❌

Reading view

ClickPulse – Find high-earning affiliate programs with performance data


ClickPulse is an analytics and intelligence platform for affiliate marketers. It analyzes thousands of affiliate programs to show real conversion rates, EPC, order values, and effective commissions, helping you choose programs that actually make money. Use the niche insights and performance data to discover overlooked opportunities, compare alternatives, estimate earnings, and quickly replace underperforming partners. ClickPulse provides the data needed to make better decisions faster.

View startup

Manager Toolkit – Manage your team in one connected workspace


Manager Toolkit brings your team's work into one connected workspace with dashboards, actions, catchups, meetings, surveys, retrospectives, and targets so you stay prepared and follow through. It links every action to its source, surfaces key themes, and uses AI summaries to highlight what matters.

Start each day with clear priorities, run better 1:1s and retros, track development goals, and collect pulse feedback. Privacy first and secure by design, it helps people leaders save time and manage with clarity.

View startup

StockExpertAI – See real risks before you invest


Stock Expert AI is a financial intelligence platform with dual-engine AI forecasting that combines technical models with Facebook Prophet's seasonality detection for predictions on over 6,000 stocks. Its Moon AI chat assistant delivers live charts, prediction overlays, analyst consensus, and a proprietary Moon Score in one confidence rating.

Ask anything about a stock in the War Room and get rich, interactive widgets, not just text. From sector heatmaps to portfolio scoring, it brings hedge-fund-level insight to retail investors.

View startup

The Weekly Byte – Track competitors with an AI-synthesized email and dashboard


The Weekly Byte delivers a Monday email summarizing your competitors' funding, launches, hires, and partnerships, plus a web dashboard for market-wide trends and company timelines. It pulls signals from news, blogs, job boards, social channels, and product updates and synthesizes them into fast, scannable briefs. Replace the hour you spend Googling competitors every week. Pick companies to track, organize them into competitive sets, and get AI-generated landscape summaries, filters, and full profiles. Free for up to 3 companies, or upgrade to Core for $5/month for sets and AI summaries.

View startup

AMD RDNA 5 Secret Sauce Unveiled? Improved Dual-Issue could double performance

AMD LLVM patch unveils HUGE improvement for RDNA 5 graphics cards A new LLVM patch from AMD for its gfx1310 (RDNA 5) graphics architecture has unveiled some of the secret sauce behind AMD’s next-generation graphics cards. Through a feature called β€œDual Issue”, two operations can be completed during a single clock cycle, doubling the performance […]

The post AMD RDNA 5 Secret Sauce Unveiled? Improved Dual-Issue could double performance appeared first on OC3D.

Save over $100 on this feature-rich Asus AM5 motherboard with Wi-Fi 7, USB4 & DDR5-8000 support β€” TUF Gaming X870-Plus is on sale for just $170

Asus launched this motherboard at $310 in late 2024, and since then its MSRP has gone down to $280, but you can score it for 40% less than that on Amazon right now. For that price, nothing really comes close in terms of performance, features, and reliability.

PromptBrake – Run 60+ attack prompts to secure your LLM endpoints


PromptBrake tests LLM endpoints using 60+ attack prompts across 12 security checks. It identifies prompt injection, data leaks, tool misuse, and policy bypasses, then provides clear pass/warn/fail results with evidence so teams can resolve issues quickly. It is compatible with OpenAI, Claude, and Gemini APIs, and integrates into CI/CD processes before release.

View startup

This $155 ESP32-powered DIY drone can hit 67 mph


ESP-Blast centers on an ESP32 board serving as the flight controller, a role usually handled by dedicated flight-control hardware in high-performance drones. The ESP32, which integrates a dual-core processor along with Wi-Fi and Bluetooth, is typically used in connected devices and hobbyist electronics rather than in high-speed aerial platforms.

Read Entire Article

AMD's upcoming RDNA 5 GPUs might improve dual-issue execution & use shader units more efficiently β€” LLVM patch adds new FMA instruction to ease compiling

A new LLVM patch has added V_FMA_F32, a 3-operand fused multiply-add (FMA instruction and introduced the VOPD3 instruction format for RDNA 5. Both of these changes should make it easier for compilers to use dual issue execution, working around the strict pairing rules that would otherwise limit max FP32 throughput in certain workloads.

Zorpa – Automate your content pipeline with images, captions, and scheduling


Zorpa automates your social content pipeline by generating images, writing platform-optimized captions, and scheduling and posting across multiple accounts from one dashboard. It includes 50+ AI Studio templates with face consistency, an 8-platform caption writer, a content calendar, and a daily review process to ensure nothing posts without approval. Choose BYOK for low cost or managed credits with features like Autopilot, Competitor Spy, multi-account management, Ghost Mode, DMCA protection, CRM, campaigns, and a 24/7 DM chatbot.

View startup

Postmarkr – Upload a PDFβ€”we print, stamp, and mail via USPS First-Class


Postmarkr lets you send physical USPS mail from your computer. Upload a PDF, enter a recipient address validated against USPS, choose your mail class, and we handle printing, enveloping, and postage. Real-time USPS tracking is included for every piece.

Built for small businesses, law firms, and healthcare providers who need to send letters, certified mail, or priority mail without the post office. Pay per piece starting at $1.50 β€” no subscriptions or monthly fees.

View startup

YaadBooks – Run accounting, POS, and payroll offline with local tools


YaadBooks is an all-in-one business operating system built for Jamaica that keeps your business running even without internet. It unifies accounting, POS, payroll, inventory, bank reconciliation, and TAJ-ready reporting, with JMD as the default currency and multi-currency support. You can add industry modules for retail, restaurants, and salons to manage loyalty, tables, or appointments, all syncing automatically after outages. Higher tiers offer unlimited users and an employee PIN portal to keep teams moving offline, while an AI assistant delivers insights and local tax guidance.

View startup

This Week in Gaming (Week 12)

Welcome to week 12 and a jam cram packed week of rather big new releases. We kick off this week with a AAA release which is an open-world action-adventure that supposedly takes place in a desert, yet apparently not at all. This is followed by a very blocky looking RPG, several deadly sins, a cube of some kind, a walking simulator that apparently takes place on a beach and finally an ARPG that takes place after an asteroid destroyed the world. There are some additional games this week as well, as there were just too many releases this week.

Crimson Desert / This week's AAA release / Thursday 19 March
On the continent of Pywel, the rival factions of Pailune maintained a tense balance. At Kliff's side stood his steadfast Greymane comrades, but a devastating ambush in the deep of night by their sworn enemies the Black Bears leaves members of the Greymanes dead or scattered across the continent. Kliffβ€”having lost his family in all but nameβ€”is determined to reunite with the survivors and rebuild the faction. Steam link

Apple's MacBook Neo modded to a 1 TB SSD, breaking the firm's 512 GB barrier β€” base 256 GB model gets modded in expert NAND swap surgery

DirectorFeng, an expert technician in China has just performed what is likely the first hardware mod of its kind on the new MacBook Neo. He has birthed the world's only MacBook Neo with a terabyte of storage by physically changing the NAND chip on the logic board. Whether this mod is worth the price given the Neo's target audience, well, that's for you to decide.

ASRock launches new Frankensteined motherboard with one DDR4 slot and two DDR5 slots β€” Intel board signals the RAM apocalypse is truly nigh

A motherboard that can accept both DDR4 and DDR5 memory can be the difference between you being able to build a new PC or putting it off till the shortage is over. ASRock's new H610M Combo II is otherwise pretty barebones, not even featuring PCIe 4.0 storage, but it does have enough to get you by before things settle down.

Kapitol.ai – Track Congressional stock trades before the market


Kapitol.ai tracks, scores, and analyzes Congressional stock trades in real time, allowing you to follow high-conviction moves with clear context. It monitors all 535 members, ingests STOCK Act filings, and rates each trade by committee influence, position size, timing, and sector relevance. Subscribers receive instant email alerts, live P&L tracking, and insider score ratings, helping them turn public disclosures into actionable investment intelligence.

View startup

ContentForge – Repurpose long-form content into platform-native posts


ContentForge repurposes long-form content into publish-ready posts for LinkedIn, X, email, Instagram, Threads, Facebook, ad copy, and blog outlines. Paste a blog, transcript, or notes, pick platforms, and get native tone, length, and structure in under a minute. Built for B2B creators, founders, and marketing teams, it cuts drafting time by 80%, supports tone options, and offers plans from free to business with brand voice training, bulk processing, API access, and team accounts.

View startup

askt – Build a private journaling habit with secure prompts


askt is a private journaling app that gives you one thoughtful question each day and invites you to answer it. Write as much as you like, tag your mood, and return later to reflect on patterns and year-over-year growth. Entries save automatically and use end-to-end encryption, so only you can read them. It’s ad-free, not social, and collects no email. Start with a passkeyβ€”no passwords or credit card requiredβ€”and build a daily practice that sticks.

View startup

Revive – Recover failed payments and re-engage canceled subscribers


Revive gets back lost MRR with smart payment retries and win-back campaigns. Connect Stripe, Lemon Squeezy, Gumroad, Paddle, or Polar.sh via read-only OAuth and start recovering in minutes. Its engine analyzes decline codes to time retries, sends high-converting dunning emails with one-click card updates, and re-engages canceled users at 7, 14, and 30 days. You can track recovered revenue in a real-time dashboard and pay a flat $49 per month with no revenue share.

