

How AI helps:

Artlist AI lets you scale creative volume without scaling your budget, so you can test and learn faster.Β 

  • Images and thumbnails at scale. Rapid asset generation means marketers can refresh visuals for each social cycle without starting from zero.
  • Micro-edits for micro-audiences. Teams can test small creative differences β€” intro clips, CTAs or captions β€” against audience segments and measure results quickly.
  • Faster learning loops. Instead of waiting for quarterly reports, marketers can identify high-performing creative in real time and reallocate spend to proven variants.

Creative volume matters less than creative velocity. When teams can produce, test, and iterate at social speed, they turn marketing from a guessing game into a measured system of learning.

4) Measuring creative impact with real feedback loops

The problem: Marketers still rely heavily on vanity metrics, for example, views, likes and impressions, but they say little about actual persuasion. Traditional testing cycles are slow, and connecting creative choices to downstream results is often guesswork.

What’s working: AI analytics tools can now correlate creative elements like color palettes, pacing, tone or voice style, with engagement and conversion metrics. Instead of waiting for a quarterly attribution report, teams can see which versions perform best in near real time.

In 2024, Mondelez used AI-based video analysis to study over 12,000 ad variants across brands like Oreo and Cadbury. The company found that ads with warmer narration tones and moderate pacing drove 19% higher recall and 11% stronger purchase intent. Those insights were rolled back into production templates, improving both speed and consistency across markets. Mondelez also recently disclosed plans to reduce production costs by 30–50% using its generative-AI tool, with an investment of over U.S. $40 million and target rollout of AI-generated TV ads by the 2026 holiday season.

How AI helps:

  • Content analysis at scale. Vision and audio models scan thousands of creative variants to detect which stylistic traits correlate with stronger brand recall.
  • Real-time dashboards. Campaign teams get immediate feedback on performance by geography, audience, or platform instead of waiting for end-of-month reports.
  • Creative optimization guidance. AI tools surface which combinations of voice, script length, or visual tone work best, turning subjective preferences into measurable variables.

For the first time, creative decisions such as voice choice, image framing and script tempo can be validated by behavioral data, not just opinion. That feedback loop helps marketers spend smarter and produce more resonant campaigns.

The takeaway

Marketers don’t need a grand reset to benefit from AI. The immediate wins are practical: faster video, full production cycles, steady brand voice across regions, more creative tests per month, tighter compliance and a relieved creative team. In a year when budgets are steady rather than expanding, those gains matter. The smartest teams ship smaller, learn quicker and document everything, turning AI from a headline into a dependable part of how they make and run campaigns.Β 

If you’re ready to modernize your workflow and unlock real creative speed, talk to Artlist’s experts. Join 33 million creators using Artlist to produce high-volume, studio-level content without the studio cost, and see how Artlist AI can transform the way you work.Β 

Fix SOC Blind Spots: See Threats to Your Industry & Country in Real Time

Modern security teams often feel like they’re driving through fog with failing headlights. Threats accelerate, alerts multiply, and SOCs struggle to understand which dangers matter right now for their business. Breaking out of reactive defense is no longer optional. It’s the difference between preventing incidents and cleaning up after them. Below is the path from reactive firefighting to a

China-Linked Ink Dragon Hacks Governments Using ShadowPad and FINALDRAFT Malware

The threat actor known as Jewelbug has been increasingly focusing on government targets in Europe since July 2025, even as it continues to attack entities located in Southeast Asia and South America. Check Point Research is tracking the cluster under the name Ink Dragon. It's also referenced by the broader cybersecurity community under the names CL-STA-0049, Earth Alux, and REF7707. The

Samsung seems to be focusing on foldables in 2026 – here are 3 possible reasons why

With the Galaxy Z Trifold on the way and rumors of a small upgrade for the Galaxy S26 series, it seems that Samsung is betting big on its folding phones for next year.

AMD unveils new Radeon RX 9060 XT LP GPU

17 December 2025 at 15:48

AMD unveils new low-power RX 9060 XT GPU model with 16GB of memory AMD has officially unveiled a new RDNA 4 GPU model: a low-power variant of the company’s Radeon RX 9060 XT. This new RX 9060 XT LP GPU was first spotted by ITHOME, though it has also been listed on AMD’s website. This […]

The post AMD unveils new Radeon RX 9060 XT LP GPU appeared first on OC3D.

Nvidia plans heavy cuts to GPU supply in early 2026

17 December 2025 at 14:53

Nvidia reportedly plans 30-40% cut in GeForce GPU production in early 2026 Recent reports have claimed that Nvidia intends to reduce its production capacity for GeForce RTX 50 series GPUs in the first half of 2026. These cuts are reportedly due to shortages of memory, not just GDDR7, but all memory types. 30-40% of Nvidia’s […]

The post Nvidia plans heavy cuts to GPU supply in early 2026 appeared first on OC3D.

OpenAI just made ChatGPT images 4x faster and easier on the eyes β€” now CEO Sam Altman faces the GPU crunch

OpenAI has unveiled ChatGPT Images with new capabilities, which Sam Altman described as β€œsomething fun.” The update promises greater precision, enhanced creativity, and is said to be four times faster.

(PR) Rapidus Unveils New AI Design Tools for Advanced Semiconductor Manufacturing

17 December 2025 at 16:18
Rapidus Corporation today announced a new suite of AI design tools to support its Rapidus AI-Assisted Design Solution (Raads), a key initiative to helping to realize its Rapid and Unified Manufacturing Service (RUMS) concept. The technology approach will also be rebranded to Rapidus AI-Agentic Design Solution with multiple tools being released starting in 2026 and will be provided to customers together with a process design kit and reference flows. The new tools include:
  • Raads Generator: An EDA tool based on large-scale language models (LLMs). When designers input semiconductor specifications, Raads Generator outputs register-transfer level (RTL) design data optimized for Rapidus' 2 nm manufacturing process.
  • Raads Predictor: A tool for RTL debugging and optimization for physical design, placement and routing. Raads Predictor will provide a power performance area (PPA) estimate in a short period of time.
Designers can use Raads Generator to write design ideas and desired specifications as RTL source code and then input data into Raads Predictor along with Synopsys Design Constraints (SDC), allowing them to predict the PPA of the silicon manufactured by Rapidus.

(PR) D-Link Unveils Next-Generation Enterprise Wireless Lineup

17 December 2025 at 15:54
D-Link Corporation (TWSE: 2332), a global leader in networking and intelligent connectivity, today announced the expansion of its enterprise wireless portfolio with the launch of two next-generation access points: the DAP-E9560 BE9500 Wi-Fi 7 Ceiling-Mount Access Point and the DAP-X3060W AX3000 Wi-Fi 6 Wall-Plate Access Point. Designed to address diverse deployment scenarios in hospitality, education, enterprise offices, and multi-dwelling units (MDUs), the new lineup advances D-Link's vision of delivering secure, scalable, and intelligently managed wireless solutions for modern business environments.

Built with D-Link's engineering precision and commitment to Made in Taiwan quality, creative innovation, solution-oriented, one-stop service, and sustainable design, both products reinforce the company's mission of "One Connection * Infinite Possibilities," offering enterprises a more seamless and efficient path to next-generation network transformation.

Netrinos – Zero-config mesh VPN. Connect devices anywhere, no servers needed


Tired of fighting firewalls, port forwarding, and VPN servers just to access your own devices? Netrinos creates a private mesh network that connects all your devices as if they were in the same room. Install, log in, connected. WireGuard encryption, zero configuration. New in Pro: invite team members, control who can access what, and set up gateways to reach your NAS, printers, or home network from anywhere. 60-day free trial, no credit card required.

View startup

Enthusiast modder stuffs an entire gaming PC inside a gutted Commodore PET 2001 β€” replaced the screen with an iPad Retina LCD, but the original keyboard still works

A Redditor found a "pre-gutted" Commodore PET 2001 that they repurposed as a fully-fledged gaming PC, while keeping the Commodore's keyboard intact and functional. The internals are relatively modest, but they can still play most modern games on the retrofitted screen, which is a Retina LCD from an iPad.

New calculator helps evaluate the economics of datacenters in space β€” running the numbers on orbital computing reveals a brutal reality

17 December 2025 at 15:00
Orbital data centers have been proposed as a way around the power, cooling, and regulatory requirements of building on Earth, but new analysis reveals that the costs of deploying such installations would be astronomical.

The Unseen Threat in Your Browser: Why AI Demands a Security Reckoning

17 December 2025 at 15:18

The post The Unseen Threat in Your Browser: Why AI Demands a Security Reckoning appeared first on StartupHub.ai.

Gartner’s recent advisory, urging organizations to ban AI browsers from the workplace, has ignited a critical conversation within the cybersecurity community. This provocative stance, explored in a recent episode of IBM’s Security Intelligence podcast by host Matt Kosinski and panelists Austin Zeizel, Evelyn Anderson, and Ryan Anschutz, underscores a fundamental tension: the rapid innovation of […]

The post The Unseen Threat in Your Browser: Why AI Demands a Security Reckoning appeared first on StartupHub.ai.

GhostPoster Malware Found in 17 Firefox Add-ons with 50,000+ Downloads

A new campaign named GhostPoster has leveraged logo files associated with 17 Mozilla Firefox browser add-ons to embed malicious JavaScript code designed to hijack affiliate links, inject tracking code, and commit click and ad fraud. The extensions have been collectively downloaded over 50,000 times, according to Koi Security, which discovered the campaign. The add-ons are no longer available.

Windows 11 Will Ask for Permission Before AI Agents Access Personal Files

17 December 2025 at 13:43
Microsoft has implemented a mandatory consent framework for Windows 11 that stops AI agents from accessing personal files without explicit user permission. The company has updated its agentic AI features documentation for experimental preview builds, specifying six protected folders as off-limits by default: Desktop, Documents, Downloads, Music, Pictures, and Videos. Each AI assistant must request access individually, rather than receiving system-wide permissions. This opt-in design ensures that standard installations remain unaffected unless users decide to activate the feature and approve specific agents.

The permission system functions on a per-agent basis, meaning that approval for one AI tool does not automatically apply to others that might be installed. When an agent tries to access files, Windows presents a consent interface where users can choose to grant permanent access, require reauthorization for each interaction, or deny requests entirely. Each AI assistant has its own settings portal, allowing users to modify these permissions later if needed. Additionally, the platform is testing discrete connectors that manage interactions with system applications like File Explorer and Settings, separate from core folder permissions. This modular design allows users to, for instance, permit an agent to adjust system settings while restricting access to personal photos or documents.

With New Mozilla CEO, Firefox Will Become a "Modern AI Browser"

17 December 2025 at 13:00
Mozilla, the company behind one of the last non-Chromium browsers, Firefox, has appointed Anthony Enzor-DeMeo as the new CEO. Under his leadership, the company plans to transform Firefox into a "modern AI browser." This announcement provides a glimpse into the future direction of the Firefox browser, sparking concerns within the community that one of the last independent browser projects will adopt AI features that may not be widely desired. While modernization of browsers is generally positive, the Firefox user base tends to prefer fast and lightweight solutions over AI-enhanced software. Below, you can find the entire new message from the new CEO Anthony Enzor-DeMeo:

AI Presentation Maker Transforms Research Workflow

17 December 2025 at 11:18

The post AI Presentation Maker Transforms Research Workflow appeared first on StartupHub.ai.

Google's NotebookLM integrates an AI presentation maker, powered by Nano Banana Pro, to transform raw research into polished, visually engaging slide decks.

The post AI Presentation Maker Transforms Research Workflow appeared first on StartupHub.ai.

From Models to Agents The Next AI Frontier

17 December 2025 at 11:15

The post From Models to Agents The Next AI Frontier appeared first on StartupHub.ai.

The true inflection point in artificial intelligence is not merely the advent of large language models, but their rapid evolution into autonomous agents, capable of understanding context, intent, and orchestrating complex tasks. This profound shift, from static models to dynamic, decision-making entities, heralds a new era of enterprise AI, moving beyond mere chatbots to intelligent […]

The post From Models to Agents The Next AI Frontier appeared first on StartupHub.ai.

Apple May Refresh iMac Pro Desktop AIO, But Without Top-Tier M5 Ultra SoC

17 December 2025 at 09:12
An Ultra variant of Apple's M5 SoC line-up has been rumored for a while now, specifically when it comes to Apple's Mac Studio, which has more or less become the brand's de-facto high-end desktop workstation lineup. Given that the M3 Ultra occupied the top position in Apple's chip line-up, it was expected that if Apple ever released an update to the Apple iMac Pro, it would feature the Ultra model of the current SoC family. However, thanks to a leaked kernel debug tool, courtesy of MacRumors, we now have tentative confirmation that the Apple iMac Pro will likely top out at an M5 Max instead of the rumored Ultra SoC. The software refers to an iMac identified by the codename J833c with a platform that goes by H17C, which has been associated with the M5 Max SoC. This suggests that the M5 Max iMac Pro is not only planned, but has already made it into testing.

The current generation iMac tops out at the M4 SoC, which is still an impressive performer, for what it is, but if the difference between the M4 and the M4 Max is anything to go by, an M5 Max-powered iMac Pro would mean stepping up from a 10-core GPU and 10-core CPU to a 16-core CPU and 40-core GPU, and that's just in terms of processing power. The iMac Pro will likely see a revival of the bigger 27-inch display, and it's also more than likely that Apple will also upgrade the display brightness and perhaps even its HDR performance, since rumors point to the iMac Pro being destined for professional and creative workspaces. Apple only recently released the base model M5 SoC, but the M5 Pro, Max, and Ultra are expected to launch at some point in Spring 2026.

BitterBot AI – The AI agent that executes tasks, not just explains them


BitterBot is a free AI agent that actually does the work for you. Instead of getting instructions on how to code something, BitterBot just builds it. It can create websites, analyze data, automate tasks, and manage files - all on its own. Just tell it what you want done and it figures out how to do it, no technical skills needed. Try it free at https://bitterbot.ai/

View startup

SteamOS Will Now Warn Users About Incompatible Xbox Controllers

17 December 2025 at 07:13
Valve has put a lot of development work into SteamOS to make the Steam Deck experience as seamless as possible, but there are still some limitations to the Linux-based OS, and specifically with third-party hardware or software. Tools like Steam Input and the Proton compatibility layer go a long way to ensuring compatibility, but they're not perfect. Case in point, Valve's latest Steam Deck Beta Client update has added a warning to notify users when they pair or connect an Xbox controller with incompatible firmware to the Deck via Bluetooth. This should resolve some of the frustration regarding any potential failed pairing attempts or issues

Unfortunately, at the time of writing, there is nothing users can do from the Steam Deck or SteamOS itself when encountering this issue. As Valve acknowledges in the update notes, the only way to correct the incompatibility is to connect the controller to a Windows PC and update firmware via the Xbox Accessories app. The same update also makes some changes to the default accessibility options, making it easier to access the Quick Access menu with the guide button, and it adds options to report and block offensive group chat members directly from the group chat. This update is still in the beta channel, but it should arrive in the mainline channel within a few weeks.

Google Cloud’s Agent Development Kit: Orchestrating Autonomous AI

17 December 2025 at 04:45

The post Google Cloud’s Agent Development Kit: Orchestrating Autonomous AI appeared first on StartupHub.ai.

The promise of artificial intelligence has long extended beyond simple query-response systems to truly autonomous entities capable of complex reasoning and action. Annie Wang, an AI expert, articulated this vision succinctly when she stated, β€œAn agent is essentially an LLM that can reason, act, and observe.” This fundamental shift, from a static language model to […]

The post Google Cloud’s Agent Development Kit: Orchestrating Autonomous AI appeared first on StartupHub.ai.

Tesla engaged in deceptive marketing for Autopilot and Full Self-Driving, judge rules

17 December 2025 at 05:14
The judge ruled Tesla should have to suspend manufacturing and sales in California for 30 days, but the DMV stayed those rulings for 90 days to allow Tesla to comply.