View startup

Epic Scout – AI-filtered freelance contracts for developers and designers


Epic Scout is a ruthless AI job filter for freelance developers and designers. Instead of losing 20% to Upwork or scrolling through irrelevant posts, our AI does the hunting. It sweeps remote job boards every two hours, eliminates junk listings, and emails you only the high-ticket, direct-client contracts that match your specific skills and hourly rate. You apply first, deal directly with the client, and keep 100% of your money.

View startup

Lead-Centric – Maximize your leads to retain clients longer


Lead-Centric is a white-label client portal platform designed for marketing agencies. It addresses the main reason agencies lose clients: lack of visibility. Each client receives their own branded portal where they can view every lead, every follow-up, and every deal in real time. No more 'what am I paying you for?' calls or spreadsheet reports. Clients feel informed, involved, and confident that their money is working. This allows you to keep more clients longer while spending less time proving your value.

View startup

Gigflow – Track gigs and automate invoices for UK freelance performers


Gigflow is a gig management app for UK freelance singers, musicians, and DJs. After every gig, most performers are stuck updating spreadsheets, guessing mileage, and chasing payments across multiple agencies. Gigflow handles all of this from your phone by auto-calculating commissions, logging HMRC-compliant mileage, generating invoices, and tracking what you're owed. It exports tax-year reports, so you're ready for self-assessment and Making Tax Digital without the admin chaos.

View startup

HireFlow – Ace every interview with AI-driven practice tailored to your role


HireFlow delivers realistic AI mock interviews tailored to your job description, CV, and role. It adapts questions in real time, runs voice or text sessions, and grades each answer with clear feedback and scores. Use detailed reports, coaching, and multilingual support to target weak areas and build confidence before the real interview. You can practice for any role, track your progress, and rely on GDPR-compliant privacy controls.

View startup

CertSim – Practice exam platform with AI-driven features


CertSim helps you pass IT certification exams with AI-driven practice tests and analytics. It delivers expert-validated questions for AWS, Azure, GCP, and more, with instant explanations, real-time translation, and timed simulations that mirror the real exam. You can track strengths and weaknesses on a dashboard, set a target date, and study on any device with continuously updated content.

View startup

Rogue Traders LIVE – Join a live trading floor and absorb pro decisions in real time


Rogue Traders broadcasts a fully operational trading floor in real time. Subscribers watch experienced traders run live books across assets, debate setups, and manage risk with no scripts or hindsight. You get live execution, cross-asset signal alignment, inflection analysis, desk notes, and portfolio-level perspective, delivered with the raw culture of prop floors. This is immersive access to how professionals think and act, not a course or copy-trading.

View startup

OpenClaw AI Agent Flaws Could Enable Prompt Injection and Data Exfiltration

China's National Computer Network Emergency Response Technical Team (CNCERT) has issued a warning about the security stemming from the use of OpenClaw (formerly Clawdbot and Moltbot), an open-source and self-hosted autonomous artificial intelligence (AI) agent. In a post shared on WeChat, CNCERT noted that the platform's "inherently weak default security configurations," coupled with its

ChatWithDr – 24/7 telehealth text-based consultations for $39.99


ChatWithDr is a digital health platform delivering 24/7 asynchronous telehealth consultations across all 50 U.S. states. Patients submit a secure online medical intake, and licensed physicians review each case to provide medical guidance, treatment recommendations, and prescriptions when appropriate. Designed to improve healthcare accessibility, ChatWithDr eliminates scheduling barriers by offering flat-rate pricing and text-based care without requiring video visits or insurance. The HIPAA-compliant platform focuses on common urgent care and primary care conditions while enabling scalable access to timely medical care nationwide.

View startup

VidShift – Convert videos and audio in your browser without uploading


VidShift is a free in-browser video and audio converter that runs entirely on your device using WebAssembly and WebCodecs. You can convert between MP4, WebM, MKV, MOV, and export audio as MP3, AAC, WAV, FLAC, or OGG with no uploads, watermarks, or file size limits. You can choose quality and resolution, trim, rotate, mute, change speed, compress files, or batch convert multiple videos. After the initial load, it works offline, keeping your media private while delivering fast conversions.

View startup

Shopper scores $1,000 in PC hardware for just $86 in a shocking pricing glitch β€” Newegg shrugs off massive loss and responds with a thumbs-up emoji

A Redditor only paid $86.98 for a Ryzen 5 7600X, a Gigabyte B850 Eagle motherboard, and a 32GB DDR5-6000 Corsair Vengeance RGB kit from Newegg. The CPU even included a free 240mm Cooler Master AIO worth $85 on its own. All these parts would typically cost $1,012, but this customer scored a combo deal for the ages.

Verve – Video-only dating app focusing on live connections


Verve is a video-only dating app launching exclusively in Washington, DC. It eliminates bios, photos, and endless texting. Users can browse short, authentic video profiles and, upon matching, choose to start a live or scheduled video call or go straight to an in-person date. Optional, slightly unhinged prompts, inspired by Cards Against Humanity, help enhance profile videos and live connections. This app is designed for people who confidently engage in person and are less concerned about superficial details. We are launching in DC first, focusing on a small, curated audience while we build a platform worth scaling.

View startup

GlassWorm Supply-Chain Attack Abuses 72 Open VSX Extensions to Target Developers

Cybersecurity researchers have flagged a new iteration of the GlassWorm campaign that they say represents a "significant escalation" in how it propagates through the Open VSX registry. "Instead of requiring every malicious listing to embed the loader directly, the threat actor is now abusing extensionPack and extensionDependencies to turn initially standalone-looking extensions into transitive

US gov't revokes controversial AI hardware export rule that would mandate investments from foreign companies β€” new export rules are still in the works, though

U.S. Department of Commerce withdraws draft export rule for AI accelerators that would give the U.S. government ultimate power over export controls and mandate investments of foreign firms in the U.S. AI sector.

qtrl – Modern QA platform that combines test management with AI automation


qtrl is a QA platform that unifies test management, automation, and AI-driven execution so teams can scale quality without losing oversight. It supports autonomous QA agents with browser execution, centralized test cases with full traceability, and progressive automation that you can review and approve at every step.

qtrl integrates with requirements, runs across multiple environments with secure secrets, and builds adaptive memory to perform smarter testing. It embeds governance, permissioned autonomy, and auditability for enterprise needs.

View startup

NUROSPICY – Gamified meal prep for neurodivergent minds


NUROSPICY is a meal planning web app that deals you dinner like a card game. With one button, you receive a hand of recipes, can reroll what you don't like, and get a combined shopping list grouped by aisle, a step-by-step cooking plan, and hands-free cooking mode. There's no sign-up required, no browsing, and no decisions to make. It's designed for neurodivergent people who love to eat but dislike cooking.

View startup

β€˜Everything seemed to go right’: Kurt Russell might β€˜never make his peace’ with Tombstone β€” but The Madison on Paramount+ is his Western redemption

New Paramount+ show The Madison marks Kurt Russell’s triumphant return to the Western genre β€” and he’s convinced it’s the β€˜best thing’ Taylor Sheridan has ever made.

Fiber internet provider says it can detect leaking water pipes using existing infrastructure, prevented loss of 2 million liters a day over three months β€” Lightsonic tech detects underground vibrations, machine learning isolates the source

U.K. fiber network provider Openreach used startup firm Lightsonic's technology to detect leaks in Affinity Water's service area, helping save over 2 million liters of drinking water daily.

Enthusiast rebuilds AA-battery-powered PC, sextuples run time to 30 minutes with 64 batteries β€” uses three voltage regulators in parallel to achieve stability, runs computer for over 30 minutes on 64 AA cells

YouTube creator ScuffedBits re-did their experiment and was able to eke out 30 minutes of game time (and a benchmarking session!) while running their desktop PC on 64 AA batteries.

Mylar Bag Packaging Template Generator – Creates accurate mylar bag templates with instant preview and export


Packaging Template Generator is a browser-based tool that lets brands and designers create production-ready templates for custom pouches and flexible packaging without using CAD software. Users enter dimensions and select features like gussets or zippers, and the template updates instantly with a live preview.

The generator produces accurate dielines that can be downloaded and used for printing, helping small brands and product makers design packaging faster without hiring a packaging engineer or designer. It is designed for food, supplements, cosmetics, and other products that use custom pouch packaging.

View startup

Athel AI – Streamline clinical operations with compliant AI automation


Athel AI builds precision automation for healthcare that streamlines hospital workflows, accelerates compliance tasks, and reduces manual bottlenecks. Its intelligent assistants manage intake, scheduling, communication, and document summarization while delivering real-time insights. The platform integrates securely, meets HIPAA, PHIPA, and PIPEDA standards, and stores data in Canada.

View startup

Keychron Launches $59.99 C0 HE 8K One-Handed Gaming Keyboard

Keychron's recent move to the HE 8K line of analog keyboards has seen it launch a number of new high-performance gaming keyboards with the new Ultra-Fast Lime switches and wired-only connectivity. Now, the keyboard maker has announced the C0 HE 8K as the latest installment in that line-up of gaming keyboards. The C0 HE 8K is a left-handed gaming-oriented wired Hall effect gaming keyboard with a built-in wrist rest and an aggressive angular design. The Keychron C0 HE 8K is launching at $59.99 on Keychron's online store and on Amazon. The Keychron C0 HE 8K is an affordable take on a one-handed gaming keyboard, so there is no gasket mount or wireless connectivity to be had, although it does have many of the same features of Keychron's higher-end keyboards.