Outlex – AI-powered legal OS for European startups


Outlex is an AI-first legal operating system built for European startups navigating 27 jurisdictions. Our hybrid platform combines intelligent automation for routine legal work with seamless handoffs to vetted specialist lawyers when the stakes are high. Replace fragmented tools and €50K+ annual legal spend with a unified infrastructure that delivers 70% savings, 10x faster turnarounds, and complete transparencyβ€”giving founders peace of mind to focus on building

View startup

Wall Street’s Data Center Dilemma: A Capital Expenditure Reckoning for AI

17 December 2025 at 04:15

The post Wall Street’s Data Center Dilemma: A Capital Expenditure Reckoning for AI appeared first on StartupHub.ai.

Wall Street has reached a stark conclusion regarding the massive buildout of AI data centers: companies are β€œpaying too much money to build out the data centers.” This pronouncement, delivered with characteristic fervor by Jim Cramer on CNBC’s Mad Money, signals a critical shift in investor sentiment, moving away from the unbridled enthusiasm that has […]

The post Wall Street’s Data Center Dilemma: A Capital Expenditure Reckoning for AI appeared first on StartupHub.ai.

$80 Xbox Games May Be Back on the Table After The Outer Worlds 2 Debacleβ€”Eventually

17 December 2025 at 03:32
Following Nintendo's move to $80 AAA games earlier in 2025, Microsoft tried the same trick with The Outer Worlds 2, initially announcing the launch price at $79.99 but later dropping the price back down to $69.99. Microsoft blamed the price reduction on "market conditions," but speculation was that the actual reason behind walking back the price was backlash from the gaming community. Microsoft is also not the first game studio to sit out the $80 game price shift, with EA earlier announcing a similar decision. Now, it seems as though the $80 AAA game is back on the table for Xbox Game Studios, according to recent comments by Matt Booty, Microsoft's President of Game Content and Studios, in an interview with Variety.

Booty subtly acknowledged the backlash to the initial pricing of The Outer Worlds 2, commenting that "We've reacted in the last year, and I think for us, the real focus is going to beβ€”I'll come to that phrase meeting people where they are." He also said that Microsoft's gaming division will focus on "delivering player satisfaction and delivering player value," although he says that, for the time being, there are no updates to Microsoft's game pricing strategy. The implication seems to be that Microsoft hasn't completely written off the idea of the $80 game, but that it is simply waiting for the right time. He also went on to say that he thinks "there's going to be less of a focus on what's that top line price of a game, as people start to engage in different ways with games," and implied that Microsoft would explore different monetization strategies and "listen to the feedback from fans" as it did so.

Tencent Invests in Unrecord, the Viral Bodycam Game, As Development Enters Full Swing

17 December 2025 at 02:59
Unrecord, the bodycam shooter that went viral in 2023 thanks to its hyperrealistic graphics, has been in production for a while already, but Drama Studios, the studio behind the game, has just announced that it has secured the requisite funding to "build the best game possible." This isn't the first round of funding that Drama Studios has received for Unrecord, with the outfit previously being awarded $2.5 million from The Games Fund in 2024. The new round of funding means that the game's development is in full swing, and the studio will be growing yet again. The game is currently available to wishlist on Steam, and it seems to remain popular, having amassed over 130,000 followers since 2023. There is still no release date attached to the game, however.

According to GamesIndustry.biz, some of this funding came from Tencent, who purchased a minority stake in the French game studio, although there aren't many details on how much exactly Tencent has invested in the indie studio. However, a recent post on the game's X account says that the studio has grown from two developers to 10 and is and that the company is now hiring more developers to work on the game. It has also promised an updated look at gameplay footage for Unrecord in 2026. Unrecord has not been without controversy, with dissenters labeling the game's premise exploitativeβ€”specifically because it hinges on realistic bodycam-style visuals and depictions of violent police interactions and abuse of power.

Education’s AI Reckoning: Anthropic Navigates Light and Shade

17 December 2025 at 02:46

The post Education’s AI Reckoning: Anthropic Navigates Light and Shade appeared first on StartupHub.ai.

Artificial intelligence is not merely changing education; it is acting as a β€œforcing function that makes everyone deal with it now,” as Maggie Vo, Head of Anthropic’s Ministry of Education, succinctly put it. This sentiment encapsulates the urgent yet nuanced conversation unfolding among the Anthropic team about AI’s profound impact on learning. In a recent […]

The post Education’s AI Reckoning: Anthropic Navigates Light and Shade appeared first on StartupHub.ai.

AI Funding Fears Overstated, Says Goldman Sachs’ Sung Cho

17 December 2025 at 02:16

The post AI Funding Fears Overstated, Says Goldman Sachs’ Sung Cho appeared first on StartupHub.ai.

The prevailing market anxiety surrounding artificial intelligence funding, often fueled by dramatic shifts in corporate valuations and perceived vulnerabilities, is largely overstated. This was the central, reassuring message from Sung Cho, Co-head of Public Tech Investing and U.S. Fundamental Equity at Goldman Sachs Asset Management, during a recent discussion on CNBC’s β€œClosing Bell” with Scott […]

The post AI Funding Fears Overstated, Says Goldman Sachs’ Sung Cho appeared first on StartupHub.ai.

EU probes Google for "illegally" scraping the web to catch up to its AI rivals

17 December 2025 at 00:05

European regulators have opened a new antitrust investigation into Google, centering on whether the company illegally scraped online content to shore up its artificial intelligence products as it struggled to keep pace with faster-moving rivals. Fortune reports that the probe targets Google's use of publisher content and user-uploaded material to...

Read Entire Article

Mutually – Mutually connects you with others based on your digital footprint


Mutually isn’t just another swipe app. It’s a smarter way to meet people who match your mindset, taste, and energy. Mutually connects you with others based on your digital footprint. Whether you’re looking for new friends or something more, you’ll discover people who share your passions, playlists, and personality.

Why Mutually stands out β€’ Find people who actually share your interests β€’ Authentic profiles powered by your favorite platforms β€’ Quick matches through QR codes β€’ Separate modes for friends and dating β€’ Built-in chat for real conversations, not small talk Discover your crowd. Deepen your connections. Find your Mutualize.

View startup

AI Data Centers Face Rising Political Heat Over Energy Costs

16 December 2025 at 23:45

The post AI Data Centers Face Rising Political Heat Over Energy Costs appeared first on StartupHub.ai.

The burgeoning demand for artificial intelligence infrastructure is colliding with a formidable, rapidly evolving political headwind: the escalating energy consumption of data centers and its direct impact on household utility costs. This friction, highlighted in a recent CNBC discussion between TechCheck Anchor Deirdre Bosa and anchor Kelly Evans, underscores a critical shift in how the […]

The post AI Data Centers Face Rising Political Heat Over Energy Costs appeared first on StartupHub.ai.

Salesforce Agentic AI Redefines Sales Productivity

16 December 2025 at 23:18

The post Salesforce Agentic AI Redefines Sales Productivity appeared first on StartupHub.ai.

Salesforce Agentic AI is poised to redefine sales productivity by empowering teams with intelligent automation and data-driven insights, especially for SMBs.

The post Salesforce Agentic AI Redefines Sales Productivity appeared first on StartupHub.ai.

ContextForge MCP Gateway: the MCP router for AI agents

16 December 2025 at 23:16

The post ContextForge MCP Gateway: the MCP router for AI agents appeared first on StartupHub.ai.

IBM’s open-source ContextForge MCP Gateway positions itself as an enterprise-ready MCP router for AI agents, sitting between LLM-driven applications and the tools and data they need. Framed as a secure Model Context Protocol gateway, it turns a sprawl of MCP servers and REST endpoints into a single, governed interface that AI agents can call without […]

The post ContextForge MCP Gateway: the MCP router for AI agents appeared first on StartupHub.ai.

AI’s Concentrated Talent War and a Precarious Labor Market

16 December 2025 at 23:15

The post AI’s Concentrated Talent War and a Precarious Labor Market appeared first on StartupHub.ai.

The ongoing β€œAI talent war continues in tech without generating many jobs,” according to Diane Swonk, Chief Economist at KPMG, highlighting a significant paradox at the heart of the current economic landscape. This keen observation from Swonk, made during a recent appearance on CNBC’s β€œThe Exchange” alongside Kelly Evans and Scott Wapner, cuts through the […]

The post AI’s Concentrated Talent War and a Precarious Labor Market appeared first on StartupHub.ai.

Google: Exact match keywords won’t block broad match in AI Max

16 December 2025 at 23:54
Why phrase match is losing ground to broad match in Google Ads

Ginny Marvin, Google’s Ads Liaison, is clarifying how keyword match types interact with AI Overviews (AIO) and AI Mode ad placements β€” addressing ongoing confusion among advertisers testing AI Max and mixed match-type setups.

Why we care. As ads expand into AI-powered placements, advertisers need to understand which keywords are eligible to serve β€” and when β€” to avoid unintentionally blocking reach or misreading performance.

Back in May. Responding to questions from Marketing Director Yoav Eitani, Marvin confirmed that an ad can serve either above or below an AI Overview or within the AI Overview β€” but not both in the same auction:

  • β€œYour ad could trigger to show either above/below AIO or within AIO, but not both at this time.” Marvin confirmed.

While both exact and broad match keywords can be eligible to trigger ads above or below AIO, only broad match keywords (or keywordless targeting) are eligible to trigger ads within AI Overviews.

What’s changed. In a follow-up exchange with Paid Search specialist Toan Tran, Marvin clarified that Google has updated how eligibility works. Previously, the presence of an exact match keyword could prevent a broad match keyword from serving in AI Overviews. That is no longer the case.

  • β€œThe presence of the same keyword in exact match will not prevent the broad match keyword from triggering an ad in an AI Overview, since the exact match keyword is not eligible to show Ads in AI Overviews and hence not competing with the broad match keyword.” Marvin said.

Since exact and phrase match keywords are not eligible for AI Overview placements, they do not compete with broad match keywords in that auction β€” meaning broad match can still trigger ads within AIO even when the same keyword exists as exact match.

The big picture. Google is reinforcing a clear separation between traditional keyword matching and AI-powered intent matching. Ads in AI Overviews rely on a deeper understanding of both the user query and the AI-generated content, which is why eligibility is limited to broader targeting signals.

The bottom line. Exact and phrase match keywords won’t show ads in AI Overviews β€” but they also won’t block broad match from doing so. For advertisers leaning into AI Max and AIO placements, broad match and keywordless strategies are now essential to unlocking reach in Google’s AI-driven surfaces.

The Samsung Galaxy Z Trifold's best feature won't be its hardware – here's why I'm more excited for DeX

The Samsung Galaxy Z Trifold is due to merge the mobile and tablet experience like never before – and if Samsung plays its cards right, it could revolutionize on-the-go productivity.

Obscura VPN just aced its first independent audit

Obscura VPN, a relatively new VPN service for iOS and macOS, has just passed its first independent audit. The VPN prioritizes not just user security, but also complete browsing privacy.

Larian's Swen Vincke Says Divinity Will be Turn-based RPG, Hints About "Next Level" Experience

16 December 2025 at 23:36
The unveiling of Divinity at this year's edition of The Game Awards was widely celebrated. Since releasing Baldur's Gate 3 back in 2023, Larian Studios has moved on from working within the constraints of Dungeons and Dragons (D&D) license. RPG enthusiasts have long-pondered over the prospect of revisiting the Divinity universe. Last week, the Belgian-headquartered development house did not reveal many details regarding their next flagship project. Their pagan ritual-filled cinematic trailer did not contain any gameplay footage. Online gaming communities have debated about Divinity's gameplay systems; will it be a real-time action experience, or a turn-based affair? Recently, Larian head honchoβ€”Swen Vinckeβ€”sat down with Bloomberg's Jason Schreier.

In a Q&A session, the company's founder acknowledged that his team will try to build on the success of Baldur's Gate 3. This will be a big challenge, Vincke elaborated: "it's more. It's more pressure. The weight of the expectations weighs high. We're trying not to think about it, because we have to make our own thing." Larian's last in-house fantasy IPβ€”Divinity: Original Sin IIβ€”was released back in 2017. Continuing his train of thought he opined: "Baldur's Gate 3 was a good game, and I'm proud of it, but I think this one (Divinity) is going to be way better...This is going to be us unleashed, I think. It's a turn-based RPG featuring everything you've seen from us in the past, but it's brought to the next level." Additionally, Vincke confirmed that Divinity will be released via an early access model; likely following a similar trail pioneered by BG3's staggered release roadmap.

(PR) Sucker Punch Productions Co-Founder Stepping Down, New Leadership Duo Revealed

16 December 2025 at 22:59
After nearly three decades helping bring to life iconic franchises like Sly Cooper, inFAMOUS, Ghost of Tsushima, and Ghost of Yōtei, Brian Fleming has announced he's passing Sucker Punch studio leadership on to a new generation. Over the past year, Brian has worked closely with PlayStation Studios to ensure that Sucker Punch was in the best hands moving forward with a strong foundation for the studio's continued success. Starting January 1, longtime creative and technical leaders Jason Connell and Adrian Bentley will step into new roles as studio heads, continuing to guide the team's focus on ambitious, character-driven experiences that define PlayStation Studios.

New Studio Heads: Jason Connell and Adrian Bentley
Jason and Adrian have been instrumental in shaping the creative and technical direction of Sucker Punch in recent years. Jason, who served as Co-Creative Director on the Ghost franchise, has helped define the studio's visual identity and storytelling style, bringing cinematic depth and emotional resonance to PlayStation players around the world. Adrian, as Technical Director, has led the studio's engineering and production efforts, driving innovation across development tools and game systems that make Sucker Punch's worlds so immersive. Together, they represent the perfect blend of creative vision and technical excellence that has always set the studio apart.

Core Ultra 7 365 Test Sample Gets Geekbenched, Scores Indicate Generational Regression

16 December 2025 at 22:48
As of this morning, a mysterious "Lenovo 4810X90100" test platform has produced slightly concerning benchmark scores. The freshly-archived database entry shows a next-gen laptop/notebook being driven by an unreleased Intel Core Ultra 7 365 "Panther Lake-H" mobile processor. Overall Geekbench 6.3 scoresβ€”2451 (single-core) and 9714 (multicore)β€”were swiftly noted down by keen trackers of emerging PC hardware. According to Geekbench browser data, Team Blue's upcoming mid-range SKU is an 8-core design, with two contained clusters consisting of four units (each). This leak seems to align with official Intel preview material that describes a generic product's 4 Performance-core + 4 Low Power-Efficiency-core configuration.

The leaked "Panther Lake-H" part's Geekbench tallies were compared to scores generated by similarly positioned predecessors: Core Ultra 7 268V and Core Ultra 7 258V. Surprisingly, the older "Lunar Lake-H" mobile CPUsβ€”on averageβ€”outperform the younger sibling. When weighing up the differences between Core Ultra 7 268V and Core Ultra 7 365 SKUs, Wccftech acknowledged clock speed disparities and contrasting TDP ranges. Unfortunately, the leaked "Panther Lake-H" candidate trails behind in terms single-core and multicore of scoresβ€”6% and 7%, respectively. Notebookcheck reckoned that the evaluated Core Ultra 7 365 APU offers underwhelming "12th-gen-like performance." When cross-referenced against the Core Ultra 7 258V, Intel's newer chip lagged behind in both single-core (11%) and multicore scenarios (10%). It is plausible that the "Lenovo 4810X90100" prototypeβ€”sporting 64 GB of non-specific system memoryβ€”was set up with pre-release drivers, therefore putting its hosted Core Ultra 7 365 processor in a disadvantaged position.