The C0 HE 8K has 35 keys in total, making it technically a 35% keyboard, although it's only the left half of the keyboard, replete with the num row and F row. The keyboard's chassis itself is made of ABS plastic, while the keyboard features an aluminium plate and a tray-mount design. The keycaps are double-shot shine-through ABS in the OEM profile, which should feel familiar with many gamers coming from a standard gaming keyboard. Like all of Keychron's other HE keyboards, the C0 HE 8K offers customization of everything from the switch actuation distance to SOCD, DKS, Rapid Trigger, and Mod Tap through Keychron Launcher. The C0 HE 8K's Ultra-Fast Magnetic Lime switches feature a very lightweight 25 g actuation force, 3.35 mm of travel, and an all-POM construction. One of the stranger aspects of the C0 HE 8K is its "ergonomic angled space bar," that Keychron says is designed to align with the natural curvature of human fingers. The C0 HE 8K also features a built-in silicone palm rest for improved comfort. While the C0 HE 8K is largely marketed at gamers, Keychron also pitches it as a control surface for creative workflows.

Memflect – Turn your memories into audio reflections in your own voice


Memflect turns personal memories into 5-minute AI-voiced audio reflections. Share a memory, choose a soundscape, and record a short voice sample for cloning to receive a cinematic audio piece that sounds like a movie about your life.

Built for anyone going through somethingβ€”breakups, loss, starting over, or just wanting to process what matters. It also works as a gift: record a heartfelt message for someone you love, delivered in your own voice with cinematic music. One-time purchase of $3.99. No subscriptions or accounts are required to try.

View startup

Microsoft Will Add Gaming Copilot To Xbox Series Consoles in 2026

Microsoft Gaming Copilot has been out in the open and available as a beta for PC, mobile, and the ROG Xbox Ally since October 2025, but at a recent GDC conference, Sonali Yadav, Xbox Gaming AI partner group project manager, commented that Gaming Copilot would be landing on current-gen Xbox Series X|S consoles before the end of 2026. Gaming Copilot purports to help its users "get to the fun, faster," with gaming recommendations based on what you play, "get help without leaving your game," by providing players with advice based on their current place in the game, and "brainstorm strategies and get tips or insights with personalized coaching." The service operates similarly to Microsoft's original Copilot service in that it takes screenshots of your gameplay and analyzes them to provide the aforementioned recommendations.
I'm excited to announce that later this year, we will bring Gaming Copilot to the current-generation consoles, and we will continue to bring it to more services that players are playing.

APIPod – Access 100+ top AI models through a single, high-performance API


APIPod unifies access to over 100 AI models across text, image, video, and audio through a single OpenAI-compatible API. It handles routing, retries, and circuit-breaker protection, and provides real-time logs, token and cost analytics, and centralized governance so you can ship faster and control spend. With multi-channel routing, low global latency, and 99.99% uptime, APIPod supports everything from chatbots and content pipelines to enterprise AI gateways.

View startup

AMD Sees Agent Computers as the Next Step in AI PCs

AMD has published a guide for running OpenClaw locally on Windows through two distinct hardware paths it calls "RyzenClaw" and "RadeonClaw," both built around its own silicon and designed to keep AI agent workloads off the cloud entirely. The push is part of AMD's broader "Agent Computer" narrative, where it argues that not every AI workload belongs in a data center, people and businesses want control over their data, affordable always-on AI without usage limits, and the confidence that their models are running locally rather than on someone else's infrastructure. The setup runs through WSL2 with LM Studio handling local LLM inference via llama.cpp, and supports Memory.md through local embeddings, no cloud dependency required. AMD says the environment can be configured in under an hour, targeting early adopters and developers experimenting with personal AI agents.

The RyzenClaw path is built around a Ryzen AI Max+ system with 128 GB of unified memory, AMD specifically recommends reserving 96 GB as variable graphics memory for this use case. Running the Qwen 3.5 35B A3B model, that configuration delivers around 45 tokens per second, processes 10,000 input tokens in roughly 19.5 seconds, supports a 260K token context window, and can run up to six agents concurrently. AMD is positioning this as capable of "agent swarm" experimentation on consumer hardware. RadeonClaw takes a different approach, pairing OpenClaw with the Radeon AI PRO R9700, a workstation-class card with 32 GB of VRAM. That setup is considerably faster, around 120 tokens per second with the same model, and 10,000 input tokens processed in about 4.4 seconds. The tradeoff is a smaller 190K token context window and support for only two concurrent agents, compared to six on the Ryzen AI Max+ path.

ReditX – Create AI images and videos from text in seconds


ReditX is an AI-powered platform for instant image and video generation. It turns simple prompts into high-quality visuals with text-to-image, text-to-video, and image-to-video workflows, along with avatar video creation and advanced image enhancement. Creators and businesses can produce cinematic results quickly and manage output with coin-based plans across starter to enterprise tiers. Based in India and available worldwide, ReditX streamlines content production from idea to export.

View startup

Getaway Glow – Plan group trips with anonymous surveys and AI-crafted itineraries


Getaway Glow AI is a group travel planning tool made for women who are tired of being the unpaid travel agent for every trip. Before any bookings, the organizer sends an anonymous preference survey to the group, letting each person share their real budget, dietary needs, and activity preferences privately without fear of being singled out. The free tier handles surveys and the alignment report. For $39 per trip, the AI uses that data to generate custom itinerary options with destination recommendations and budget breakdowns based on what your group wants. The group then gets access to a trip planning dashboard for collaboration.

View startup

PlayStation 6 VS Xbox Next (Project Helix) – Leaked Specifications Compared

PS6 VS Next-Gen Xbox – Which console has the strongest specifications? The next generation of consoles is coming. Both Sony and Microsoft have started teasing what’s coming next. Specifications for Sony’s PlayStation 6 β€œOrion” silicon and Microsoft’s Xbox β€œProject Helix” (Magnus) chips have leaked. That means we can compare both systems early. What’s worth noting […]

The post PlayStation 6 VS Xbox Next (Project Helix) – Leaked Specifications Compared appeared first on OC3D.

Epomaker Officially Launches Glyph Typewriter-Style Mechanical Keyboard alongside Aula F75 Ultra, Luma100, and RT100 Pro

Epomaker recently showed off its typewriter-style Glyph mechanical keyboard at CES 2026, although there was no word on when the keyboard would launch, the exact specifications, or what the price would be. Now, the keyboard maker has officially released the Glyph keyboard in all its retro glory, alongside three other keyboards: the $119 Luma100, which we covered previously, the RT100 Pro, and the Epomaker x Aula F75 Ultra.

The Epomaker Glyph is a 75% mechanical keyboard with 83 keys, a programmable knob, a 2.79-inch full-color TFT screen, and a built-in tablet stand. The highlight of the Glyph is mostly its retro design, which is emphasized by the round, uniform-height keycaps that look a little like bottle caps. The keyboard itself is made of ABS plastic, while the keycaps are dye-sublimated PBT, and the plate is made of aluminium. The Glyph comes with either Epomaker Wisteria V2 switches, which are linear switches with a 45 gf operating force, POM+PTFE stems, PC top housings, and PA66 bottom housings. It also features 2.4 GHz and Bluetooth connectivity and an 8,000 mAh battery. The Glyph is available on Epomaker's website for $139.

New "Zombie ZIP" attack can evade most antivirus scanners


Known as Zombie ZIP, the method hides malware inside a deliberately malformed compressed archive. According to its creators, most antivirus engines currently fail to detect the threat, potentially giving attackers a new delivery mechanism. At the same time, some researchers argue the technique is less a vulnerability and more a...

Read Entire Article

Windows 11 is getting support for 1,000 Hz+ monitors soon as part of Insider builds β€” Microsoft has reportedly increased the refresh rate limit to 5,000 Hz

Microsoft says "monitors can now report refresh rates higher than 1000 Hz" in the patch notes for its latest Windows 11 Insider builds. These updates are part of the Release Preview channel, which means they're very close to final, public release. On the other hand, Nvidia has also pushed the first update for its G-Sync Pulsar displays.

Chloe Varnfield talks sneaky Google Ads settings and tanking performance

Chloe Varnfield, a digital marketing specialist at Atelier Studios with nearly eight years in PPC, joined me to share the mistakes that shaped her career β€” and the lessons every advertiser should take from them.

When Google sneaks settings past you

Chloe’s first story centers on Google’s account-level automated assets setting β€” a feature so well hidden that many advertisers don’t know it exists until a client sends a screenshot asking why their headline looks completely wrong. The setting, buried behind a three-dot menu, defaults to β€œon”, meaning Google can automatically generate and serve headlines advertisers never wrote or approved. The takeaway: always audit your account-level settings, and treat every Google update as a potential default you’ll need to turn off.

Why you should never make changes on a Friday

A client asked Chloe to narrow their campaign’s location targeting mid-call. She made the change quickly β€” and accidentally excluded the UK entirely while targeting only the desired regions. Campaigns stopped delivering. It took three days of head-scratching before she audited the full campaign and found the culprit. The lesson she now swears by: never make significant changes on a Friday, and when something stops working, go straight to a full audit rather than waiting for the algorithm to β€œfix itself.”

The time she listened to a Google rep β€” and tanked performance for two months

Chloe’s most costly story involves a campaign that was performing at its best in years. A Google rep recommended switching bid strategy from Maximise Conversions to Maximise Conversion Value. She made the switch β€” and performance collapsed. For small to medium-sized businesses that already struggle to hit the conversion volume thresholds needed for smart bidding to work effectively, changing bid strategy is a high-stakes decision that shouldn’t be made on the spot. It took two months to recover, with the pressure of a major seasonal sale looming. She fixed it β€” but the lesson stuck: don’t let enthusiasm or a rep’s insistence override your judgment. Sit on big decisions. Trust your gut.