Microsoft updates the Xbox Wireless Headset with Bluetooth LE Audio

16 December 2025 at 22:38
In a blog post on Xbox Wire, Microsoft has announced a new firmware update for the Xbox Wireless Headset that will add Bluetooth LE Audio support for compatible Windows 11 devices. The blog post specifically mentions the ASUS ROG Xbox Ally and Xbox Ally X, but any PC with Windows 11 and Bluetooth LE/LC3 audio support will be able to take advantage of the new features. Microsoft recently announced support for super wideband stereo voice support for Windows 11 and the Xbox Wireless Headset is the first headset we're aware of that now supports that feature.

Microsoft also promises that the firmware update will bring with it lower latencyβ€”this is part of the LC3 audio codec when compared to the old SBC audio codecβ€”and improved battery life, although the company didn't say by how much. Finally the firmware will bring with it a preview of a broadcast audio feature that will allow "Windows 11 insiders on select PCs" to share the game audio with other Bluetooth devices, but the blog post doesn't go into any details of how this will work. The Xbox Wireless Headset will also be able to take advantage of super wideband stereo audio in Windows 11, which Microsoft claims will offer better in-game audio, although for those that already own a wireless headset with a dedicated dongle, this might not bring any real world advantages. To update the Xbox Wireless Headset you either need to use the Xbox Accessories app on an Xbox console or in Windows 11 and the update only applies to the 2024 version of the headset.

NVIDIA Powers Nearly 72% of Cloud Accelerator Locations, AMD at 5.8%

16 December 2025 at 22:14
The cloud accelerator market share study was conducted at UBS bank, one of the leading research hubs and close followers of market trends. According to UBS, NVIDIA commands 71.2% of the cloud accelerator location share, which includes GPUs and other types of ASICs. The study also included AMD, and other ASICs. For the AMD front, the situation is entirely different as the data shows AMD only captures 5.8% of the cloud accelerator location share. The remaining portion is split between ASICs designed by the likes of AWS, Meta, Broadcom, MediaTek, and many others, which command 22.3%.

For NVIDIA, the company has managed to penetrate its GPUs in 258 datacenter locations, serving as cloud GPU providers. This is a significant number considering that some of these customers are ordering tens and even hundreds of thousands of GPUs at once, especially as the demand expands. NVIDIA is still the backbone of the entire AI infrastructure, and even its older GPUs, which include A100 and H100, are capturing a significant part of the GPU cloud market share. As "Blackwell" SKUs, designed for a million GPU scale systems, are scaled, and we enter the "Rubin" era, NVIDIA's lead may only expand.
Before yesterdayTech

ChatGPT Images Unleashes Unprecedented Visual Agility

16 December 2025 at 22:45

The post ChatGPT Images Unleashes Unprecedented Visual Agility appeared first on StartupHub.ai.

The latest iteration of ChatGPT Images, powered by OpenAI’s new flagship image generation model, GPT Image 1.5, signals a profound shift in the accessibility and flexibility of visual content creation. This release, demonstrated through a dynamic visual showcase, moves beyond simple image generation to offer sophisticated editing and stylistic transformations that were once the exclusive […]

The post ChatGPT Images Unleashes Unprecedented Visual Agility appeared first on StartupHub.ai.

The 2026 AI predictions: Why infrastructure will fail, but apps will fly.

16 December 2025 at 22:32

The post The 2026 AI predictions: Why infrastructure will fail, but apps will fly. appeared first on StartupHub.ai.

While Big Tech faces supply chain bottlenecks and AGI timelines push into the 2030s, AI application startups are set to achieve unprecedented scale in 2026.

The post The 2026 AI predictions: Why infrastructure will fail, but apps will fly. appeared first on StartupHub.ai.

AI’s Unseen Cost: Political Pressure Mounts on Data Center Energy Demands

16 December 2025 at 21:45

The post AI’s Unseen Cost: Political Pressure Mounts on Data Center Energy Demands appeared first on StartupHub.ai.

The burgeoning computational demands of artificial intelligence are rapidly colliding with public policy and local politics, as highlighted in a recent CNBC β€œMoney Movers” segment. CNBC Business News TechCheck Anchor Deirdre Bosa reported on growing political pressure stemming from the massive energy consumption of AI data centers, revealing a new front of risk for the […]

The post AI’s Unseen Cost: Political Pressure Mounts on Data Center Energy Demands appeared first on StartupHub.ai.

Your Support Team Should Ship Code – Lisa Orr, Zapier

16 December 2025 at 21:15

The post Your Support Team Should Ship Code – Lisa Orr, Zapier appeared first on StartupHub.ai.

Lisa Orr, Product Leader at Zapier, shared a compelling narrative about how her company is leveraging artificial intelligence to transform its support operations, enabling the support team to actively ship code. The core problem was the sheer volume of support tickets generated by API changes, overwhelming traditional support workflows. Zapier’s journey began with a clear […]

The post Your Support Team Should Ship Code – Lisa Orr, Zapier appeared first on StartupHub.ai.

Google AI Overviews surged in 2025, then pulled back: Data

16 December 2025 at 21:00

Google rapidly expanded AI Overviews in search during 2025, then pulled back as they moved into commercial and navigational queries. These findings are based on a new Semrush analysis of more than 10 million keywords from January to November.

AI Overviews surged, then retreated. Google didn’t roll out AI Overviews in a straight line in 2025. A mid-year spike gave way to a pullback, suggesting Google moved fast to test the feature, then eased off based on user data:

  • January: 6.5% of queries triggered an AI Overview
  • July: AI Overview visibility peaked, appearing in just under 25% of queries.
  • November: Coverage fell back to less than 16% of queries.

Zero-click behavior defied expectations. Surprisingly, click-through rates for keywords with AI Overviews have steadily risen since January. AI Overviews don’t automatically reduce clicks and may even encourage them.

  • AI Overviews still appear more often on searches that already tend to drive no clicks.
  • But when Semrush compared the same keywords before and after an AI Overview appeared, zero-click rates fell from 33.75% to 31.53%.

Informational queries no longer dominate. Early 2025 AI Overviews were almost entirely informational:

  • January: 91% informational
  • October: 57% informational

Now, AI Overviews are appearing for commercial and transactional queries:

  • Commercial queries: Increased from 8% to 18%
  • Transactional queries: Increased from 2% to 14%

Navigational queries are rising fast. In an unexpected shift, AI summaries are increasingly intercepting brand and destination searches:

  • Navigational AI Overviews grew from under 1% in January to more than 10% by November.

Google Ads + AI Overviews. Earlier this year, ads rarely appeared next to AI Overviews. Now they’re common:

  • Ads alongside AI Overviews rose from about 3% in January to roughly 40% by November.
  • Ads show at the bottom of around 25% of AI Overview SERPs.

Science is the most impacted industry. By keyword saturation, Science leads all verticals for AI Overviews at 25.96%. Computers & Electronics follows at 17.92%, with People & Society close behind at 17.29%.

  • Since March, Food & Drink has seen the fastest growth in AI Overviews of any category.
  • Meanwhile, Real Estate, Shopping, and Arts & Entertainment remain lightly affected, with AI Overviews appearing on fewer than 3% of keywords.

Why we care. AI Overviews are unevenly and persistently reshaping click behavior, commercial visibility, and ad placement. Volatility is likely to continue, so closely monitor performance shifts tied to AI Overviews.

The report. Semrush AI Overviews Study: What 2025 SEO Data Tells Us About Google’s Search Shift

Dig deeper. In May, I reported on the original version of Semrush’s study in Google AI Overviews now show on 13% of searches: Study.

Compromised IAM Credentials Power a Large AWS Crypto Mining Campaign

An ongoing campaign has been observed targeting Amazon Web Services (AWS) customers using compromised Identity and Access Management (IAM) credentials to enable cryptocurrency mining. The activity, first detected by Amazon's GuardDuty managed threat detection service and its automated security monitoring systems on November 2, 2025, employs never-before-seen persistence techniques to hamper

Amazon UK's last-minute gifts sale is live – I've hand-picked the 29 best deals on TVs, appliances, headphones, and more

Amazon's UK last-minute sale continues this week, so I've gathered together 29 of the best deals if you're shopping for some gift inspiration or hunting down some final tech bargains of the year.

UK MPs target VPNs in latest Online Safety Act debate

The debate around age verification and VPNs heats up as MPs confirm their commitment to review the role these providers should have in preventing their service from being misused to evade the law.

Nixxes updates Ghost of Tsushima to enable FSR ML Frame Generation

16 December 2025 at 20:29

Hotfix 8 for Ghost of Tsushima adds FSR ML Frame Generation to the game Nixxes Software has released its β€œPatch 8 Hotfix” for Ghost of Tsushima’s PC version, adding support for AMD FSR ML Frame Generation. This new Frame Generation technique is part of AMD’s FSR β€œRedstone” update. With this update, users of AMD’s Radeon […]

The post Nixxes updates Ghost of Tsushima to enable FSR ML Frame Generation appeared first on OC3D.

(PR) "A Game About Feeding A Black Hole" Available Now on PC Windows & Linux Platforms

16 December 2025 at 21:46
It's finally here. Celebrate A Game About Feeding A Black Hole's launch with a limited-time 30% discount on Steam. Thank you to everyone who joined us along the way. As always, keep feeding the black hole.

Feeders, Andy and I set out to build a small project along the lines of an incremental game that made financial sense for the time we'd put in. For myself, it has been a rollercoaster of excitement, stress, worry, and more. I would also describe it as a rocketship, and every time we think things are calming down...BAM! It feels like THIS: absolute chaos, want to keep going, and can't take your eyes off the moment.

Minisforum EU & UK Stores Start Selling EOP4A OCuLink External GPU Card

16 December 2025 at 21:34
Minisforum's UK and Europe webstores have listed a brand-new EOP4A OCuLink external GPU card. The tinyβ€”80 mm x 40 mm (not including bracket)β€”adapter's interface is advertised as offering native high-performance. According to the manufacturer's typo and error-laden promo material, this design: "delivers lossless PCIe (4.0) to OCuLink conversion," thus "unleashing the full potential of external NVMe storage and compute accelerators." The Chinese tech specialist's EOP4A model seems destined for deployment in desktop setups, with externally connected graphics card boosting performance across content creation, AI development, and gaming applications.

The item's product specification sheet mentions a TO OCuLink SF8611/8612 interface, compatible PCIe slot types (x1, x4, x8, & x16), and support for (up to)
PCIe 4.0 Γ—4 bandwidths. The adapter is bundled with low-profile and full-height brackets. Plug-and-play functionalityβ€”e.g, driverless operationβ€”is a big selling point. Currently, the EOP4A is listed with a tidy discountβ€”for prospective UK buyers, Β£20 has been deducted from the standard price Β£89 tag (including VAT). A similar launch promotion is in effect at the Minisforum EU webshop, albeit with local taxes applied at checkout. The premium constructed EOP4A can be picked up for Β£69 (~$93 USD) or €55 + VAT (~$65 USD). By contrast, Minisforum's North American webstore directs potential customers to a placeholder product page. At the time of writing, this listing bears no pricing or stock availability info.

FSP Launches S210 mATX Small Tower Case Series

16 December 2025 at 21:24
FSP has introduced the S210, a compact microATX small tower case with a 23-liter volume. The case measures 350 Γ— 212 Γ— 310 mm and can accommodate full-length graphics cards up to 340 mm and CPU air coolers up to 170 mm in height. It's made of 0.7 mm SPCC steel, paired with a single tempered glass side panel. A metal mesh section covering the front and top of the case improves airflow, while a built-in carry handle adds basic portability. As with many small form factor cases, power supply support is limited to standard ATX units up to 140 mm in length. Cooling options include space for two 120 mm fans or a 240 mm radiator at the top, along with a single 120 mm rear fan.

Storage support consists of one 3.5-inch drive (shared with an SSD mount) and two dedicated 2.5-inch SSD mounts. For expansion, it features four PCIe slots while the front I/O panel is minimal, offering two USB 3.0 ports, an audio jack and a power button. The S210 supports microATX and Mini-ITX motherboards and is currently listed in black as the S210-B variant, so we can expect other color options in the future. Pricing and exact availability haven't yet been announced by FSP.

Kingston Warns SSD Shortage Will Get Worse in the Next 30 Days

16 December 2025 at 21:23
The modern-day SSD business is divided into two main perspectives: some predict that NAND flash shortages will ease, while others speculate that the situation will worsen. Kingston belongs to the latter group, with the company's datacenter SSD business manager, Cameron Crandall, discussing this on The Full Nerd Network podcast. According to him, the NAND flash shortage is expected to significantly worsen in the next 30 days, leading to an increase in SSD prices from current levels. In 2025 alone, Kingston reports that NAND flash prices have surged by 246% from the first quarter until now, with 70% of that increase occurring in just the last 60 days. This indicates that many price hikes have already been integrated into the supply chain.

Edward Crisler, the PR manager for Sapphire, advised PC gamers and potential buyers not to rush into panic buying. He stated, "The good news is, I don't think the real pain we're experiencing now and for the next six months or so will last much longer than that." However, he also highlighted that the main challenge is the uncertainty of the situation. This uncertainty is causing OEMs to increase NAND flash contracts, where no one besides NAND flash manufacturers can predict what is exactly going on in the supply chain, and just how long the massive demand will outpace supply.

(PR) Worldwide Server Market Revenue Increased 61% in Q3 of 2025, According to IDC

16 December 2025 at 20:58
According to the International Data Corporation (IDC) Worldwide Quarterly Server Tracker, the server market reached a record $112.4 billion dollars in revenue during the third quarter of the year. This quarter showed another high double digit-growth rate by reaching a year-over-year (YoY) increase of 61% in vendor revenue compared to the same quarter of 2024. Revenue generated from x86 servers increased 32.8% in 2025Q3 to $76.3 billion while Non-x86 servers increased 192.7% YoY to $36.2 billion.

Revenue for servers with an embedded GPU in the third quarter of 2025 grew 49.4% year-over-year representing more than half of the server market revenue. The fast pace at which hyperscalers and cloud service providers have been adopting servers with embedded GPUs has fueled the server market growth which almost doubled in size compared to 2024 with revenue of $314.2 billion dollars for the first three quarters of 2025.

(PR) Team Cherry Introduces "Sea of Sorrow" - Hollow Knight: Silksong's First Expansion

16 December 2025 at 20:58
Heya Everyone! With Hollow Knight: Silksong's launch year coming to a close, we wanted to take a moment and share a peek at what we've been working on, what's available now, and what's coming in 2026. But firstly, and most importantly, we wanted to say a huge thank you to all the players who've braved Silksong's distant and dangerous lands. That's over seven million of you who've purchased the game, alongside millions more playing on Xbox Game Pass!

It's a truly staggering number of players, more than we could have ever expected (enough to crash all of the storefronts!). Watching the community grow, seeing the amazing art, the mods, the unexpected strategies, and the support between players through the game's challenges has been hugely rewarding for us here at home. Your continued enthusiasm remains a massive motivator as we work towards expanding the game even further. That first big expansion is already well underway!

AMD Ryzen AI 9 465 APU Geekbenched, "Gorgon Point" Offers Negligible Performance Bump

16 December 2025 at 20:46
AMD's not-yet-launched "Ryzen AI 400 series" seems to be nearing finalized form, given an uptick of leaks across the latter half of 2025. Yesterday morning, an ASUS Vivobook S 15 (M5650GA) test platform posted further "Gorgon Point" details within the Geekbench browser database. The sort of next-gen ultra-thin laptop was driven by a Ryzen AI 9 465 APU. This particular SKU also turned up via a CrossMark entry, late last month. Looking at fresh Geekbench 6.5 scores, the evaluated ASUS device achieved overall 2780 (single-core) and 12001 (multicore) digits. When compared to a direct predecessor, the unreleased mobile chip only holds an advantage in single-core stakes.

It is possible that the Ryzen AI 9 465 mobile processor was put through its paces under unfavorable conditionsβ€”e.g. running on immature drivers and firmware. The direct forebear, Team Red's readily available Ryzen AI 9 365 "Strix Point" mobile processor holds a 474 point advantage in multicore scenarios. Industry observers reckon that the "Gorgon Point" generation will deploy with improved onboard XDNA 2 NPUs, albeit combined with familiar "Strix Point" credentials. The forthcoming Ryzen AI 9 465 SKU seems to share its predecessor's base specifications: 10 cores/20 threads, maximum 5.0 GHz boost/turbo clock, and a Radeon 880M iGPU. Amusingly, Geekbench has identified the leaked part as belonging to the "Strix Point" mobile familyβ€”it is widely believed that "Gorgon Point" will emerge as a mild refresh over the "previous" generation.

Rec'd – Discover places you'll love through people you trust


Rec'd is a social discovery platform, turning trusted social signals into personalised recommendations. Right now people use multiple apps to discover places, save them, verify them and book. Rec'd integrates this process into one, powerful, AI based app that lets people discover the way they want, saving into one clean and intelligent platform.

View startup

AI Liberation: Unlocking Potential Beyond β€œSecurity Theater”

16 December 2025 at 20:17

The post AI Liberation: Unlocking Potential Beyond β€œSecurity Theater” appeared first on StartupHub.ai.

The prevailing narrative around artificial intelligence often centers on the race for capability, but a recent discussion on the Latent Space podcast unveiled a contrasting, equally vital perspective: the imperative of liberation and radical transparency in AI development. Pliny the Liberator, renowned for his β€œuniversal jailbreaks” that dismantle the guardrails of frontier models, and John […]

The post AI Liberation: Unlocking Potential Beyond β€œSecurity Theater” appeared first on StartupHub.ai.

Greylock’s Enduring Legacy: People, Principles, and the AI Frontier

16 December 2025 at 19:48

The post Greylock’s Enduring Legacy: People, Principles, and the AI Frontier appeared first on StartupHub.ai.

Greylock Partners, a venture capital firm celebrating its 60th anniversary this year, offers a compelling study in enduring success through relentless adaptation and an unwavering commitment to core principles. In a recent episode of Uncapped with Jack Altman, General Partner Saam Motamedi, one of Greylock’s youngest partners, delved into the foundational elements that have allowed […]

The post Greylock’s Enduring Legacy: People, Principles, and the AI Frontier appeared first on StartupHub.ai.

What We Learned Deploying AI within Bloomberg’s Engineering Organization – Lei Zhang, Bloomberg

16 December 2025 at 19:16

The post What We Learned Deploying AI within Bloomberg’s Engineering Organization – Lei Zhang, Bloomberg appeared first on StartupHub.ai.

β€œThe reality of applying AI at scale inside a mature engineering organization is far more complex and nuanced,” stated Lei Zhang, Head of Technology Infrastructure Engineering at Bloomberg, during a recent discussion. Zhang, speaking about Bloomberg’s extensive experience integrating AI into the workflows of over 9,000 software engineers, offered a candid look at the practical […]

The post What We Learned Deploying AI within Bloomberg’s Engineering Organization – Lei Zhang, Bloomberg appeared first on StartupHub.ai.

The AI Memory Wars Heat Up Around Video

16 December 2025 at 19:00

The post The AI Memory Wars Heat Up Around Video appeared first on StartupHub.ai.

Video has become a dominant signal on the internet. It powers everything from Netflix’s $82 billion Warner Bros Discovery acquisition to the sensor streams feeding warehouse robots and city surveillance grids. Yet beneath this sprawl, AI systems are hitting a wall in that they can tag clips and rank highlights, but they struggle to remember […]

The post The AI Memory Wars Heat Up Around Video appeared first on StartupHub.ai.

New EWA study results suggest on-demand pay boosts income

16 December 2025 at 18:57

The post New EWA study results suggest on-demand pay boosts income appeared first on StartupHub.ai.

Independent EWA study results analyzing over one million EarnIn users suggest that flexible access to earned wages increases monthly income by 11.5 percent.

The post New EWA study results suggest on-demand pay boosts income appeared first on StartupHub.ai.

The enterprise blueprint for winning visibility in AI search

16 December 2025 at 20:00
The enterprise blueprint for winning visibility in AI search

We are navigating the β€œsearch everywhere” revolution – a disruptive shift driven by generative AI and large language models (LLMs) that is reshaping the relationship between brands, consumers, and search engines.

For the last two decades, the digital economy ran on a simple exchange: content for clicks.Β 

With the rise of zero-click experiences, AI Overviews, and assistant-led research, that exchange is breaking down.

AI now synthesizes answers directly on the SERP, often satisfying intent without a visit to a website.Β 

Platforms such as Gemini and ChatGPT are fundamentally changing how information is discovered.Β 

For enterprises, visibility increasingly depends on whether content is recognized as authoritative by both search engines and AI systems.

That shift introduces a new goal – to become the source that AI cites.

A content knowledge graph is essential to achieving that goal.Β 

By leveraging structured data and entity SEO, brands can build a semantic data layer that enables AI to accurately interpret their entities and relationships, ensuring continued discoverability in this evolving economy.

This article explores:

  • The difference between traditional search and AI search, including the concept of comprehension budget.
  • Why schema and entity optimization are foundational to discovery in AI search.
  • The content knowledge graph and the importance of organizational entity lineage.
  • The enterprise entity optimization playbook and deployment checklist.
  • The role of schema in the agentic web.
  • How connected journeys improve customer discovery and total cost of ownership.

The fundamental difference between traditional and AI search

To become a source that AI cites, it’s essential to understand how traditional search differs from AI-driven search.

Traditional search functioned much like software as a service.Β 

It was deterministic, following fixed, rule-based logic and producing the same output for the same input every time.

AI search is probabilistic.Β 

It generates responses based on patterns and likelihoods, which means results can vary from one query to the next.Β 

Even with multimodal content, AI converts text, images, and audio into numerical representations that capture meaning and relationships rather than exact matches.

For AI to cite your content, you need a strong data layer combined with context engineering – structuring and optimizing information so AI can interpret it as reliable and trustworthy for a given query.

As AI systems rely increasingly on large-scale inference rather than keyword-driven indexing, a new reality has emerged: the cost of comprehension.Β 

Each time an AI model interprets text, resolves ambiguity, or infers relationships between entities, it consumes GPU cycles, increasing already significant computing costs.

A comprehension budget is the finite allocation of compute that determines whether content is worth the effort for an AI system to understand.

4 foundational elements for AI discovery

For content to be cited by AI, it must first be discovered and understood.Β 

While many discovery requirements overlap with traditional search, key differences emerge in how AI systems process and evaluate content.

AI discovery - foundational elements

1. Technical foundation

Your site’s infrastructure must allow AI engines to crawl and access content efficiently.Β 

With limited compute and a finite comprehension budget, platform architecture matters.Β 

Enterprises should support progressive crawling of fresh content through IndexNow integration to optimize that budget.

Ideally, this capability is native to the platform and CMS.

2. Helpful content

Before creating content, you need an entity strategy that accurately and comprehensively represents your brand.Β 

Content should meet audience needs and answer their questions.Β 

Structuring content around customer intent, presenting it in clear β€œchunks,” and keeping it fresh are all important considerations.

Dig deeper: Chunk, cite, clarify, build: A content framework for AI search

3. Entity optimization

Schema markup, clean information architecture, consistent headings, and clear entity relationships help AI engines understand both individual pages and how multiple pieces of content relate to one another.Β 

Rather than forcing models to infer what a page is about, who it applies to, or how information connects, businesses make those relationships explicit.

4. Authority

AI engines, like traditional search engines, prioritize authoritative content from trusted sources.Β 

Establishing topical authority is essential. For location-based businesses, local relevance and authority are also critical to becoming a trusted source.

The myth: Schema doesn’t work

Many enterprises claim to use schema but see no measurable lift, leading to the belief that schema doesn’t work.Β 

The reality is that most failures stem from basic implementations or schema deployed with errors.

Tags such as Organization or Breadcrumb are foundational, but they provide limited insight into a business.Β 

Used in isolation, they create disconnected data points rather than a cohesive story AI can interpret.

The content knowledge graph: Telling AI your story

The more AI knows about your business, the better it can cite it.Β 

A content knowledge graph is a structured map of entities and their relationships, providing reliable information about your business to AI systems.

Deep nested schema plays a central role in building this graph.

entity-lineage-for-deep-nested-schema

A deep nested schema architecture expresses the full entity lineage of a business in a machine-readable form.

In resource description framework (RDF) terms, AI systems need to understand that:

  • An organization creates a brand.
  • The brand manufactures a product.
  • The product belongs to a category.
  • Each category serves a specific purpose or use case.

By fully nesting entities – Organization β†’ Brand β†’ Product β†’ Offer β†’ PriceSpecification β†’ Review β†’ Person – you publish a closed-loop content knowledge graph that models your business with precision.

Dig deeper: 8 steps to a successful entity-first strategy for SEO and content

Get the newsletter search marketers rely on.


The enterprise entity optimization playbook

In β€œHow to deploy advanced schema at scale,” I outlined the full process for effective schema deployment – from developing an entity strategy through deployment, maintenance, and measurement.

Automating for operational excellence

At the enterprise level, facts change constantly, including product specifications, availability, categories, reviews, offers, and prices.Β 

If structured data, entity lineage, and topic clusters do not update dynamically to reflect these changes, AI systems begin to detect inconsistencies.

In an AI-driven ecosystem where accuracy, coherence, and consistency determine inclusion, even small discrepancies can erode trust.

Manual schema management is not sustainable.

The only scalable approach is automation – using a schema management solution aligned with your entity strategy and integrated into your discovery and marketing flywheel.

Measuring success: KPIs for the generative AI era

As keyword rankings lose relevance and traffic declines, you need new KPIs to evaluate performance in AI search.

  • Brand visibility: Is your brand appearing in AI search results?
  • Brand sentiment: When your brand is cited, is the sentiment positive, negative, or neutral?
  • LLM visibility: Beyond branded queries, how does your performance on non-branded terms compare with competitors?
  • Conversions: At the bottom of the funnel, are conversion metrics being tracked and optimized?

Dig deeper: 7 focus areas as AI transforms search and the customer journey in 2026

From reading to acting: Preparing for the agentic web

The web is shifting from a β€œread” model to an β€œact” model.

AI agents will increasingly execute tasks on behalf of users, such as booking appointments, reserving tables, or comparing specifications.

To be discovered by these agents, brands must make their capabilities machine-callable. Key steps to prepare include:

  • Create a schema layer: Define entity lineage and executable capabilities in a machine-readable format so agents can act on your behalf.
  • Use action vocabularies: Leverage Schema.org action vocabularies to provide semantic meaning and define agent capabilities, including:
    • ReserveAction.
    • BookAction.
    • CommunicateAction.
    • PotentialAction.
  • Establish guardrails: Declare engagement rules, required inputs, authentication, and success or failure semantics in a structured format that machines can interpret.

Brands that are callable are the ones that will be found. Acting early provides a compounding advantage by shaping the standards agents learn first.

The enterprise entity deployment checklist

Use this checklist to evaluate whether your entity strategy is operational, scalable, and aligned with AI discovery requirements.

  • Entity audit: Have you defined your core entities and validated the facts?
  • Deep nesting: Does your JSON-LD reflect your business ontology, or is it flat?
  • Authority linking: Are you using sameAs to connect entities to Wikidata and the Knowledge Graph?
  • Actionable schema: Have you implemented PotentialAction for the agentic web?
  • Automation: Do you have a system in place to prevent schema drift?
  • Single source of truth (SSOT): Is schema synchronized across your CMS, GBP, and internal systems?
  • Technical SEO: Are the technical foundations in place to support an effective entity strategy?
  • IndexNow: Are you enabling progressive and rapid indexing of fresh content?

Connected customer journeys and total cost of ownership

connected-customer-discovery-flywheel

Your martech stack must align with the evolving customer discovery journey.Β 

This requires a shift from treating schema as a point solution for visibility to managing a holistic presence with total cost of ownership in mind.

Data is the foundation of any composable architecture.Β 

A centralized data repository connects technologies, enables seamless flow, breaks down departmental silos, and optimizes cost of ownership.

This reduces redundancy and improves the consistency and accuracy AI systems expect.

When schema is treated as a point solution, content changes can break not only schema deployment but the entire entity lineage.Β 

Fixing individual tags does not restore performance. Instead, multiple teams – SEO, content, IT, and analytics – are pulled into investigations, increasing cost and inefficiency.

The solution is to integrate schema markup directly into brand and entity strategy.

When structured content changes, it should be:

  • Revalidated against the organization’s entity lineage.
  • Dynamically redeployed.
  • Pushed for progressive indexing through IndexNow.

This enables faster recovery and lower compute overhead.

Integrating schema into your entity lineage and discovery flywheel helps optimize total cost of ownership while maximizing efficiency.

A strategic blueprint for AI readiness

Several core requirements define AI readiness.

ai-ready-enterprise-strategy
  • Data: Centralized, unified, consistent, and reliable data aligned to customer intent is the foundation of any AI strategy.
  • Connected journeys and composable architecture: When data is unified and structured with schema, customer journeys can be connected across channels. A composable martech stack enables consistent, personalized experiences at every touchpoint.
  • Structured content: Define organizational entity lineage and create a semantic layer that makes content machine- and agent-ready.
  • Distribution: Break down silos and move from channel-specific tactics to an omnichannel strategy, supported by a centralized data source and progressive crawling of fresh content.

Together, these efforts make your omnichannel strategy more durable while reducing total cost of ownership across the technology stack.

Thanks to Bill Hunt and Tushar Prabhu for their contributions to this article.

When Google’s AI bidding breaks – and how to take control

16 December 2025 at 19:00
When Google’s AI bidding breaks – and how to take control

Google’s pitch for AI-powered bidding is seductive.

Feed the algorithm your conversion data, set a target, and let it optimize your campaigns while you focus on strategy.Β 

Machine learning will handle the rest.

What Google doesn’t emphasize is that its algorithms optimize for Google’s goals, not necessarily yours.Β 

In 2026, as Smart Bidding becomes more opaque and Performance Max absorbs more campaign types, knowing when to guide the algorithm – and when to override it – has become a defining skill that separates average PPC managers from exceptional ones.

AI bidding can deliver spectacular results, but it can also quietly destroy profitable campaigns by chasing volume at the expense of efficiency.Β 

The difference is not the technology. It is knowing when the algorithm needs direction, tighter constraints, or a full override.

This article explains:

  • How AI bidding actually works.
  • The warning signs that it is failing.
  • The strategic intervention points where human judgment still outperforms machine learning.

How AI bidding actually works – and what Google doesn’t tell you

Smart Bidding comes in several strategies, including:

Each uses machine learning to predict the likelihood of a conversion and adjust bids in real time based on contextual signals.

The algorithm analyzes hundreds of signals at auction time, such as:

  • Device type.
  • Location.
  • Time of day.
  • Browser.
  • Operating system.
  • Audience membership.
  • Remarketing lists.
  • Past site interactions.
  • Search query.

It compares these signals with historical conversion data to calculate an optimal bid for each auction.

During the β€œlearning period,” typically seven to 14 days, the algorithm explores the bid landscape, testing bid levels to understand the conversion probability curve.Β 

Google recommends patience during this phase, and in general, that advice holds. The algorithm needs data.

The first problem is that learning periods are not always temporary.Β 

Some campaigns get stuck in perpetual learning and never achieve stable performance.

Dig deeper: When to trust Google Ads AI and when you shouldn’t

Google’s optimization goals vs. your business goals

The algorithm optimizes for metrics that drive Google’s revenue, not necessarily your profitability.

When a Target ROAS of 400% is set, the algorithm interprets that as β€œmaximize total conversion value while maintaining a 400% average ROAS.” 

Notice the word β€œmaximize.”

The system is designed to spend the full budget and, ideally, encourage increases over time.Β 

More spend means more revenue for Google.