The account mistakes that still happen in 2026

When auditing inherited accounts, Chloe consistently sees the same three problems: broken or absent conversion tracking (sometimes still pulling from Universal Analytics), broad match applied to brand campaigns β€” which makes it impossible to know whether results are genuinely driven by non-brand keywords β€” and accounts with zero negative keywords. These aren’t minor structural issues. They directly distort performance data and waste budget.

On honesty, client relationships, and not spiralling

Across all three of her own stories, Chloe’s client relationships survived because she communicated transparently β€” explaining what had gone wrong, what she was doing to fix it, and what the next step would be if that didn’t work. Her advice to anyone mid-crisis: breathe, be kind to yourself, stay calm, and remember that no one has died. The ability to fix problems under pressure is what builds expertise β€” and fixing something difficult often becomes your proudest professional moment.

The AI mistake too many marketers are making

On AI, Chloe is clear: using it to generate ad copy or proposals without reviewing or editing the output is lazy and obvious. AI should make you faster, not replace your judgment. Always put your own voice and review back into whatever it produces.

πŸ’Ύ

Chloe Varnfield shares sneaky Google settings, Friday mistakes, and Google rep advice that tanked her campaigns β€” and the hard-won lessons that came after.

Travel Hitched – Plan honeymoons and romantic trips with AI


Hitched uses AI to help couples plan unforgettable honeymoons, bachelor or bachelorette trips, mini-moons, anniversaries, and romantic getaways. Start by choosing your trip type, then get matched with a dedicated travel advisor who listens to your goals and creates a personalized plan. They support you from ideas to booking, offering stress-free planning and tailored recommendations for your celebration.

View startup

Meetings Recorder – Bot-free AI meetings recorder for Mac


MeetingRecorder is a native macOS app that records any meeting β€” Zoom, Teams, Meet, Slack β€” without adding bots to your calls. It captures both microphone and system audio locally on your Mac, then transcribes and summarizes with one click. It offers unlimited recordings and 20 free transcriptions per month. No accounts, no cloud storage, and no data sold.

View startup

Chinese Hackers Target Southeast Asian Militaries with AppleChris and MemFun Malware

A suspected China-based cyber espionage operation has targeted Southeast Asian military organizations as part of a state-sponsored campaign that dates back to at least 2020. Palo Alto Networks Unit 42 is tracking the threat activity under the moniker CL-STA-1087, where CL refers to cluster, and STA stands for state-backed motivation. "The activity demonstrated strategic operational patience and

Meta to Shut Down Instagram End-to-End Encrypted Chat Support Starting May 2026

Meta has announced plans to discontinue support for end-to-end encryption (E2EE) for chats on Instagram after May 8, 2026. "If you have chats that are impacted by this change, you will see instructions on how you can download any media or messages you may want to keep," the social media giant said in a help document. "If you're on an older version of Instagram, you may also need to update the

Nvidia upgrades its G-Sync Pulsar monitors with new 1.1.4 firmware

Nvidia improves its G-Sync Pulsar tech with optimised sub-90 FPS performance and a dedicated 60 Hz mode Nvidia has officially released firmware version 1.1.4 for its G-Sync Pulsar displays, supporting models from Acer, AOC, ASUS, and MSI. According to this firmware’s release notes, this update improves the performance of Nvidia’s monitors when operating at a […]

The post Nvidia upgrades its G-Sync Pulsar monitors with new 1.1.4 firmware appeared first on OC3D.

Windows 11 Insiders Get Support for >1,000 Hz Monitor Refresh Rate

Microsoft has released its latest Windows 11 builds, 26100.8106 and 26200.8106 (KB5079387), to Insiders in the Release Preview Channel. This update notably introduces support for monitor refresh rates exceeding 1,000 Hz, marking the first time Windows 11 officially supports four-digit refresh rates. Over the years, gaming monitors have rapidly progressed from double-digit to triple-digit refresh rates. However, reaching four-digit levels so quickly was unexpected, as the general gaming audience has been limited by current technology's ability to output such high frame rates. With modern GPUs capable of running hundreds of frames per second in FPS titles, it's logical for Windows to prepare for future advancements.

For instance, Philips and AOC have introduced the first gaming monitors capable of supporting 500 Hz at 1440p and 1,000 Hz at 720p. The Philips Evnia 27M2N5500XD and AOC AGON Pro AGP277QK monitors were showcased in China on December 6 and appear to be built around the same 27-inch panel. Both models are QHD 500 Hz displays that can switch to a lower-resolution 720p mode, doubling the refresh rate to 1,000 Hz. Generally, this resolution is required to achieve frame rates above 500 FPS, but modern GPUs like the NVIDIA GeForce RTX 5090 can easily reach these numbers in games like Counter-Strike 2. At 1080p, we observed about 726 FPS, indicating that 720p gaming can fully utilize the 1,000 Hz display.

CPU-Z v2.19 Update Brings Preliminary Intel "Wildcat Lake" Support

The popular CPU-Z utility for monitoring and hardware diagnostics has received an update in the latest version 2.19 release, bringing preliminary support for Intel's upcoming "Wildcat Lake" Core 300 series processors. This indicates that Intel's Core 300 series is on the horizon, and Intel might bring the same technology powering the Core Ultra 300 series "Panther Lake" processors to the embedded/edge sector. This lineup serves 12-25 W TDP applications with six processor cores, consisting of two "Cougar Cove" P-cores, zero "Darkmont" E-cores, and four low-power efficiency cores (2P + 0E + 4LPE). This CPU configuration is paired with two Xe3 GPU cores, which clearly shows that the product is aimed at lower-tier configurations. Additionally, the tool now includes support for AMD's Ryzen AI 7/PRO 450G/E, AI 5/PRO 440G/E & 435G/E, and AI 9 HX 470 processors. Finally, the tool now reads four-ranked CUDIMM memory in CQDIMM version.
Below is the complete changelog.

Tech titans team up to form optical interconnect alliance to solve the AI buildout's big data bottleneck β€” Nvidia, AMD, Broadcom & more set sights on building PHY to break through the limitations of copper

AMD, Nvidia, Microsoft, Broadcom, and Meta have formed an Optical Compute Interconnect Multi-Source Agreement (OCI MSA) to develop a standardized optical interconnect for AI data centers, with the aspiration to build a PHY capable of handling speeds of up to 3.2 Tb/s.

Nvidia claims 1 million times better path tracing performance is coming in future gaming GPUs β€” says current GPUs are already 10,000x faster than Pascal

At GDC 2026, Nvidia held a presentation somehow aimed at gamers and not data center clients. Still, it was sprinkled with AI pat-on-the-backs, with the company touting that its future gaming GPUs will offer 1,000,000 better path tracing performance. And the current-gen Blackwell family is apparently already 100,000x better due to dedicated Tensor and RT cores.

The ongoing Strait of Hormuz blockage will impact the semiconductor and AI industries with Aluminum, Helium and LNG shortages β€” and with no timeline for re-opening, supply chains face significant challenges

As global oil prices spike due to the ongoing pseudo-blockade of the Strait of Hormuz, other shortages are looming, and they could hit the chip and AI industries dramatically. Aluminum and helium for chip production have been impacted, but shortages of LNG could also cause problems for gas turbines powering many AI data centers.

Langulife – Social Language Practice Mobile App for A2+


Langulife is the language practice app for "post-beginners+" (A2–C1) who have the basics but struggle to bridge the gap to actual usage. We scrap the gamification trap of streaks and badges to focus on the absolute priority: your ability to express yourself. This is for the learner who is tired of playing games and ready to start communicating.

Instead of translating generic scripts, you engage with prompts that demand original thoughtβ€”the discussions you actually have in a boardroom or a bar. Express your perspective via voice and text, compare your answers with the community, and connect with language partners. Stop memorizing scripts. Start articulating what you think.

View startup

SerpApi asks court to throw out Reddit scraping complaint

Reddit vs SerpApi

SerpApi is asking a federal court to dismiss Reddit’s lawsuit over alleged scraping of Reddit content from Google Search, saying Reddit is trying to use copyright law to control user posts and public search results.

  • The motion follows Reddit’s amended complaint filed in February.
  • SerpApi says the filing still fails to show copyright ownership, circumvention of technical protections, or concrete harm.

SerpApi’s argument. SerpApi CEO Julien Khaleghy, in a blog post today, argued the lawsuit fails for several reasons:

  • Reddit doesn’t own most of the content at issue. Its user agreement states that users retain ownership.
  • Reddit holds only a non-exclusive license to user posts.
  • The snippets cited in the complaint (e.g., dates, addresses, short fragments) aren’t copyrightable.
  • SerpApi accessed Google Search pages, not Reddit itself.

DMCA. Khaleghy said Reddit claims SerpApi violated the Digital Millennium Copyright Act (DMCA) by circumventing technical protections. SerpApi disputes that claim, saying it retrieves the same search results visible to anyone who enters a query in Google. Khaleghy argued that:

  • SerpApi doesn’t break encryption or bypass authentication.
  • Accessing public webpages isn’t β€œcircumvention” under the DMCA.
  • Reddit is trying to enforce copyright protections it doesn’t own.
  • Reddit’s privacy policy states that public posts may appear in search results.

Catch up quick. Legal fights over search scraping and AI data have intensified in recent months:

Why we care. The case tests whether companies can extract information from Google’s search results without violating copyright or the DMCA. The outcome could affect SEO tools and AI training data.

What’s next. The court must decide whether Reddit’s amended complaint can proceed. If the judge dismisses the case with prejudice, Reddit’s claims against SerpApi in this lawsuit would end.

SerpApi’s blog post. Reddit’s Lawsuit is a Dangerous Attempt to Expand Platform Power

The $32B acquisition that one VC is calling the β€˜Deal of the Decade’

According to Index Ventures Partner Shardul Shah, cybersecurity startup Wiz sits β€œat the center of three tailwinds: AI, cloud, and security spend.” Those tailwinds powered what just became the largest venture-backed acquisition in history β€” Google’s $32 billion deal, finalized after a declined 2024 offer, antitrust review on both sides of the Atlantic, and an […]

Trump’s Iran war could make chip shortage worse – Analysts warn

The Iran war isn’t just affecting oil prices; it could harshly impact the semiconductor industry long-term When Israel and the United States of America attacked Iran, the first thing that analysts focused on was how this would impact the price of oil and natural gas. Iran has the power to effectively close the Strait of […]

The post Trump’s Iran war could make chip shortage worse – Analysts warn appeared first on OC3D.

(PR) Monster Hunter Stories 3: Twisted Reflection Launches Today

The journey begins today as Monster Hunter Stories 3: Twisted Reflection is now available on Nintendo Switch 2, PlayStation 5, Xbox Series X|S, and PC via Steam. Players can dive into the latest installment in the beloved RPG series, featuring the vibrant world of Monster Hunter shaped by the bonds between Riders and monsters.

A Tale of Two Kingdoms and Twisted Fates
Set two centuries after the Rift War, the kingdoms of Azuria and Vermeil stand divided. Twin Rathalos unexpectedly hatch from a single egg, igniting a story of forbidden lands and twisted destinies. A mysterious Crystal Encroachment emerges, threatening ecosystems and habitats across the land and causing conflicts between monsters. Along the journey, players will meet companions including vice-captain Gaul, monster expert Ogden, Ranger Squad co-founder Kora, the protagonist's best friend and confidante Simon, the brand-new member of the Rangers Thea, and Eleanor the Princess of Vermeil. As the protagonist, players take on the role of Prince/Princess of Azuria, and captain of the Ranger Squad to protect endangered eggs, restore ecosystems, and confront powerful monsters. Players can also customize their protagonist's appearance to fully step into the roles.

(PR) GMKtec Unveils 2nd Gen EVO-T2 Desktop AI Supercomputer with 180 TOPS

GMKtec unveiled its next-generation EVO-T2 Desktop AI Supercomputer Series during Intel's 3rd-Generation Core Ultra processor launch event in Shanghai. Powered by Intel Core Ultra X9 388H and X7 358H processors based on the Panther Lake architecture, the EVO-T2 delivers up to 180 TOPS of AI computing power, positioning it as one of the most powerful desktop AI mini PCs available.

The system integrates Intel's latest 18A (1.8nm-class) manufacturing technology, featuring RibbonFET gate-all-around transistors and PowerVia backside power delivery, marking a major milestone in Intel's semiconductor roadmap.

8BitDo Adds Commodore C64 Edition to Its Retro R8 Mouse Series

8BitDo is adding another variant to its Retro R8 mouse lineup, this time drawing inspiration from the Commodore 64. Like the previous R8 models, the C64 Edition keeps the same internals: a PAW 3395 optical sensor with six DPI levels between 50 and 26,000 DPI, and polling rates up to 8,000 Hz over wired and 4,000 Hz over 2.4G (Bluetooth is capped at 125 Hz). The symmetrical design supports both left and right-hand use, and four programmable side buttons can be configured through the 8BitDo Ultimate Software. The switches haven't been specified for this variant. The NES-inspired N Edition uses Kailh Sword GM X micro switches, but 8BitDo hasn't confirmed whether the C64 Edition carries the same.

The mouse ships with a charging dock that doubles as a wireless signal extender, which is a nice touch. Battery life is rated at 100 hours over Bluetooth or between 26 and 105 hours on 2.4G depending on polling rate, with a full charge taking around 2.5 hours. All three connection modes are supported: Bluetooth LE 5.3, 2.4G wireless, and wired USB. The 8BitDo Retro R8 C64 Edition mouse is priced at $49.99, in line with the existing NES-inspired N Edition.

PortfolioTrackr – Live portfolio tracking for stocks, crypto, and commodities


PortfolioTrackr is a clean, fast portfolio tracker built for serious investors. Track stocks across major global exchanges, crypto (BTC, ETH, and thousands of coins), and commodities like gold, silver, and oil, all with live prices in one dashboard. Set price targets (T1, T2) and stop losses on every position. Monitor your total P&L, per-position performance, and dividend income. Get price alerts delivered via email, Telegram, or WhatsApp the moment your targets are hit. No app to install. Works on any device. Free 3-day trial, then Starter at $9/month or Pro at $19/month with a Lifetime plan available.

View startup

Adverse – Own and monetize digital ad spaces you control and embed anywhere


Adverse lets you own and control digital ad spaces. Buy an Ad Space once, embed it anywhere, decide who can advertise, what appears, and what to charge while keeping 90% of the revenue. Approvals happen with a click, and smart contracts automate display and payouts. Scale layouts to fit your design, combine spaces, and choose ads that match your audience. Adverse runs privacy-first with no cookies or trackers, giving publishers creative freedom, transparency, and fair monetization.

View startup

Beyond keywords: Mastering AI-driven campaigns

Google Ads dashboard concept

The days of building campaigns around long lists of keywords are fading. Today, AI-powered Google campaigns and features like Performance Max (PMax) and AI Max are changing the rules.

These keywordless campaigns lean on automation, audience signals, and machine learning to find new opportunities, often faster and at greater scale than humans can.

At SMX Next, three PPC pros β€” Nikki Kuhlman, VP of search at Jumpfly; Brad Geddes, founder of Adalysis; and Christine Zirnheld, director of lead gen at Cypress North β€” explained where PMax and AI Max fit into your broader campaign strategy, where humans still make the difference, and how to strike the right balance between automation and control.

AI Max for Search: Best practices and what not to do

AI Max for Search is not a new campaign type. It’s a one-click opt-in setting within existing Search campaigns.

Without requiring you to switch to broad match, it expands your keywords β€” similar to broad match or Dynamic Search Ads β€” using your landing pages and other site assets. It then personalizes the ad copy and landing page the searcher sees.

The evolution from traditional setup

In the old setup, you might have used a keyword like β€œskincare for dry sensitive skin” that sent users to a moisturizer page with generic ad copy because you couldn’t capture every variation. With Google’s current matching, a specific ad group no longer guarantees that keyword will trigger that ad group.

AI Max for Search addresses this by generating ad copy based on the search query, making it more relevant and directing users to a landing page that better matches their needs.

Success with blog content

One area where AI Max for Search is seeing success beyond the norm is blog content. While DSA campaigns traditionally excluded blogs, AI Max for Search can now serve blogs as landing pagesβ€”and they’re converting. The key is that these blogs guide readers to specific products, not just general content.

The generated headlines are compelling and longer than what traditional RSAs allow, creating a more engaging user experience.

Best Practices for AI Max for Search

Do:

  • Use it on existing campaigns with history and data, not brand new campaigns
  • Test it as a 50/50 experiment instead of an outright change
  • Use it on brand campaigns with brand inclusion capabilities
  • Apply it to campaigns not hitting budget that could use more volume
  • Review landing pages and utilize URL exclusions (individual or rule-based)
  • Use landing page inclusions at the ad group level
  • Review search queries regularly and add negative search terms
  • Enable both text customization and final URL expansion for maximum value
  • Turn off AI Max at the ad group level when specific ad groups drive poor traffic

Don’t:

  • Use it on brand new campaigns without data
  • Change all campaigns at once without testing
  • Use it on brand campaigns without name recognition or brand inclusion ability
  • Apply it to budget-constrained campaigns
  • Turn off both URL expansion and text customization β€” if you’re not using both features, stick with broad match and smart bidding
  • Assume it works universally β€” test on individual campaigns

Your action plan

Week 1: Pick a search campaign to test (brand with brand inclusion, with budget capability, needing more volume). Review landing page URLs and add inclusions or exclusions.

Week 2: Review search queries and add negatives.

Week 3: Continue optimization and turn off AI Max at the ad group level as needed.

Experiment checklist:

  • Ensure enough volume for a 50/50 experiment
  • Give the experiment 6 weeks to 2 months
  • Set up a custom experiment if you need to enable brand inclusion or update settings
  • For one-click experiments, change confidence level to medium and turn off auto-apply

Match type performance: What the data shows

A comprehensive study analyzing over 16,000 campaigns revealed surprising insights about match type performance across different bidding strategies.

Match type basics

  • Exact match: Should match only when the search term has the same intent as your keyword. Misspellings and word order haven’t mattered for years β€” focus on user intent.
  • Phrase match: The search intent should match your keyword, but could have additional information around it, whether modifiers, phone numbers, or websites.
  • Broad match: Shows for anything related to the search intent. The key difference is that broad match uses additional signals that exact and phrase don’t, such as content on the landing page, other keywords in the ad group, and, most powerfully, previous search history for that user.

Performance by bidding strategy

Max Conversion strategies (Max Conversions, Max Conversion Value):

Most campaigns using max bid strategies have under 30 conversions per month, giving machines limited data to work with. The findings:

  • Exact match has the best click-through rates and conversion rates
  • Broad match had the worst conversion rates but the best return on ad spend
  • Broad match also had lower CPA than phrase match
  • Phrase match performed worst overall

Recommendation: Start with exact match, then skip phrase match entirely and layer in broad match if you have more budget to spend.