Business goals are often different.Β 

You may want a 400% ROAS with a specific volume threshold.Β 

You may need to maintain margin requirements that vary by product line.Β 

Or you may prefer a 500% ROAS at lower volume because fulfillment capacity is constrained.

The algorithm does not understand this context.Β 

It sees a ROAS target and optimizes accordingly, often pushing volume at the expense of efficiency once the target is reached.

This pattern is common. An algorithm increases spend by 40% to deliver 15% more conversions at the target ROAS. Technically, it succeeds.Β 

In practice, cash flow cannot support the higher ad spend, even at the same efficiency.Β 

The algorithm does not account for working capital constraints.

Key signals the algorithm can’t understand

AI bidding works well, but it has limits.Β 

Without intervention, several factors can’t be fully accounted for.

Seasonal patterns not yet reflected in historical data

Launch a campaign in October, and the algorithm has no visibility into a December peak season.

It optimizes based on October performance until December data proves otherwise, often missing early seasonal demand.

Product margin differences

A $100 sale of Product A with a 60% margin and a $100 sale of Product B with a 15% margin look identical to the algorithm.Β 

Both register as $100 conversions. The business impact, however, is very different.Β 

This is where profit tracking, profit bidding, and margin-based segmentation matter.

Customer lifetime value variations

Unless lifetime value modeling is explicitly built into conversion values, the algorithm treats a first-time customer the same as a repeat buyer.Β 

In most accounts, that modeling does not exist.

Market and competitive changes

When a competitor launches an aggressive promotion or a new entrant appears, the algorithm continues bidding based on historical conditions until performance degrades enough to force adjustment.Β 

Market share is often lost during that lag.

Inventory and supply chain constraints

If a best-selling product is out of stock for two weeks, the algorithm may continue bidding aggressively on related searches because of past performance.Β 

The result is paid traffic that cannot convert.

This is not a criticism of the technology. It’s a reminder that the algorithm optimizes only within the data and parameters provided.Β 

When those inputs fail to reflect business reality, optimization may be mathematically correct but strategically wrong.

Warning signs your AI bidding strategy is failing

The perpetual learning phase

Learning periods are normal. Extended learning periods are red flags.

If your campaign shows a β€œLearning” status for more than two weeks, something is broken.Β 

Common causes include:

  • Insufficient conversion volume – the algorithm typically needs at least 30 to 50 conversions per month.
  • Frequent changes that reset the learning period.
  • Unstable performance with wide day-to-day fluctuations.

When to intervene

If learning extends beyond three weeks, either:

  • Increase the budget to accelerate data collection.
  • Loosen the target to allow more conversions.
  • Or switch to a less aggressive bid strategy like Enhanced CPC.Β 

Sometimes the algorithm is simply telling you it does not have enough data to succeed.

Budget pacing issues

Healthy AI bidding campaigns show relatively smooth budget pacing.Β 

Daily spend fluctuates, but it stays within reasonable bounds.Β 

Problematic patterns include:

  • Front-loaded spending – 80% of the daily budget gone by 10 a.m.
  • Consistent underspending, such as averaging 60% of budget per day.
  • Volatile day-to-day swings, like spending $800 one day, $200 the next, then $650 after that.

Budget pacing is a proxy for algorithm confidence.Β 

Smooth pacing suggests the system understands your conversion landscape.Β 

Erratic pacing usually means it is guessing.

The efficiency cliff

This is the most dangerous pattern. Performance starts strong, then gradually or suddenly deteriorates.

This shows up often in Target ROAS campaigns.Β 

  • Month 1: 450% ROAS, excellent.Β 
  • Month 2: 420%, still good.Β 
  • Month 3: 380%, concerning.Β 
  • Month 4: 310%, alarm bells.

What happened?Β 

The algorithm exhausted the most efficient audience segments and search terms.Β 

To keep growing volume – because it is designed to maximize – it expanded into less qualified traffic.Β 

Broad match reached further. Audiences widened. Bid efficiency declined.

Traffic quality deterioration

Sometimes the numbers look fine, but qualitative signals tell a different story.Β 

  • Engagement declines – bounce rate rises, time on site falls, pages per session drop.Β 
  • Geographic shifts appear as the algorithm drives traffic from lower-value regions.Β 
  • Device mix changes, often skewing toward mobile because CPCs are cheaper, even when desktop converts better.Β 
  • Time-of-day misalignment can also emerge, with traffic arriving when sales teams are unavailable.

These quality signals do not directly influence optimization because they are not part of the conversion data.Β 

To address them, the algorithm needs constraints: bid adjustments, audience exclusions, or ad scheduling.

The search terms report reveals the truth

The search terms report is the truth serum for AI bidding performance.Β 

Export it regularly and look for:

  • Low-intent queries receiving aggressive bids.
  • Informational searches mixed with transactional ones.
  • Irrelevant expansions where the algorithm chased conversions into entirely different intent.

A high-end furniture retailer should not spend $8 per click on β€œfree furniture donation pickup.” 

A B2B software company targeting β€œproject management software” should not appear for β€œproject manager jobs.” 

These situations occur when the algorithm operates without constraints.Β 

Keyword matching is also looser than it was in the past, which means even small gaps can allow the system to bid on queries you never intended to target.

Dig deeper: How to tell if Google Ads automation helps or hurts your campaigns

Get the newsletter search marketers rely on.


Strategic intervention points: When and how to take control

Segmentation for better control

One-size-fits-all AI bidding breaks down when a business has diverse economics.Β 

The solution is segmentation, so each algorithm optimizes toward a clear, coherent goal.

Separate high-margin products – 40%+ margin – into one campaign with more aggressive ROAS targets, and low-margin products – 10% to 15% margin – into another with more conservative targets.Β 

If the Northeast region delivers 450% ROAS while the Southeast delivers 250%, separate them.Β 

Brand campaigns operate under fundamentally different economics than nonbrand campaigns, so optimizing both with the same algorithm and target rarely makes sense.

Segmentation gives each algorithm a clear mission. Better focus leads to better results.

Bid strategy layering

Pure automation is not always the answer.Β 

In many cases, hybrid approaches deliver better results.

  • Run Target ROAS at 400% under normal conditions, then manually lower it to 300% during peak season to capture more volume when demand is high.Β 
  • Use Maximize Conversion Value with a bid cap if unit economics cannot support bids above $12.Β 
  • Group related campaigns under a portfolio Target ROAS strategy so the algorithm can optimize across them.Β 
  • For campaigns with limited conversion data or volatile performance, Enhanced CPC offers algorithmic assistance without full black box automation.

The hybrid approach

The most effective setups combine AI bidding with manual control campaigns.

Allocate 70% of the budget to AI bidding campaigns, such as Target ROAS or Maximize Conversion Value, and 30% to Enhanced CPC or manual CPC campaigns.Β 

Manual campaigns act as a baseline. If AI underperforms manual by more than 20% after 90 days, the algorithm is not working for the business.

Use tightly controlled manual campaigns to capture the most valuable traffic – brand terms and high-intent keywords – while AI campaigns handle broader prospecting and discovery.Β 

This approach protects the core business while still exploring growth opportunities.

COGS and cart data reporting (plus profit optimization beta)

Google now allows advertisers to report cost of goods sold, or COGS, and detailed cart data alongside conversions.Β 

This is not about bidding yet, but seeing true profitability inside Google Ads reporting.

Most accounts optimize for revenue, or ROAS, not profit.Β 

A $100 sale with $80 in COGS is very different from a $100 sale with $20 in COGS, but standard reporting treats them the same.Β 

With COGS reporting in place, actual profit becomes visible, dramatically improving the quality of performance analysis.

To set it up, conversions must include cart-level parameters added to existing tracking.Β 

These typically include item ID, item name, quantity, price, and, critically, the cost_of_goods_sold parameter for each product.

Google is testing a bid strategy that optimizes for profit instead of revenue.Β 

Access is limited, but advertisers with clean COGS data flowing into Google Ads can request entry.Β 

In this model, bids are optimized around actual profit margins rather than raw conversion value.Β 

This is especially powerful for retailers with wide margin variation across products.

For advertisers without access to the beta, a custom margin-tracking pixel can be implemented manually. It is more technical to set up, but it achieves the same outcome.

Dig deeper: Margin-based tracking: 3 advanced strategies for Google Shopping profitability

When AI bidding actually works

AI bidding works best when the fundamentals are in place:Β 

  • Sufficient conversion volume.
  • A stable business model with consistent margins and predictable seasonality.
  • Clean conversion tracking.
  • Enough historical data to support learning.

In these conditions, AI bidding often outperforms manual management by processing more signals and making more granular optimizations than humans can execute at scale.

This tends to be true in:

  • Mature ecommerce accounts.
  • Lead generation programs with consistent lead values.
  • SaaS models with predictable trial-to-paid conversion paths.

When those conditions hold, the role shifts.

Bid management gives way to strategic oversight – monitoring trends, identifying expansion opportunities, and testing new structures.

The algorithm then handles tactical optimization.

Preparing for AI-first advertising

Google is steadily reducing advertiser control under the banner of automation.Β 

  • Performance Max has absorbed Smart Shopping and Local campaigns.Β 
  • Asset groups replace ad groups.Β 
  • Broad match becomes mandatory in more contexts.Β 
  • Negative keywords increasingly function as suggestions the system may or may not honor.

For advertisers with complex business models or specific strategic goals, this loss of granularity creates tension.Β 

You are often asked to trust the algorithm even when business context suggests a different decision.

That shift changes the role. You are no longer a bid manager.Β 

You are an AI strategy director who:

  • Defines objectives.
  • Provides business context.
  • Sets constraints.
  • Monitors outcomes.
  • Intervenes when the system drifts away from strategic intent.

No matter how advanced AI bidding becomes, certain decisions still require human judgment.Β 

Strategic positioning – which markets to enter and which product lines to emphasize – cannot be automated.Β 

Neither can creative testing, competitive intelligence, or operational realities like inventory constraints, margin requirements, and broader business priorities.

This is not a story of humans versus AI. It is humans directing AI.

Dig deeper: 4 times PPC automation still needs a human touch

Master the algorithm, don’t serve it

AI-powered bidding is the most powerful optimization tool paid media has ever had.Β 

When conditions are right – sufficient data, a stable business model, and clean tracking – it delivers results manual management cannot match.

But it is not magic.

The algorithm optimizes for mathematical targets within the data you provide.Β 

If business context is missing from that data, optimization can be technically correct and strategically wrong.Β 

If markets change faster than the system adapts, performance erodes.Β 

If your goals diverge from Google’s revenue incentives, the algorithm will pull in directions that do not serve the business.

The job in 2026 is not to blindly trust automation or stubbornly resist it.Β 

It is to master the algorithm – knowing when to let it run, when to guide it with constraints, and when to override it entirely.

The strongest PPC leaders are AI directors. They do not manage bids. They manage the system that manages bids.

Rogue NuGet Package Poses as Tracer.Fody, Steals Cryptocurrency Wallet Data

Cybersecurity researchers have discovered a new malicious NuGet package that typosquats and impersonates the popular .NET tracing library and its author to sneak in a cryptocurrency wallet stealer. The malicious package, named "Tracer.Fody.NLog," remained on the repository for nearly six years. It was published by a user named "csnemess" on February 26, 2020. It masquerades as "Tracer.Fody,"

Apple TV adds support for Google Cast even as Netflix ditches the feature

16 December 2025 at 19:45
The tech giant's adoption of Google Cast signals its interest in making Apple TV a more competitive streaming service at a time when the market is seeing increased consolidation.

(PR) Ubisoft Absorbs March of Giants IP & Dev Team, Formerly an Amazon Montreal Game

16 December 2025 at 20:04
Ubisoft has announced that it has acquired March of Giants, formerly developed by Amazon's Montreal Games studio. The team will join Ubisoft to continue development of the free-to-play title, and is currently working on the next major update. March of Giants puts players in some big shoes, placing them in the role of a giant combatant on a ravaged urban battlefield inspired by the technology of the early 1900s. The 4v4 tactical multiplayer online battle arena (MOBA) sees players fighting through lanes, commanding battalions of comparatively pint-sized enemies, and deploying Battleworksβ€”such as trenches, tanks, turrets, and bunkersβ€”to eliminate the enemy team. The gameplay combines elements of real-time strategy and lane-based combat with large-scale coordination and mastery of the giant heroes that make up its roster.

The team behind the game consists of a group of veteran developers, led by two former Ubisoft team members with ties to some of the company's biggest franchises. Alex Parizeau, Senior Production lead, is the former Managing Director of Ubisoft Toronto, having overseen several major AAA productions. Xavier Marquis, Creative Director, was the original Creative Director behind Rainbow Six Siege. Both bring their wealth of experience, passion, and expertise to the development of this new title.

Beelink ME Pro Teaser Showcases DIY Swappable AMD, Arm, & Intel Board System

16 December 2025 at 19:53
Beelink has started marketing a slick-looking ME Pro NAS system; past weekend updatesβ€”mostly aimed at the Chinese marketβ€”contain only scant details. Western audiences were treated to a 35-second-long teaser trailer that shows off some of the next-gen device's internals. The Shenzhen-based manufacturer advertises the ME Pro model as a NAS/mini PC hybridβ€”the standard version seems to accommodate two drives. In the recent past, Beelink has hinted about a larger 4-bay variant. According to the company's Weibo blog, the ME Pro's unique selling point (USP) is: "a groundbreaking DIY drawer-style design, modularizing the motherboard and allowing you to upgrade the core components as needed, truly realizing 'My device, my control,' solving big problems with a small investment."

Beelink's promo video shows AMD, Arm, and Intel board options being swapped in and out of the compact hybrid NAS/PC system. Elaborating on Team Blue hardware, one press outletβ€”NAS Comparesβ€”reportedly attended a recent special preview event at Beelink's headquarters. Representatives allegedly outlined board designs that featured Intel N95 "Alder Lake-N" and N150 "Twin Lake" processors. The manufacturer has not released an ME Pro specification sheet, so it is difficult to judge the product's proportions. In a second Weibo post, Beelink's description reads as follows: "(our design) uses a unibody metal casing, replacing the traditional multi-component structure, which significantly reduces size while improving structural strength. Compared to mainstream dual-bay NAS devices, the overall size is reduced by nearly 50%, making it more suitable for desktop and home environments." A vague "coming soon" tagline could hint at an eventual proper unveiling at next month's CES trade show.

Minisforum Intros BD895i SE MoDT ITX Motherboard with Ryzen 9 8945HX

16 December 2025 at 19:20
Minisforum has introduced the BD895i SE, a new MoDT (Mobile on Desktop) Mini-ITX motherboard built around AMD's Ryzen 9 8945HX mobile processor. Similar in concept to the earlier BE7XXi HX series, the BD895i SE targets compact, high-performance desktop builds by combining a flagship mobile CPU with a full-featured ITX (170 Γ— 170 mm) layout and a large, pre-installed CPU heatsink. The board features a 16-core, 32-thread Ryzen 9 8945HX processor, running at a 2.5 GHz base clock with boost speeds of up to 5.4 GHz. The CPU is rated at a 100 W TDP, with boost power reaching up to 120 W. Minisforum included a low-profile heatsink with four heatpipes and a total height of 37 mm leaving users the choice to add their own 120 mm fan if needed.

The BD895i SE motherboard features a full PCIe 5.0 x16 slot for a dedicated graphics card, as the integrated AMD Radeon 610M is not really suitable for gaming. Then there are two DDR5-5200 SO-DIMM slots supporting up to 96 GB and dual PCIe 4.0 x4 M.2 2280 slots for NVMe storage. Another M.2 2230 E-Key slot is reserved for Wi-Fi or Bluetooth cards. In terms of connectivity, it offers HDMI 2.1, DisplayPort 1.4, and USB-C with DisplayPort Alt Mode (all up to 8K at 60 Hz or 4K at 120 Hz). A 2.5 GbE LAN port alongside a mix of USB 2.0, USB 3.2 Gen 1, and USB-C Gen 2 ports and full analog audio jacks on the rear I/O complements the specs. Power is supplied through standard ATX connectors, making the board compatible with conventional desktop PSUs. Pricing is set at around $423.90 in the US (regular price $529), while European pricing is listed at €439 (regular price €549). Minisforum BD895i SE is positioned as a compact but premium option for users looking to build a powerful small-form-factor system around AMD latest high-end mobile CPU.