Target Bid Strategies (Target CPA, Target ROAS):

Most campaigns using these strategies have over 30 conversions per month, with many at 50 or 100+, giving machines substantially more data. The findings:

  • Exact match is again the best match type
  • Phrase match comes second
  • Broad match is third
  • Phrase match performs better with more data

Recommendation: Start with exact match, layer in phrase match with more budget, then add broad match if additional budget is available.

The phrase match puzzle

Why does phrase match perform poorly with limited data but better with more data?

Broad match uses additional signals, particularly previous search queries, to determine bids. When conversion data is limited (under 30 conversions monthly), broad match’s ability to leverage previous search history makes it much stronger than phrase match.

However, with sufficient data (50–100+ conversions), Google can properly match phrase match keywords using machine-learning pattern matching.

Brand vs. non-brand considerations

When you combine brand and non-brand data, exact match becomes even more powerful, delivering significantly higher click-through rates, higher conversion rates, lower CPAs, and much higher return on ad spend. That’s why segmenting keywords by brand and non-brand is crucial when determining your match type strategy.

Ecommerce exception

For ecommerce companies, broad match (and sometimes phrase match) can produce higher average order values than exact match. When someone searches for a specific product, and you carry that exact item, conversion rates are high, but they’re usually buying a single product with a lower checkout value.

When shoppers haven’t decided on a product, they tend to match broader keywords and build larger carts β€” resulting in lower conversion rates but higher order values.

Performance Max for lead generation

There’s a common misconception that Performance Max only works for ecommerce and is too difficult for lead generation. That couldn’t be further from the truth.

The critical success factor

The biggest mistake you can makeβ€”one you should avoid entirelyβ€”is optimizing campaigns for form submissions alone. If you treat every form submission as your campaign goal, you’ll end up with spammy submissions and frustrated sales teams.

The solution: integrate your Google Ads account with your CRM and import bottom-of-funnel leadsβ€”sales-qualified leads (SQLs), marketing-qualified leads (MQLs), opportunities, or even customers if the sales cycle is short.

When you tell Google Ads what you actually want and set it as your campaign goal, Performance Max can cast a wide net while still bringing in qualified prospects.

Available controls for regulated industries

Performance Max has significantly more controls now than at launch, making it viable for highly regulated industries:

  • Brand exclusions: Exclude all brand traffic from Performance Max campaigns
  • Campaign-level negative keywords: Exclude unwanted search terms directly
  • Search term reports: See what’s triggering your ads and exclude accordingly
  • Channel reporting: View spending and performance across different networks
  • Page feeds: Control where you send traffic on your site
  • Final URL expansion toggle: Turn it off completely if needed
  • Text enhancement controls: Optional feature that can be disabled entirely
  • Text guidelines: Specify words to avoid (e.g., β€œdiscount” or β€œdirectory”)

Device control: The secret weapon for B2B

One of the most underutilized levers for B2B and regulated industries is device control, introduced at the beginning of 2025. You can turn off any device from your Performance Max campaign.

A B2B SaaS example demonstrates the impact: Before device segmentation in January, the account had 224 SQLs from desktop at an acceptable CPA, but 33 from mobile at $319 CPA (above goal). After creating separate mobile campaigns with more aggressive target CPAs, they achieved 190 desktop SQLs and 37 mobile SQLs in a shorter month, with mobile CPA dropping to $204 and overall Performance Max CPA declining from $238 to $204.

Real Performance Max results for B2B SaaS

Despite lower conversion rates from Performance Max compared to search campaigns (due to broader reach), the results speak for themselves. In September 2025, one B2B SaaS account achieved:

  • Search Campaigns: 150 SQLs at $237 CPA
  • Performance Max: 204 SQLs at $220 CPA

Performance Max cast a wider net with cheaper CPCs, bringing in not just more leads but more sales-qualified leads at a lower cost.

How they did it:

  • Optimized for SQLs, not form submissions
  • Set lower target CPAs in Performance Max than search (to control spend while casting wider net)
  • Created separate campaigns for off-hours to control weekend spending
  • Turned off final URL expansion and text enhancements (client preference)
  • Implemented separate mobile and tablet campaigns with aggressive target CPAs

AI Max for Search in lead generation

AI Max for Search brings the power of Performance Max to the search network, where bottom-of-funnel intent is strongest. This is especially valuable for lead generation accounts that spend on other networks in Performance Max but don’t generate leads from them.

Early results: Higher ed financial services

A higher education financial client (loan products) showed promising early results:

Approved applications (primary KPI):

  • Standard Search: 86 approved applications at $660 CPA
  • AI Max: 70 approved applications at $579 CPA

AI Max brought in qualified leads cheaper despite the highly competitive keyword environment.

Down-funnel performance

Beyond the initial conversion action (soft credit check), AI Max showed superior performance throughout the funnel:

  • 42% of AI Max form submissions resulted in soft pulls vs. 36% for standard search
  • 9.9% of AI Max form submissions resulted in bookings vs. 5.58% for standard search

AI Max isn’t just bringing more qualified prospects at the topβ€”lead quality remains higher throughout the entire funnel.

How they did it:

  • Optimized for approved applications, not form submissions
  • Set lower target CPAs in AI Max than standard search
  • Used high-performing bottom-of-funnel keywords with broad match types
  • Kept final URL expansion and text enhancements disabled (still worked well without them)

Win with AI without losing control

PPC success requires embracing AI-driven campaigns while maintaining strategic human oversight. Whether you use AI Max for Search, Performance Max for lead generation, or adjust match types based on bidding methods and data volume, the key is understanding how these tools work and applying best practices aligned with your business goals.

The data is clear: exact match remains powerful across scenarios, but phrase and broad match perform differently depending on bidding strategy and data volume. For lead generation, the game changer is optimizing for true bottom-of-funnel conversions rather than form submissions, combined with strategic device controls and proper campaign segmentation.

The future of PPC depends on knowing when β€” and how β€” to apply automation and control for maximum impact.

πŸ’Ύ

PPC experts explain how AI Max and Performance Max reshape search, when automation wins, and where human strategy still drives better leads.

INTERPOL Dismantles 45,000 Malicious IPs, Arrests 94 in Global Cybercrime

INTERPOL on Friday announced the takedown of 45,000 malicious IP addresses and servers used in connection with phishing, malware, and ransomware campaigns, as part of the agency's ongoing efforts to dismantle criminal networks, disrupt emerging threats, and safeguard victims from scams. The effort is part of an international law enforcement operation that involved 72 countries and territories.

Storm-2561 Spreads Trojan VPN Clients via SEO Poisoning to Steal Credentials

Microsoft has disclosed details of a credential theft campaign that employs fake virtual private network (VPN) clients distributed through search engine optimization (SEO) poisoning techniques. "The campaign redirects users searching for legitimate enterprise software to malicious ZIP files on attacker-controlled websites to deploy digitally signed trojans that masquerade as trusted VPN clients

Automatic Super Resolution is coming to the ROG Xbox Ally X next month

Automatic Super Resolution to come to the ROG Xbox Ally X next month At GDC 2026, Microsoft has confirmed that Automatic Super Resolution (Auto SR) is coming to the ROG Xbox Ally X in April. This is a Windows-level upscaling solution designed to improve the image quality of games when rendering at a sub-native resolution. […]

The post Automatic Super Resolution is coming to the ROG Xbox Ally X next month appeared first on OC3D.

(PR) Fresh Look at Hello Sunshine Unveiled at Future Games Show Spring Showcase

Red Thread Games and Megabit Publishing today shared some fresh details about their highly anticipated survival mystery Hello Sunshine at the Future Games Show Spring Showcase. This latest look at Hello Sunshine focuses on the systems that drive the players' journey through a scorched, war-torn world and the relationship that they'll build with their giant robot companion along the way.

In Hello Sunshine, there is one simple rule: stay out of the sun. Step outside the robot's shade during the day and players will find that the punishing heat will eventually kill them; remain close, and they might last until nightfall.

Creality Sermoon P1 3D Scanner review: 3D scanning on the go

The Creality Sermoon P1 is a standalone 3D scanner that allows users to go through the entire scanning workflow without having to touch a computer. The 22 crossed laser line mode is surprisingly fast, and the single laser line mode was able to scan deep holes during testing. The NIR (near-infrared) mode is capable of scanning without requiring any markers, and the on-device editing allows users to make on-location edits without needing to stop. The Sermoon P1 handled everything we threw at it, but the color texture was the only disappointing part of the testing.

Why surface-level SEO tactics won’t build lasting AI search visibility

Google search monolith crumbling

A recent Harvard Business Review piece echoes the shift we’re sseeing in the SEO industry: at a macro level, LLMs and Google’s AI-powered SERP features, such as AI Overviews, aren’t just creating a zero-click environment, but also changing user journeys and behavior.

They’re collapsing what used to be multi-touch customer journeys into a single synthesized answer.

For a more visual and emphatic metaphor, the monolith of β€œSearch” is crumbling.

When that happens, brands lose many of the touchpoints they once owned, and your marketing strategy must change accordingly. HBR captures this moment well, arguing that marketing now has a new audience and that algorithms increasingly shape first impressions.

That said, while the article points in the right direction on the broader trend, its tactical advice is generic and falls back on shallow tactics.

Much of the guidance returns to familiar marketing playbook ideas that sound strategic and innovative but lack real operational depth. That gap matters for the longevity and sustainability of visibility.

The narrative may be easy for you to understand and repeat at the executive level, but it glosses over the deeper structural changes you must actually make to adapt to the new search ecosystem.

The problem with flock tactics

The HBR article centers on schema, authorship signals, and branded concepts. These recommendations risk becoming what I call β€œflock tactics.”

These ideas spread quickly because they’re easy to explain, but they offer little lasting competitive advantage once everyone adopts them.

SchemaΒ 

Schema has been one of the most debated topics in LLM and AI optimization. Microsoft Bing confirmed it uses schema for its LLMs, but the relationship between Google’s models and third-party LLMs isn’t as straightforward.

While it isn’t necessarily wrong to recommend schema as part of your overall search optimization activities (SEO and AI), positioning it as a table-stakes tactic ignores diminishing returns once competitors implement similar markup and it becomes standard.

Another gap is the role of external knowledge systems, such as Wikidata or authoritative publishers. Much of the information LLMs rely on comes from those sources rather than a single company’s website.

This is less linear to understand, explain, and demonstrate as a single line item on an activity tracker, but these are nuances you now have to deal with, whether you like it or not.

What’s also missing is any exploration β€” or even a nod β€” to how models ingest and prioritize structured data compared with the many unstructured signals they rely on.

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

E-E-A-T β€” shallow authorship signals

Attaching the names, credentials, and biographies of real experts follows familiar E-E-A-T logic and represents reasonable hygiene.

The problem is that the treatment remains superficial. It risks pushing you to focus on cosmetic signals such as bios, headshots, and credential lists without strengthening the underlying expertise pipeline.

There is a meaningful difference between placing an author bio on a page and cultivating a genuine expert entity whose work appears in conferences, third-party publications, standards committees, or academic collaborations.

Only the latter produces signals that models are more likely to recognize and trust.

Vanity concepts

The article also suggests creating branded frameworks or concepts β€” for example, something like β€œThe Acme Index” β€” to help models associate ideas with your company. In theory this sounds appealing, but in practice it’s extremely difficult to execute.

Unless those ideas spread into the trusted datasets LLMs tend to prioritize, they rarely gain traction.

You need those concepts and frameworks adopted and discussed by entities other than yourself, including academic journals, technical standards, widely used software ecosystems, and other prominent entities in your category.

What often results instead is a proliferation of branded labels that remain largely invisible to the models they were meant to influence.

The structural blind spots

Beyond these tactical issues, the analysis overlooks deeper structural challenges. It treats AI primarily as an external platform shift.

The implication is that you must simply adapt to it rather than actively shaping your own environment.

Internalizing AI infrastructure

HBR never seriously considers the possibility of building AI into your own infrastructure. You can deploy assistants, RAG systems, and domain-specific agents within your own products and customer experiences.

These systems operate in logged-in, transactional contexts where first-party data and controlled interfaces still matter enormously.

In those environments, traditional concerns such as site architecture, structured data, and product design remain deeply relevant, though they operate differently from public search optimization.

It’s not just SEO

The discussion also frames SEO primarily as a page-ranking problem tied to discovery.

That perspective misses the broader shift toward entity-level knowledge management (things, not strings).

Visibility within LLMs increasingly depends on how well you structure entities, taxonomies, and knowledge graphs, and on how those systems connect with external data sources.

Most LLMs don’t process data at the petabyte scale Google uses to understand entity relationships. There is a strong correlation that when something ranks well on Google, third-party LLMs often correlate and β€œtrust” Google’s guidance on which brands to show, for what, and when.

HBR’s phrase β€œengineering recall” points directly to this deeper data engineering work, yet the implications aren’t expanded.

LLM model heterogeneity

Another major omission is the diversity of AI systems themselves.

Different AI assistants and models rely on different training datasets, refresh cycles, retrieval mechanisms, and safety layers.

That heterogeneity means you can’t assume a single optimization strategy will work across all AI surfaces.

It also doesn’t explore the risk of broad-stroke approaches. If you try to increase visibility within AI models without accounting for safety filters, attribution errors, or hallucinations, you may gain visibility in ways that are inaccurate or reputationally damaging.

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

Surface-level tactics won’t build AI visibility

HBR’s article works well as a high-level explanation of how AI is changing marketing. It helps you understand that traditional SEO alone is no longer enough and that you must consider how AI systems see and describe your brand.

As a practical guide, however, the advice is thin. Most recommendations focus on surface-level tactics that many companies will quickly copy, reinforcing the echo chamber of flock tactics that are easy to sell and quantify, but risk narrowing your focus to short-term wins at the expense of longer-term strategy.

The real challenge is deeper. You need clear entity definitions, structured knowledge systems, reliable data in trusted sources AI models use, testing across how different models represent you, and AI-powered experiences within your own products.

β€œWinning” in the AI era will depend less on cosmetic SEO improvements and more on the harder structural work behind the scenes.

Only 15% of pages retrieved by ChatGPT appear in final answers: Report

AI search fan out

ChatGPT retrieves far more webpages than it cites. A new AirOps analysis found that 85% of discovered sources never appear in the final answer.

Why we care. If you want your content cited in AI-generated answers, discovery isn’t enough. Most retrieved pages never become visible to users.

Key finding. In AI answers, retrieval doesn’t equal citation. Your page can rank and be retrieved yet still lose the citation to a source that better matches the prompt or supporting context.

  • This shifts optimization toward earning selection inside the AI synthesis processβ€”not just appearing in search results, per the report.

By the numbers:

  • 82,108 citations appeared in final responses.
  • Only 15% of retrieved pages were cited.
  • 85% of pages surfaced during research never appeared in answers.

Citation rates also varied by query type:

  • 18.3% for product discovery queries
  • 16.9% for how-to queries
  • 11.3% for validation searches

Fan-out queries. ChatGPT often expands prompts with additional internal searches while generating an answer, creating what the report calls a β€œsecond citation surface.” Across the dataset:

  • 89.6% of prompts triggered two or more follow-up searches.
  • Fan-out searches expanded 15,000 prompts into 43,233 queries.
  • 32.9% of cited pages appeared only in fan-out resultsβ€”not the original prompt.
  • 95% of fan-out queries had zero traditional search volume.

Google ranking correlation. High Google rankings strongly correlated with citations:

  • 55.8% of cited pages ranked in Google’s top 20.
  • Pages ranking in Position 1 were cited 3.5 times more often than pages outside the top 20.

About the data. AirOps analyzed 548,534 pages retrieved across 15,000 prompts to examine how ChatGPT expands queries and selects citations.

The study. The Influence of Retrieval, Fan-out, and Google SERPs on ChatGPT Citations

Investigating a New Click-Fix Variant

Disclaimer: This report has been prepared by the Threat Research Center to enhance cybersecurity awareness and support the strengthening of defense capabilities. It is based on independent research and observations of the current threat landscape available at the time of publication. The content is intended for informational and preparedness purposes only. Read more blogs around threat

Windows 11 update enables 1000 Hz+ monitors

Microsoft is getting ready for the era of hyper-fast gaming displays With its newest Windows Insider build, Microsoft has enabled support for refresh rates above 1000 Hz. With builds 26100.8106 and 26200.8106, Windows 11 Insiders can now utilise screens that push past the kHz barrier. According to Blur Busters, who have pushed for this support, […]

The post Windows 11 update enables 1000 Hz+ monitors appeared first on OC3D.

(PR) Frore Systems Introduces LiquidJet Nexus - Defining the AI Thermal Stack for Next-Generation Β½ U AI Server Systems like NVIDIA Kyber

To overcome the thermal barriers limiting AI performance, Frore Systems today unveiled LiquidJet Nexus, a new integrated liquid cooling system designed for the AI era. As AI compute densities surge, heat is emerging as a primary constraint, elevating what experts are calling the "AI Thermal Stack" - the infrastructure responsible for extracting heat from advanced computing systems and rejecting it into the atmosphere. The AI Thermal Stack is increasingly determining the compute density, system efficiency, and overall performance of AI Data Centers.

LiquidJet Nexus is designed for liquid cooling Β½ U compute trays like NVIDIA Kyber, directly cooling GPUs, CPUs, DPUs, NICs, DC-DC power converters, and VRs. It integrates multiple multi-stage 3D short-loop jetchannel liquid cooling LiquidJet coldplates into a unified system. LiquidJet Nexus delivers 75% higher heat transfer efficiency, keeps components 8Β°C cooler, eliminating hoses, connectors, and manifolds, and reducing thermal stack weight by 65%. This unified system addresses the thermal demands of next-generation AI hardware.

Stop paying for traffic: The enterprise CMO’s guide to ROI-driven SEO

Vanity metrics vs revenue

The standard agency reporting call is broken. Budgets are under extreme scrutiny, yet you still invest in vendors that celebrate arbitrary traffic gains while your sales pipeline stays flat.

Optimizing for raw traffic volume is a legacy mindset that hides real commercial performance. The new mandate is to build an acquisition engine that influences buyers and protects your profit and loss (P&L) long before the transaction.

To survive as a marketing leader today, you must ruthlessly challenge your internal teams and external agencies. Stop accepting reports on operational output and demand hard financial accountability: pipeline contribution, customer lifetime value (LTV) to customer acquisition cost (CAC) ratios, and reduced paid media dependency.

The new path to purchase: Why traffic is bleeding your budget

Chasing top-of-funnel informational traffic is a trap. If the users clicking your links aren’t actively buying, you’re paying for vanity metrics, not business outcomes.