(PR) Aspyr Postpones Launch of Deus Ex Remastered, Refunds Existing Pre-orders

16 December 2025 at 19:20
Thank you to the community for your feedback following the reveal of Deus Ex Remastered. We've listened to what you had to say, and in order to better meet fan expectations and deliver the best possible experience for players, Deus Ex Remastered will no longer launch on Feb. 5, 2026. We look forward to sharing more updates with you when the time comes. All existing pre-orders will be fully refunded. Thank you again for your patience and support.

About Deus Ex Remastered
The year is 2052. Societies are teetering on collapse, plagues spread unchecked, and shadow governments shape the future. You are JC Dentonβ€”a nano-augmented UNATCO operative tasked with protecting order. But the deeper you dig, the more lies you uncover. What begins as a mission becomes a fracture point in human history.

Apple's "Baltra" AI Chip Reportedly Launching in 2027, Could be Inference-only Design

16 December 2025 at 19:10
About a year ago, industry insiders let slip early details regarding Apple's mysterious "Baltra" AI chip. At the time, Broadcom was reportedly assisting with the design's network technology. In addition, the "Baltra" project was linked to TSMC 3 nm N3E or N3P node processes. Recent leaks suggest an adjusted gestation schedule; instead of a predicted start of "mass production by 2026," timelines seem to be extended by another year. The reasons behind this rumored delay are not clear, but the North American giant hasβ€”officially, as of October 2025β€”bolstered its Apple Intelligence and Private Cloud Compute-oriented data centers with US-made advanced server equipment, likely running on familiar M2 Ultra processors. Industry moles reckon that M4-based variants are being readied for distribution.

Apple's futuristic in-house "Baltra" AI setup is expected to be inference-oriented, as predicted by Max Weinbach. The tech industry analyst imagined: "basically, I doubt they'll do a massive cluster, but maybe something closer to a GB300 style, with ~64 chips all to all with larger high bandwidth LPDDR memory. Should be significantly cheaper than most current chips, and match the needs." Last month, Mark Gurman claimed that Apple had signed up for a custom Google Gemini-powered LLM. The notorious tipster believes that the claimed $1 billion per year deal will drive next-gen personalized Siri and AI web search functions. As mentioned before, Apple's in-progress "Baltra" data center chip could be designed for efficiency and cost-effectiveness, rather than the raw processing power required for training.

(PR) Arctic Announces MX-7 Thermal Interface Material

16 December 2025 at 18:11
ARCTIC introduces MX-7, a new thermal paste that delivers reliable performance in both classic desktop CPU setups and demanding direct die applications on GPUs, laptops, and console processors. Thanks to a new formula, MX-7 achieves even lower thermal resistance and offers a measurable increase in performance. The thermal paste is neither electrically conductive nor capacitive, eliminating the risk of short circuits or discharges.

MX-7 is a relatively viscous thermal paste that cannot be applied with a spatula. An even layer thickness is achieved through contact pressure when mounting the cooler. The MX-7 impresses not only with its performance, but also with its new syringe design. The improved composition with high cohesiveness reduces the pump-out effect during repeated heat cyclesβ€”ideal for long-lasting performance. The Authenticity Check also allows the authenticity of each individual product to be verified.

PayPal wants to become a bank in the US

16 December 2025 at 19:27

The San Jose-based firm submitted applications to both the Federal Deposit Insurance Corporation and the Utah Department of Financial Institutions to form an industrial loan company chartered in Utah. The proposed entity, to be known as PayPal Bank, would allow the company to expand its small-business lending operations and offer...

Read Entire Article

Billionaire CRE Developer Warns of Data Center Finance Risks

16 December 2025 at 18:17

The post Billionaire CRE Developer Warns of Data Center Finance Risks appeared first on StartupHub.ai.

The burgeoning demand for artificial intelligence has ignited a gold rush in data center development, but for seasoned commercial real estate (CRE) billionaire Fernando De Leon, this frenetic activity bears unsettling resemblances to past market excesses. De Leon, CEO of Leon Capital Group, recently spoke with CNBC Senior Real Estate Correspondent Diana Olick on the […]

The post Billionaire CRE Developer Warns of Data Center Finance Risks appeared first on StartupHub.ai.

Lazard CEO: U.S. economy increasingly a levered bet on AI

16 December 2025 at 17:45

The post Lazard CEO: U.S. economy increasingly a levered bet on AI appeared first on StartupHub.ai.

Lazard CEO Peter Orszag joined CNBC’s β€œSquawk Box” to discuss the current state of the economy, the impact of the AI boom, and the broader implications for businesses and employment. He articulated a bifurcated economic landscape where AI-driven sectors are experiencing significant growth, contrasted with other areas that are not seeing the same level of […]

The post Lazard CEO: U.S. economy increasingly a levered bet on AI appeared first on StartupHub.ai.

Data 360 Powers Trusted AI: A New Foundation for ISVs

16 December 2025 at 17:18

The post Data 360 Powers Trusted AI: A New Foundation for ISVs appeared first on StartupHub.ai.

Salesforce's Data 360, now generally available, provides ISVs with a critical unified data foundation essential for developing and deploying trusted AI solutions.

The post Data 360 Powers Trusted AI: A New Foundation for ISVs appeared first on StartupHub.ai.

A 3-tier framework for Shopify integrations that drive conversions

16 December 2025 at 18:00
A 3-tier framework for Shopify integrations that drive conversions

Shopify powers more than 6 million live ecommerce websites, supported by a robust app ecosystem that can extend nearly every part of the customer journey.Β 

Anyone can develop an app to perform virtually any function.Β 

But with so many integrations to choose from, ecommerce teams often waste time testing add-ons that promise revenue gains but fail to deliver.

Having worked across a wide range of Shopify implementations, I’ve seen which tools consistently improve checkout completion, recover abandoned carts, and increase revenue.Β 

Based on that experience, I’ve organized the most effective integrations into three tiers by priority – so you can implement the essentials first, then move on to more advanced optimization.

Tier 1: Mobile-first, frictionless buying

With 54.5% of holiday purchases happening on mobile, the ecommerce experience must be seamless and flexible.Β 

As a result, every Shopify site should have two components integrated into its storefront:Β 

  • A digital wallet compatibility.
  • A buy now, pay later (BNPL) option.Β 

Without these in place, Shopify users introduce unnecessary friction into the purchase journey and risk sending customers to competitors.Β 

The good news is that both components integrate natively with Shopify, requiring no custom development.

Why you need digital wallets

Digital wallets, such as Apple Pay, Google Pay, and PayPal, autofill delivery and payment information with a single click, eliminating the friction of typing on a small screen.Β 

This ease of use can shorten the purchase journey to just a few clicks between a social ad and checkout.

Adoption is accelerating. Up to 64% of Americans use digital wallets at least as often as traditional payment methods, and 54% use them more often.

Eliminate price objections with BNPL

Beyond payment convenience, customers also expect flexibility.Β 

BNPL providers, including Klarna and Afterpay, allow buyers to spread payments over time, reducing price objections at checkout.Β 

These options contributed $18.2 billion to online spending during last year’s holiday season – an all-time high, according to Adobe.

Together, digital wallets and BNPL form the foundation of a modern, mobile-first checkout experience.Β 

With these essentials in place, Shopify users can focus on tools that re-engage customers and bring them back to complete their purchases.

Dig deeper: The ultimate Shopify SEO and AI readiness playbook

Tier 2: The re-engagement power players

The second tier focuses on re-engagement – tools designed to bring back customers who have already shown intent.Β 

These integrations improve abandoned-cart recovery, increase repeat purchases, and build trust through social proof.

Re-engage customers with email and SMS

Email remains one of the most effective channels for re-engaging customers at every stage of the journey.Β 

Klaviyo and Attentive are strong options for Shopify users because both offer deep platform integration with minimal setup.

Both platforms also support SMS, allowing Shopify sellers to send automated text messages directly to customers’ mobile devices.Β 

SMS consistently delivers higher open, click-through, and conversion rates than email, making it especially effective for re-engagement use cases such as abandoned-cart recovery.

Together, these tools enable targeted campaigns and sophisticated automated flows that drive incremental revenue.Β 

However, CAN-SPAM and TCPA regulations require explicit opt-in for email and SMS marketing, respectively.Β 

As a result, sellers can only use these channels to contact customers who have agreed to receive marketing messages.

Use human-centered SMS outreach

While Attentive and Klaviyo effectively reach customers who have opted in to marketing, CartConvert helps sellers engage the 50% to 60% of shoppers who have not.Β 

The platform uses real people to contact cart abandoners via SMS. Because the outreach is not automated, TCPA restrictions do not apply.

CartConvert agents have live conversations with potential customers about their shopping experience.Β 

They are familiar with the products and can guide buyers back toward a purchase by suggesting alternatives or offering discounts.Β 

Running CartConvert alongside Klaviyo or Attentive ensures both subscribers and non-subscribers are included in re-engagement efforts.

Get the newsletter search marketers rely on.


Demonstrate social proof through reviews

Human-centered marketing also plays a role in building buyer confidence.Β 

Today’s online shoppers rely heavily on reviews when making purchasing decisions.Β 

When reviews are integrated directly into the shopping experience, they help establish trust and legitimacy, which in turn drive higher conversion rates.Β 

A product with five reviews is 270% more likely to be purchased than one with no reviews, research from the Spiegel Research Center at Northwestern University found.

Shopify users can choose from several review aggregators that pull Google reviews into product pages.Β 

Sellers should prioritize aggregators that also sync with Google Merchant Center, which powers Google Ads.Β 

Tools such as Okendo, Yotpo, and Shopper Approved integrate smoothly with both Shopify and Google’s ecosystem.

When reviews sync with Merchant Center, they can appear in Google Shopping ads, improving ad performance.Β 

While these tools add cost, they are also proven to generate incremental revenue that offsets the investment.

Dig deeper: How to make ecommerce product pages work in an AI-first world

Tier 3: Advanced optimization

The final tier includes more advanced integrations designed to help sellers optimize their sales funnel and performance at scale.

Attribution and analytics: Triple Whale

GA4’s changes to reporting, session logic, and interface have made attribution more difficult for many ecommerce teams.Β 

As a result, sellers are increasingly seeking clearer, independent performance insights.

Since 2023, Triple Whale has emerged as a leading alternative to Google Analytics, offering third-party attribution tools that integrate seamlessly with Shopify.Β 

The platform supports multiple attribution models – including first-click, last-click, and linear – along with cross-platform cost integration.

It also provides real-time data, which Google Analytics does not.Β 

This capability becomes especially valuable during high-pressure sales periods, such as Black Friday, when delayed reporting can lead to missed opportunities.

Although Triple Whale can cost up to $10,000 annually for mid-size brands, the improved data quality often justifies the investment for teams scaling paid acquisition.

Landing page customization: Replo

For sellers focused on improving conversion rates, landing page testing is essential.Β 

While Shopify is relatively easy to use, making changes to a live storefront for A/B testing carries the risk of breaking the site.

Replo allows Shopify users to build custom landing pages that can be tested at scale without coding.Β 

These pages typically provide a better user experience than default Shopify themes.Β 

It can also use site data to personalize landing pages based on a shopper’s browsing history.Β 

As a result, Replo-built pages often convert at higher rates than static site pages.

TikTok ads integration

TikTok continues to grow as a paid media channel, but it has traditionally presented a higher barrier to entry for advertisers.Β 

Previously, sellers needed an active TikTok account and could only purchase ads within the app, adding complexity and cost.

TikTok’s Shopify integration allows sellers to create ads that link directly to their websites, rather than keeping users inside the app.Β 

This change has lowered the barrier to entry and expanded access to the platform.Β 

Early testing shows promise for use cases such as cart abandonment, making the integration worth exploring despite its relative immaturity.

Dig deeper: Ecommerce SEO: Start where shoppers search

Prioritizing Shopify integrations for maximum impact

Shopify is a powerful platform for ecommerce, but maximizing results requires going beyond its default features.Β 

  • Start with essentials such as digital wallets and BNPL to reduce checkout friction.Β 
  • Then layer in email, SMS, and review integrations to re-engage interested shoppers.Β 
  • Finally, add analytics, attribution, and landing-page testing to optimize performance at scale.

Sellers do not need to implement every solution at once.Β 

Instead, conduct a quick audit of the existing stack against this framework, identify gaps, and prioritize the tools that improve conversion and re-engagement.Β 

Shopify’s flexibility is its greatest strength, and its app ecosystem enables sellers to turn more visitors into buyers.

Google says doing optimization for AI search is β€˜the same’ as doing SEO for traditional search

16 December 2025 at 17:22

Optimizing for AI search is β€œthe same” as optimizing for traditional search, Google SVP of Knowledge and Information Nick Fox said in a recent podcast. His advice was simple: build great sites with great content for your users.

More details. Fox made the point on the AI Inside podcast, during an interview with Jason Howell and Jeff Jarvis. Here is the transcript from the 22 minute mark:

Jarvis: β€œIs there guidance for enlightened publishers who want to be part of AI about how they should view, should they view their content in any way differently now?”

Fox: β€œThe short answer is no. The short answer is what you would have built and the way to optimize to do well in Google’s AI experiences is very similar, I would say the same, as how to perform well in traditional search. And it really does come down to build a great site, build great content. The way we put it is: build for users. Build what you would want to read, what you would want to access.”

Why we care. Many of you have been practicing SEO for many years, and now with this AI revolution in Search, you should know you are very well equipped to perform well in AI Search with many, if not all, of the skills you learned doing SEO. So have at it.

The video. Is AI Search Hurting The Open Web? With Google’s Nick Fox // AI InsideΒ #104

πŸ’Ύ

Build great sites, great content, for your users, according to Nick Fox, SVP of Knowledge and Information at Google.

Help us shape SMX Advanced 2026. You could win an All Access pass!

16 December 2025 at 17:00

We celebrated a major milestone in June: the return of SMX Advanced as an in-person event. It was our first since 2019.

More than a conference, SMX Advanced 2025 was a reunion. Search marketers from around the world came together to connect, exchange ideas, and learn the most current and advanced insights in search.

But search never stands still. With rapid shifts in AI SEO, constant algorithm changes, and the challenge of balancing generative AI with a human touch, the need for truly advanced, actionable education has never been greater.

Help shape SMX Advanced 2026

We’re committed to making the SMX Advanced 2026 program our most relevant, advanced, and exciting deep-dive experience yet. And we can’t do it without you – the expert community that makes this event legendary.

We’re inviting you to directly shape the curriculum for 2026.

Help us build a program that tackles the biggest challenges and opportunities on your radar by completing our short survey. Tell us:

  • What advanced topics are most critical to your professional growth right now.
  • Which recent search changes or complexities are keeping you up at night.
  • Which search industry experts and innovators you need to hear from.
  • Which session formats – from deep-dive clinics to lightning talks and interactive panels – will help you learn more and retain what you learn.

Fill out the survey here.

Be entered to win an All Access pass

To thank you for your time and insights, everyone who completes the survey will have the opportunity to enter an exclusive drawing.

One lucky participant will win a coveted All Access pass to SMX Advanced 2026, taking place June 3-5 at the Westin Boston Seaport.

Submit a session pitch

Beyond shaping the agenda, we also invite you to submit a session pitch. If you have a breakthrough strategy, an innovative case study, or next-level insights, this is your chance to help lead the industry conversation.

Read our guide to speaking at SMX for more details on how to submit a session idea. When you’re ready, create your profile and send us your session pitch.

We look forward to your submissions and insights! If you have any questions, feel free to reach out to me at kathy.bushman@semrush.com.

Amazon Exposes Years-Long GRU Cyber Campaign Targeting Energy and Cloud Infrastructure

Amazon's threat intelligence team has disclosed details of a "years-long" Russian state-sponsored campaign that targeted Western critical infrastructure between 2021 and 2025. Targets of the campaign included energy sector organizations across Western nations, critical infrastructure providers in North America and Europe, and entities with cloud-hosted network infrastructure. The activity has

India’s Digantara raises $50M for space-based missile defense tech

16 December 2025 at 18:44
With this funding, Digantara has raised $64.5 million in total as it moves beyond space situational awareness into missile tracking, citing growing demand from governments for space-based defense capabilities.

Databricks raises $4B at $134B valuation as its AI business heats up

16 December 2025 at 18:39
The data intelligence company has just raised more than $4 billion in a Series L funding round at a $134 billion valuation β€” up 34% from the $100 billion valuation that it achieved just three months ago.

Adobe Firefly now supports prompt-based video editing, adds more third-party models

16 December 2025 at 18:00
Adobe is updating its AI video-generation app, Firefly, with a new video editor that supports precise prompt-based edits, as well as adding new third-party models for image and video generation, including Black Forest Labs' FLUX.2 and Topaz Astra.

AMD AM4 CPU pricing spike as PC market forces alternative upgrades

16 December 2025 at 16:50

High DDR5 pricing is causing AMD/AM4 CPU price rises Consumer-grade DDR5 memory modules have seen priceΒ increases ofΒ 178-258%, forcing PC builders to consider alternative upgrade paths. Instead of moving to newer DDR5 platforms, some PC builders are upgrading to AMD’s DDR4-based AM4 platform. Why? The answer is simple: they can keep using the DDR4 memory they […]

The post AMD AM4 CPU pricing spike as PC market forces alternative upgrades appeared first on OC3D.

Microsoft exec’s β€œsuperhuman” praise of Musk highlights the outsized influence leaders wield in tech

Microsoft AI CEO Mustafa Suleyman revealed that he keeps close tabs and is in frequent contact with his peers in the AI space, including OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei, and Google DeepMind's AI Chief Demis Hassabis.

(PR) Call of Duty: Black Ops 7 Season 01 Enhanced with AMD FSR Redstone Features

16 December 2025 at 18:03
Call of Duty: Black Ops 7 Season 01 kicks off with a bang: remastered classics like Nuketown 2025 and Standoff, new maps like Fate, Utopia, and Odysseus, and a wild holiday variant, Sleighjacked. Multiplayer fans get a buffet of modesβ€”Prop Hunt, Sticks and Stones, One in the Chamber, Sharpshooter, and Gun Gameβ€”each tuned for maximum chaos and competition. Endgame received colossal new World Events. And for Zombies fans? The new Astra Malorum Map and Survival Map: Exit 115 are waiting, with new enemies (hello, O.S.C.A.R.), The Mule Kick Perk-a-Cola, and a new Wonder Weapon to discover.

AMD FSR: Ready to Run
Season 01 is the first to launch with AMD's FSR Redstone features. What does that mean for players? In short: smoother performance, sharper visuals, and a more immersive experienceβ€”no matter how intense the firefight gets. FSR is built on machine learning, delivering next-gen upscaling, frame generation, and ray regeneration. Whether you're pushing for 240 Hz on a Radeon RX 9000 series card or just want your game to look and feel incredible at any frame rate, Redstone has you covered

(PR) SEMI Reports Global Semiconductor Equipment Sales to Reach $156 Billion in 2027

16 December 2025 at 16:51
Global sales of total semiconductor manufacturing equipment by original equipment manufacturers (OEMs) are forecast to reach a record high of $133 billion in 2025, growing 13.7% year-on-year, SEMI announced today in its Year-End Total Semiconductor Equipment Forecast - OEM Perspective at SEMICON Japan 2025. Growth in semiconductor manufacturing equipment sales is expected to continue in the two following years of the forecast period, with projections of $145 billion in 2026 and $156 billion in 2027. This growth will be driven primarily by investments related to AI, particularly in leading-edge logic, memory, and the adoption of advanced packaging technologies.

"Global semiconductor equipment sales show robust momentum, with both the front-end and back-end segments projected to see three consecutive years of growth, culminating in total sales surpassing $150 billion for the first time in 2027," said Ajit Manocha, SEMI president and CEO. "Investments to support AI demand have been stronger than anticipated since our midyear forecast, leading us to boost the outlook for all segments."

(PR) D-Link Launches Nuclias Unity Network Management Platform

16 December 2025 at 16:42
D-Link Corporation, a global leader in networking and connectivity solutions, announced the launch of Nuclias Unity, a next-generation license-free cloud network management platform. Purpose-built for organizations ranging from SMBs to large, multi-site enterprises, Nuclias Unity delivers unified control, simplified operations, and enterprise-grade reliabilityβ€”without the traditional cost and complexity of licensed cloud platforms.

Backed by D-Link's decades of networking expertise, Nuclias Unity empowers IT teams to manage wired and wireless networks through a unified cloud management platform, accelerating cloud adoption while ensuring visibility, consistency, and operational efficiency across all business environments.

Colorful Intros Surprisingly Small iGame RTX 5070 Mini OC and RTX 5060 Ti Mini OC Series

16 December 2025 at 16:12
Colorful today introduced its iGame Mini RTX 50-series graphics cards. Not only do these cards meet NVIDIA's SFF-Ready specification for compact graphics cards, but exceed it by miles (rather, inches). The series is led by the iGame GeForce RTX 5070 Mini OC, followed by the iGame GeForce RTX 5060 Ti 16 GB Mini OC, and its 8 GB variant. Both the RTX 5070 and RTX 5060 Ti based cards appear to have a similar cooling solution, which contributes to their tight dimensions of 18 cm length, 12.3 cm height, and strictly 2-slot thickness.

The cooler of these cards features a dense aluminium fin-stack heatsink with the fins protruding out from the cooler shroud, much like NVIDIA's Founders Edition coolers. There is a single 95 mm axial airflow fan ventilating it. The backplate is solid metal, with a matte surface. On the RTX 5060 Ti 16 GB Mini OC, it cools half the memory chips. The iGame RTX 5070 Mini OC comes with factory OC of 2557 MHz compared to 2512 MHz reference; while both the RTX 5060 Ti cards come with 2632 MHz boost frequencies compared to 2572 MHz reference. The iGame RTX 5070 Mini OC draws power from a 16-pin 12V-2x6 connector rated for 300 W input, while the RTX 5060 Ti Mini OC cards come with single 8-pin PCIe power connectors. The company didn't reveal pricing.

Sapphire rep predicts DRAM prices will begin to stabilize in the next 6-8 months, but warns 'it may not be the prices we want' β€” GPU vendor says memory crisis is similar to tariff uncertainty

Amidst the economic uncertainty ushered in by this AI boom, some folks still have conviction and are offering hope to the community. Edward Crisler, the PR manager for GPU maker Sapphire, has just said that he believes DRAM prices will start to plateau in the next few months, so don't panic buy right now.

Qualcomm’s Ventana acquisition points to a long-term RISC-V strategy to complement its Arm lineup

16 December 2025 at 16:00
Qualcomm has agreed to acquire Ventana Micro Systems, a RISC-V CPU specialist whose engineers have spent several years pushing the open instruction set toward high-performance server and edge designs.

Progress Stalls: Sheryl Sandberg Warns AI Could Exacerbate Gender Inequality

16 December 2025 at 17:17

The post Progress Stalls: Sheryl Sandberg Warns AI Could Exacerbate Gender Inequality appeared first on StartupHub.ai.

The latest Lean In-McKinsey study reveals a stark truth: progress for women in the workplace is not just slowing, it’s stalling. Sheryl Sandberg, a pivotal figure in advocating for women’s leadership, returned to the public spotlight to deliver this sobering message, underscoring how emerging technologies like artificial intelligence threaten to further widen the gender gap. […]

The post Progress Stalls: Sheryl Sandberg Warns AI Could Exacerbate Gender Inequality appeared first on StartupHub.ai.

AI Fuels Megadeal Surge, Redefining M&A Landscape

16 December 2025 at 16:45

The post AI Fuels Megadeal Surge, Redefining M&A Landscape appeared first on StartupHub.ai.

Nearly a quarter of megadeals this year were AI-driven, a stark indicator of artificial intelligence’s transformative power in the M&A landscape. This trend, highlighted by Paul Griggs, U.S. Senior Partner at PwC, during his recent interview with Frank Holland on CNBC’s Worldwide Exchange, underscores a pivotal shift where strategic positioning and technological advancement are paramount. […]

The post AI Fuels Megadeal Surge, Redefining M&A Landscape appeared first on StartupHub.ai.

Unifying AI Operations: Flexible Orchestration Beyond Kubernetes

16 December 2025 at 16:16

The post Unifying AI Operations: Flexible Orchestration Beyond Kubernetes appeared first on StartupHub.ai.

The sheer velocity of AI innovation demands an infrastructure that can adapt, not just scale. At IBM’s TechXchange in Orlando, Solution Architect David Levy and Integration Engineer Raafat β€œRay” Abaid illuminated the critical need for a paradigm shift in how AI and machine learning workloads are managed, moving beyond the traditional automation paradigms. Their discussion […]

The post Unifying AI Operations: Flexible Orchestration Beyond Kubernetes appeared first on StartupHub.ai.

House proposes bill to advance data center buildout speed

16 December 2025 at 15:45

The post House proposes bill to advance data center buildout speed appeared first on StartupHub.ai.

The proposed legislation, dubbed β€œThe SPEED Act,” seeks to significantly reduce the time required for permitting and construction of data centers and associated power infrastructure. This is a crucial development, as the voracious appetite of AI for computational power necessitates a corresponding acceleration in the physical infrastructure that supports it. The bill proposes to limit […]

The post House proposes bill to advance data center buildout speed appeared first on StartupHub.ai.

Google fixes weeks-long Search Console Performance report delay

16 December 2025 at 16:37
Screenshot of Google Search Console

Google Search Console appears to have fixed the weeks-long delay in Performance reports. After several weeks of 50+ hour lag times, the reports now seem up to date as of the past few hours.

Now up-to-date. If you check the Search Performance report now, you should see a normal delay of about two to six hours. Over the past few weeks, that delay had stretched to more than 70 hours.

This is what I see:

The delays began a few weeks ago and took roughly three weeks to fully clear, including the backlog of data.

Page indexing report. Meanwhile, the Page Indexing report delay we reported weeks ago is still unresolved. The report is now almost a month behind, and Google has not fixed it yet. Google posted a notice at the top of the report that says:

  • β€œDue to internal issues, this report has not been updated to reflect recent data”

Why we care. If you rely on Search Console data for analytics and stakeholder or client reporting, this has been extremely frustrating. The Performance reports now appear to be updating normally, but the Page Indexing report remains heavily delayed and will continue to create reporting headaches.

Meanwhile, Google released a number of new features in the past few weeks, including:

How to boost ROAS like La Maison Simons by Channable

16 December 2025 at 16:00

Managing large catalogs in Google Performance Max can feel like handing the algorithm your wallet and hoping for the best.Β 

La Maison Simons faced that exact challenge: too many products and not enough control. Then they rebuilt their segmentation with Channable Insights and turned a β€œblack box” campaign into a revenue-generating machine.

Step 1: Stop segmenting by category

Simons originally split campaigns by product category. It sounded logical – until their best-selling sweater ate the budget and newer or overlooked products never had a chance to surface.

Static segmentation meant limited visibility and slow decisions.

Marketers stayed stuck making manual tweaks while Google kept auto-prioritizing only what was already working.

Step 2: Segment by performance

Enter Channable Insights. Product-level performance data (ROAS, clicks, visibility) now powers dynamic grouping:

Chart showing product segments: "Star Products" with a star, "Zombie Products" with a zombie face, "New Arrivals" with sparkles. Each has goals and strategies.

Products automatically move between these segments as performance shifts – no manual work needed. As Etienne Jacques, Digital Campaign Manager, Simons, put it:

β€œOne super popular item no longer takes all the money.”

Step 3: Shorten your analysis window

Instead of waiting 30 days for signals, Simons switched to a rolling 14-day window.

The result: faster reactions, sharper accuracy, and less wasted spend in a fast-moving catalog.

Step 4: Push the strategy across channels

Why stop at Google? The same segmentation logic was automatically applied on:

  • Meta
  • Pinterest
  • TikTok
  • Criteo

Cross-channel consistency creates compounding optimization.

Step 5: Watch the metrics climb

Without raising ad spend, Simons unlocked:

  • ROAS growth: from ~800% to ~1500%
  • CPC decrease: $0.37 to $0.30
  • CTR lift: 1.45% to 1.86%
  • 14% increase in average order value
  • 1300% ROAS for New Arrivals campaigns
  • Faster workflows and fewer manual tweaks

Even the β€œinvisibles” turned into surprise profit drivers once they finally got the spotlight.

Step 6: Treat automation as control, not chaos

Automation restored marketing control – it didn’t remove it.

Teams can finally learn from the data and influence which products grow, instead of letting PMax run everything on autopilot.

A table with a yellow header reading 'Quick Rules to Implement.' Two columns titled 'Principle' in pink and 'Why It Matters' in blue. Four empty rows beneath, with a colorful logo in the bottom left corner.

Your action plan

  • Classify products as Stars, Zombies, and New Arrivals.
  • Automate campaign reassignment based on real-time data.
  • Refresh product insights every 14 days.
  • Roll out segmentation logic to every paid channel.
  • Scale what wins – test what hasn’t yet.

Want Simons-style ROAS gains without extra ad spend? Start by testing the quality of your product data with a free feed and segmentation audit.

Why Data Security and Privacy Need to Start in Code

AI-assisted coding and AI app generation platforms have created an unprecedented surge in software development. Companies are now facing rapid growth in both the number of applications and the pace of change within those applications. Security and privacy teams are under significant pressure as the surface area they must cover is expanding quickly while their staffing levels remain largely

Fortinet FortiGate Under Active Attack Through SAML SSO Authentication Bypass

Threat actors have begun to exploit two newly disclosed security flaws in Fortinet FortiGate devices, less than a week after public disclosure. Cybersecurity company Arctic Wolf said it observed active intrusions involving malicious single sign-on (SSO) logins on FortiGate appliances on December 12, 2025. The attacks exploit two critical authentication bypasses (CVE-2025-59718 and CVE-2025-59719

Everbloom built an AI to turn chicken feathers into cashmere

16 December 2025 at 16:00
Everbloom has also developed a chemical process to transform waste fibers and feathers into upcycled materials that resemble everything from polyester to cashmere.

It’s bad – Here’s how much DDR5 pricing has increased

16 December 2025 at 15:01

How much have DDR5 memory prices increased? We all know that DDR5 memory pricing has shot up, but how bad is the situation? Has AI-driven datacenter demand ruined the DRAM market? Yes, but how much is it hitting our wallets? Today we have looked at today’s DRAM pricing and have compared it to 30 days […]

The post It’s bad – Here’s how much DDR5 pricing has increased appeared first on OC3D.

Thermaltake Intros TH360 V3 Ultra ARGB Sync Snow Edition CPU Cooler

16 December 2025 at 15:56
Thermaltake today introduced the TH360 V3 Ultra ARGB Sync Snow Edition, a premium all-in-one liquid CPU cooler that's positioned a notch below the company's Minecube Ultra series. The cooler features a cube-shaped pump-block, with a larger 4-inch, square true-color display floating on top. The cubical region stays recessed to ensure clearance around the CPU socket area. The 4-inch square display comes with a decent resolution of 720 x 720 pixels, and since it's square, the display-head can be rotated via software without any moving parts. The display connects to TT RGB Plus software, which puts out system monitoring info pulled from ACPI, and also lets you change backgrounds, and the cooler's lighting. Speaking of which, you get an ARGB diffuser framing the screen from the sides, and each of the three included 120 mm fans comes with RGB lighting of its own.

The pump turns at speeds ranging between 800 and 2,500 RPM. The cooler comes with 46 cm long tubing. Each of the three included 120 mm fans ventilating the 360 mm radiator turns at 500 to 2,500 RPM, pushing up to 85.29 CFM of airflow at 3.86 mm Hβ‚‚O static pressure, and up to 37.8 dBA of noise. All current desktop CPU socket types are supported, including AM5, AM4, LGA1851, LGA1700, and LGA115x/LGA1200. Thermaltake claims that the cooler can handle thermal loads of up to 365 W TDP, making it suitable for enthusiast-segment processors. The company didn't reveal pricing.

(PR) Enermax Launches REVOLUTION III S 1000 W Platinum ATX 3.1 PSU

16 December 2025 at 15:50
ENERMAX, an industry-leading force dedicated designing high-performance computer hardware and cooling solutions, proudly announces the launch of the REVOLUTION III S 1000 W, a premium fully modular power supply designed to meet the latest standards in performance, reliability, and aesthetics. Available in both black and white, the REVOLUTION III S 1000 W offers enthusiasts and professionals the flexibility to build powerful systems without compromise. The REVOLUTION III S comes with a 13-year warranty that highlights ENERMAX's confidence in the product's reliability and long-term durability.

The newest addition to the REVOLUTION line is 80 Plus, Cybenetics, and PPLP Platinum certified, ensuring outstanding efficiency and reduced power consumption. Fully compliant with the Intel ATX 3.1 standard, the REVOLUTION III S guarantees stable and reliable power delivery for not only the latest generation of high-performance graphics cards and CPUs, but also the upcoming generations too. Equipped with one 12V-2x6 interface, the REVOLUTION III S 1000 W can deliver up to 600 W of dedicated power to the latest GPU, making it future-ready for enthusiast builds.

DRAM Price Hikes Have Minimal Impact on PC OEMs, Notes Report

16 December 2025 at 15:05
The global DRAM shortage has driven up the prices of individual RAM kits, but PC OEMs have largely remained unaffected. Acer and ASUS indicated this week that rising memory costs are starting to influence notebook pricing, though retail price changes are still limited for now. Some brands, such as Dell, may begin raising prices on select high-end and business models, but neither Acer nor ASUS has officially changed their MSRPs in any way. Company executives warned that as new orders enter the market in the coming quarters, memory inflation will increasingly be reflected in end-product pricing. However, pricing remains stable for the time being, and there are no price changes.

Much of the near-term price stability comes from long-term supply agreements that protect OEMs and ODMs from paying spot market rates. Acer's CEO noted that memory historically accounted for about 8% to 10% of a PC's bill of materials, and that a 30% to 50% rise in memory costs so far has resulted in only an approximate 2% to 3% impact on the total BOM cost. Since many manufacturers secure memory through contracts that renew on quarterly or multi-year cycles, wider price fluctuations effects are likely to appear gradually, with more noticeable changes expected from the second quarter and into the third quarter of 2026 as contracts reset.

Samsung denies ending SATA SSD production amid NAND squeeze

16 December 2025 at 15:15

The controversy comes amid escalating demand for semiconductor memory driven by the growth of artificial intelligence infrastructure. Much of the industry's available NAND flash, once destined for consumer hardware such as SSDs, is now being redirected toward hyperscalers and AI labs. That shift has created one of the most constrained...

Read Entire Article

The data center cooling state of play (2025) β€” Liquid cooling is on the rise, thermal density demands skyrocket in AI data centers, and TSMC leads with direct-to-silicon solutions

The rise of AI and hyperscale computing is driving a global shift from air-based to liquid and embedded cooling as various companies are developing silicon-integrated systems capable of handling multi-kilowatt system-in-packages that can be commercialized by 2027.

Physical AI’s Off-Screen Revolution: Sanjit Biswas on Scaling Real-World Impact

16 December 2025 at 14:16

The post Physical AI’s Off-Screen Revolution: Sanjit Biswas on Scaling Real-World Impact appeared first on StartupHub.ai.

The next transformative wave of artificial intelligence is unfolding not in the digital ether, but in the tangible, messy reality of the physical world. This was the central thesis articulated by Sanjit Biswas, CEO of Samsara, in a recent discussion with Sequoia Capital’s Sonya Huang and Pat Grady. Biswas, a serial founder known for scaling […]

The post Physical AI’s Off-Screen Revolution: Sanjit Biswas on Scaling Real-World Impact appeared first on StartupHub.ai.

Samsung refutes consumer SSD phase-out rumours

16 December 2025 at 14:03

Samsung denies SATA SSD phase-out rumours, calling them false Samsung has officially denied reports that it plans to phase out its SATA SSDs and other consumer products. This follows recent rumours that Samsung planned to wind down its SATA SSD production to free up manufacturing capacity for data centre and AI customers. With Micron killing […]

The post Samsung refutes consumer SSD phase-out rumours appeared first on OC3D.

Intel Installs ASML TWINSCAN EXE:5200B High-NA EUV Machine for 14A Node

16 December 2025 at 14:24
Intel Foundry announced that it has managed to install the world's most advanced EUV machineβ€”ASML's TWINSCAN EXE:5200B High-NA EUV scannerβ€”in its facilities. The company is producing its 14A node using High-NA EUV lithography, marking the first industry transition from Low-NA. In collaboration with ASML, Intel has completed acceptance testing at Intel Foundry for its 14A node to enhance wafer output. The TWINSCAN EXE:5200B is ASML's second version of High-NA EUV scanners, following the TWINSCAN EXE:5000, which Intel initially used for its 14A trial runs. Intel previously reported processing over 30,000 wafers in a single quarter, achieving simplified manufacturing by reducing the steps needed for a specific layer from 40 to fewer than 10, resulting in significantly faster cycle times.

The new TWINSCAN EXE:5200B achieves an output of 175 wafers per hour in standard conditions, where Intel plans to tune it to over 200 wafers per hour. The machine also advances overlay precision, enabling accurate alignment of distinct lithography layers down to 0.7 nanometers. This achievement builds on Intel's High NA EUV experience, which began in 2023 with the installation of the industry's first commercial High NA tool at its Oregon research and development facility. Intel is currently shipping 14A PDK 0.5 to customers, who are reportedly very satisfied with the node's development. The company itself has praised the 14A node development as it has been achieving far better yield and performance parameters at this stage of development than the 18A node.

(PR) AAEON's BOXER-6648-ARS Delivers Intel Core Ultra Series 2 Power in Rugged Box PC Form

16 December 2025 at 13:58
Leading provider of industrial PC solutions AAEON (Stock Code: 6579), today introduced the BOXER-6648-ARS, its first fanless embedded Box PC featuring the new Intel Core Ultra Processors (Series 2) range (formerly Arrow Lake). Available in two SKUs offering either the Intel H810 (A1) or Intel Q870 Chipset (A2), the system offers a choice of Intel Core Ultra 9 Processor 285/285T, Intel Core Ultra 7 Processor 265/265T, or Intel Core Ultra 5 Processor 225/225T CPUs. As a result, the system can provide up to 24 cores of processing power alongside up to 36 TOPs of AI inferencing performance via new Intel platform's integrated CPU, GPU and NPU die architecture.

Primarily designed for more complex or AI-driven industrial automation applications, the BOXER-6648-ARS is equipped with six DB-9 ports for RS-232/422/485, an 8-bit DIO terminal block, and three LAN ports (two 2.5GbE, one 1GbE). The model based on the Intel H810 Chipset adds four USB 3.2 (5 Gbps) and four USB 2.0 ports to this selection, while the Intel Q870 Chipset model is more expansive, with six USB 3.2 (10 Gbps) ports and two USB 2.0. The other main differentiator between the two SKUs is Intel Active Management Technology support on the two 2.5GbE LAN ports, which is reserved for the Intel Q870 Chipset (A2) model only.

Intel Selects Pushkar Ranade as Interim Chief Technology Officer

16 December 2025 at 13:56
Intel announced significant changes in its senior leadership today, particularly in a key visionary role second only the CEO. Since Sachin Katti's departure a few weeks ago, the chief technology officer position has been vacant. Pushkar Ranade has now been appointed as the interim CTO. He will reportedly "help formulate the company's advanced technology strategy and to consolidate and develop critical emerging technologies, such as quantum computing, advanced interconnects, and novel materials within the new CTO Office." Pushkar Ranade has contributed to various Intel node developments and has been with the company for more than a decade, working on projects from the 65 nm node development to 7 nm SoCs. As interim CTO, he will assist CEO Lip-Bu Tan in executing his vision for a revitalized Intel.

Intel also made a few other leadership changes. Robin Colwell assumes the role of leading Intel's senior government affairs, where she will have worldwide engagement with policymakers, regulators, and industry stakeholders. Additionally, Annie Shea Weckesser joins as senior vice president and chief marketing and communications officer, leading the company's newly integrated global marketing and communications organization. She will unify corporate reputation, brand strategy, and market engagement. Intel's CEO Lip-Bu Tan pointed out that that all new executives offer the specialized knowledge and strategic vision essential for Intel's sustained success.

VT Chat – Minimal AI Chat with Deep Research Features


Introducing VT Chat, a privacy-first AI chat application that keeps all your conversations local while providing advanced research capabilities and access to 15+ AI models including Claude 4 Sonnet and Claude 4 Opus, O3, Gemini 2.5 Pro and DeepSeek R1.

Research features: Deep Research does multi-step research with source verification, Pro Search integrates real-time web search with grounding web search powered by Google Gemini.

There's also document processing for PDFs, a "thinking mode" to see complete AI reasoning, and structured extraction to turn documents into JSON. AI-powered semantic routing automatically activates tools based on your queries.

View startup

Salesforce AI Careers: A New Talent Pipeline Emerges

16 December 2025 at 13:18

The post Salesforce AI Careers: A New Talent Pipeline Emerges appeared first on StartupHub.ai.

Salesforce's global Workforce Development programs are actively shaping the landscape of AI careers, equipping over 120,000 learners with critical skills and certifications.

The post Salesforce AI Careers: A New Talent Pipeline Emerges appeared first on StartupHub.ai.

Amazon Upskills Workforce for Agentforce AI Era

16 December 2025 at 12:17

The post Amazon Upskills Workforce for Agentforce AI Era appeared first on StartupHub.ai.

Amazon is strategically investing in employee 'Agentforce AI' skills through Salesforce Trailhead, preparing its workforce for the agentic AI era.

The post Amazon Upskills Workforce for Agentforce AI Era appeared first on StartupHub.ai.

React2Shell Vulnerability Actively Exploited to Deploy Linux Backdoors

The security vulnerability known as React2Shell is being exploited by threat actors to deliver malware families like KSwapDoor and ZnDoor, according to findings from Palo Alto Networks Unit 42 and NTT Security. "KSwapDoor is a professionally engineered remote access tool designed with stealth in mind," Justin Moore, senior manager of threat intel research at Palo Alto Networks Unit 42, said in a

TechPowerUp x Chieftec Mega Giveaway: Entries Close Soon, Hurry!

16 December 2025 at 11:40
TechPowerUp partners with Chieftec to bring our readers from the European Union a chance to grab as many as six pieces of Chieftec gaming PC hardware, but you'd better hurry, entries close soon! Up for grabs are a Chieftec The Cube cube-shaped microATX case; a spacious Chieftec Hunter 3 EATX tower case; a Chieftec Apex Lumo ATX mid-tower case, a Chieftec Stealth 1000 W ATX 3.1 modular power supply; a Chieftec Iceberg 360 RGB AIO CLC, and a Chieftec Iceberg White 360 AIO CLC. Entries have been open over the past week, but close on December 19. It's really easy to join in, just fill up a little form to help us get back to you if you've won!

For more details, and to participate, visit this page.

(PR) GIGABYTE Announces Availability of AORUS Prime 5 Gaming Desktop

16 December 2025 at 11:33
GIGABYTE, the world's leading computer brand, announces that the AORUS PRIME 5 is now officially available. This high-performance desktop system is built on a new architecture with flagship-grade hardware, paired with GIGABYTE's signature cooling innovations and advanced fan technology. Designed to deliver exceptional speed and long-term reliability for gaming and multitasking, the AORUS PRIME 5 combines power, precision, and design with true plug-and-play simplicity.

The AORUS PRIME 5 not only features up to an AMD Ryzen 7 9800X3D processor and NVIDIA GeForce RTX 5080 graphics cards for multicore and next-generation AI performance, but is also built entirely with GIGABYTE products, including a 2 TB SSD and 32 GB of high-speed RGB memory for lightning-fast responsiveness and seamless multitasking. This configuration embodies GIGABYTE's DNA of proven stability, even extending that excellence into a fully integrated cooling solution.

Google to Shut Down Dark Web Monitoring Tool in February 2026

Google has announced that it's discontinuing its dark web report tool in February 2026, less than two years after it was launched as a way for users to monitor if their personal information is found on the dark web. To that end, scans for new dark web breaches will be stopped on January 15, 2026, and the feature will cease to exist effective February 16, 2026. "While the report offered general

Samsung Denies NAND Flash Exit as Sapphire PR Manager Calls for Calm

16 December 2025 at 09:20
The gaming hardware industry has been in a bit of a state of late as a result of increased demand for DRAM causing a supply shortage and massive price hikes. As a result of the aforementioned issues, Samsung has been rumored to be converting some of its HBM3E and NAND production capacity to DRAM in order to meet demand. The ensuing rumors claimed that Samsung was planning an exit from the NAND, and thus SATA SSD market, as a result of the shift in focus. Micron's recent exit from the consumer space, and the closure of its Crucial memory and SSD brand, lent credence to these rumors, however, these rumors have recently been addressed by a Samsung spokesperson directly in a response to Wccftech. The spokesperson simply said "The rumor regarding the phasing out of Samsung SATA or other SSDs is false," apparently declining to expand any further. However, this is just one indication that the DRAM crisis may not be as long-lived as some have claimed.

Around the same time, in an interview with Hardware Unboxed, Edward Crisler, the PR manager for Sapphire, cautioned PC gamers and potential buyers against panic buying, saying that, "the good news is, I don't think the real pain, that we're suffering now and for the next six months or so, is going to last much longer than that," although he goes on to say that the actual issue at hand is uncertainty of the situation. The implication is that the market will eventually begin to stabilize within the next six to eight months. The implication isn't necessarily that DRAM prices will return to normal, but rather that DRAM supply will eventually catch up to supply. It could also be the case that the massive AI and datacenter boom currently causing the shortages will slow in the coming months, which would also help to stabilize things somewhat. Crisler is careful to note that DRAM prices may still remain somewhat elevated after the market stabilizes, but he seems to be convinced that the sky-high prices we're seeing for consumer memory will fall to some degree and that the gaming industry will adapt to whatever the end result is of the market shake-up.

A Japanese startup built a speaker that's basically a sheet of fabric

16 December 2025 at 10:04

The technology originated at Japan's National Institute of Advanced Industrial Science and Technology (AIST) in 2018, where researchers demonstrated a thin, lightweight, bendable electronic textile. Sensia's new product represents the first commercial application of that research, adapting the concept into a consumer-ready format that requires no traditional speaker cones or enclosures.

Read Entire Article

Outage Owl – Stop finding out about vendor outages from angry customer tickets.


Outage Owl monitors 20+ vendor status pages in real time and alerts your team and customers when issues arise. Add a single script to show a website banner during outages and connect Slack to notify your team before tickets pile up. Create custom incidents and messages, tune alert rules, delays, and quiet hours, and keep everyone informed within seconds. Set up in under five minutes, and start free with one alert rule.

View startup

❌
❌