This happens because many buyers now use large language models (LLMs) to conduct deep research before they reach a search engine’s transactional layer. If you aren’t the cited authority during that AI-driven research phase, you’re invisible by the time buyers finalize their purchase decisions.

The 7.48% reality: The power of the educated buyer

The contrast in traffic quality is staggering when you look at the data. Across our enterprise client base, traditional organic search converts at 2.75%, while AI search converts at 7.48%.

LLMs function as the ultimate trust proxy for today’s consumers. When tools like Gemini, ChatGPT, or Perplexity synthesize dozens of reviews, whitepapers, and Reddit threads to recommend your enterprise software, users trust the LLM’s consensus more than a branded blog post.

AI engines arm consumers with comprehensive data, comparisons, and consensus. By the time a user clicks your AI citation, they’ve already made their decision based on your authority and are prepared to transact.

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

From found to cited: Architecting the default recommendation

Want to capture this 7.48% conversion rate? Your entire approach to digital asset creation must evolve. The strategy no longer centers on ranking among a list of links, but on being cited as the definitive option.

To win the AI consensus, you must translate your marketing strategy into structured capital management.

  • The old way: Publishing a 2,000-word blog post on top supply chain trends that generates 5,000 monthly visitors who bounce after reading and add zero value to your pipeline.
  • The new way: Build a generative engine optimization (GEO) hubβ€”a dedicated supply chain cost calculator page with proprietary data tables, expert author schema tagging your lead engineers, and strict answer-first formatting.

LLMs require consensus and verifiable facts to generate confident answers. By structuring your digital assets with proprietary data and verifiable entities, you become the default recommendation.

This approach may yield only 500 highly qualified visitors, but it gives LLMs what they need to cite you in vendor comparison prompts and captures buyers at the exact moment of commercial evaluation.

Strategic ROI: Using citation authority to reduce ad spend

It’s time to stop viewing SEO as a siloed traffic generator. You must treat organic citation authority as a strategic financial lever to reduce overall CAC.

Align your organic assets with your highest-CAC paid campaigns. When organic search owns the AI Overview, your paid team can confidently pull back defensive ad spend.

Here’s how to leverage paid and AI search:

  • IF your brand becomes the default AI recommendation for a high-cost commercial category, THEN your paid team must aggressively reduce defensive brand bidding to slash overall cost per acquisition (CPA).
  • IF paid search identifies a highly profitable long-tail query, THEN SEO must prioritize building a structured asset to organically capture that exact demand in the future.
  • IF an LLM cites your competitor as the superior enterprise solution, THEN your paid team must immediately deploy targeted, bottom-of-funnel conquesting ads to intercept that user before the transaction, while the organic team rapidly engineers a proprietary data asset to win back the consensus.

The monthly cannibalization review: Your immediate action item

If your Head of Search and Head of Paid Media aren’t in the same room once a month mapping organic citations against paid brand bidding, you’re burning capital.

Align your teams and channels. Routinely audit where you’re paying for clicks on terms where you already own the AI citation and the top organic spot.

Treat this cannibalization review as a strict financial audit. Identify wasted defensive ad spend and immediately reallocate those dollars toward net-new market expansion.

The enterprise scorecard: 3 questions to ask your agency tomorrow

To regain control of your P&L, you must challenge your vendors to step up. Ask your agency these three questions tomorrow morning to see if they’re true business partners or order-takers.

1. What’s our citation share of voice for our highest-margin categories?

Challenge your team to map their organic efforts directly to the AI research phase of your most profitable products.

The answer you should hear: β€œWe’ve mapped your 50 highest-margin queries. By securing the primary AI citation for these, we’ve generated $1.2 million in pipeline this quarter at a 3:1 LTV:CAC ratio.”

2. How is our citation strategy directly reducing our paid media CAC?

Require teams to prove how their organic authority captures demand that would otherwise require paid ad spend.

The answer you should hear: β€œBy capturing the definitive AI citation for [category], we paused paid bidding on those terms. This reduced our blended CAC by 18% and saved $45,000 in defensive ad spend β€” which we’ve immediately reallocated to net-new market expansion.”

3. Are our digital assets structured for LLM extraction?

Push your teams to explain their strategy for AI-driven search models. It’s no longer enough to publish standard web pages.

The answer you should hear: β€œWe’ve restructured your core commercial pages away from standard marketing copy, deploying answer-first’ frameworks, proprietary data tables, and expert author entities to ensure LLMs confidently extract and recommend your brand. This structural shift has increased our inclusion in commercial AI Overviews by 40% this quarter, directly feeding our bottom-of-funnel pipeline.”

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

Demand commercial outcomes, not operational output

In a tough economy, SEO is a measurable business unit that must defend its budget with revenue data. Don’t accept operational output as proof of commercial success.

Audit your reporting frameworks immediately. Stop accepting vanity metrics as evidence of success. Demand pipeline impact, LTV:CAC ratios, and a resilient acquisition engine.

Any agency or internal team unwilling to tie its work directly to your P&L will become obsolete. Your job as an enterprise leader is to ensure your brand is cited as the authority long before the transaction begins.

Google Search Ads in 2026 require a different kind of audit

Google Search Ads value redistribution

Brandon Ervin, Director of Product Management for Google Search Ads, recently discussed campaign consolidation, AI Max, and what advertiser control looks like in 2026 on Google’s Ads Decoded podcast. The conversation was serious and informed, and reflected a product team that understands advertiser concerns and is actively working to address them.

But the podcast is also incomplete. The gap between what Google said and what advertisers actually experience from their sales organization is large enough to warrant a direct response.

Ervin’s team is doing genuinely good work, but the platform’s structural incentives haven’t changed. Google’s evolving product is creating problems faster than it can solve them. Performance is now measured on economic standards, shaping how a search ads audit is performed.

Recent improvements to Google Search Ads

Recentish improvements are genuine:

  • Brand exclusions in Performance Max and Demand Gen.
  • Site visitor and customer exclusions from PMax campaigns.
  • Network-level reporting within bundled campaigns.
  • Improved search term visibility.
  • Brand and geo controls inside AI Max at the ad group level.
  • Semantic modeling that doesn’t anchor on campaign or ad group IDs, reducing learning period risk during consolidation.

These are meaningful. They are also solutions to issues introduced by bundling, opacity, and aggressive automation rollout.

These products have been mercilessly shopped to advertisers since 2021, and the controls that make it usable arrived years after the sales push began.

The ability to separate brand from non-brand traffic inside PMax/AI Max should not be framed as innovation. It restores a fundamental distinction that previously existed by default. The ability to see network performance inside a bundled campaign is not an expansion of control. It restores visibility that was removed.

An audit must ask whether new tools are genuinely expanding control or merely reintroducing baseline transparency.

Table stakes: What everyone agrees on

Before the real audit begins, the fundamentals. These are uncontroversial and should already be in place:

  • Run full ad extensions (sitelinks, callouts, structured snippets, image, call).
  • Use automated bidding with intentional target-setting and conversion action selection (I recognize there are still holdouts here but seems crazy to me).
  • Maintain negative keyword lists.
  • Write ads relevant to the queries they serve.
  • Audit automatically created assets for accuracy and brand safety.
  • Cut Search Partners and Display expansion from Search campaigns.
  • Separate brand and generic campaigns using brand controls.
  • Exclude site visitors and past customers from prospecting campaigns where appropriate.
  • Import offline conversion data (MQLs, SQLs, revenue, CLV, repeat rate,) to feed the algorithm downstream signals.
  • Weight conversion values by actual downstream conversion rates.
  • Account for mobile vs. desktop performance gaps.

Those are table stakes. The real audit begins after that.

What a 2026 search audit must focus on

With the prevalence of AI, advertisers need to focus on reconstructing economic visibility in systems designed around aggregation and automation.Β 

Signal architecture

In the podcast, Ervin says β€œcontrol still exists, it just looks different.” Ad controls β€” where, when, and to whom ads appear β€” are still important and changing, some think, for the worse.

The old ad controls β€” exact match, manual bids, network selection, and device modifiers β€” gave advertisers direct influence over where ads appeared and what they paid.Β 

However, the new controls are indirect. Control now lives in data quality, density, and selectivity. They influence the algorithm, but the algorithm makes the final call.

An audit should focus on three questions:

  • Quality: Are you importing revenue, pipeline stage, or qualified lead status, or only surface conversions?
  • Density: Is there enough high-quality data for the model to learn from, or is it sparse and noisy?
  • Selectivity: Are you intentionally limiting what Google can see, or are you passing everything indiscriminately?
Low prediction, high density

With these new tactics, you only pass net-new customers or high-value customers. The majority of the time, it is better to just pass the densest and most predictive conversion set.Β Β 

Incrementality

Google optimizes toward reported conversions, not incremental conversions. Brand search often captures existing demand. Retargeting often captures users already in motion. Pmax/AI Max frequently blends these signals.

Ervin was asked: Are AI-driven campaigns over-indexing on warm brand traffic to inflate blended ROAS (return on ad spend)?

He doesn’t dispute the problem, but points to partial solutions, including using brand controls, better theme your account, and looking at multi-campaign A/B testing.Β 

If incrementality is not measured, automation amplifies non-incremental signals.

Marginal returns

Google uses a blended cost-per-action (CPA). For example, the first $50K of spend might return a $30 CPA, while the next $50K might return $120.Β 

With automation, money is spent until the blended metric falls within tolerance, meaning the last dollar is not spent efficiently. The vast majority of advertisers are bidding far beyond what they should be and have no idea it is happening.

An audit must:

  • Plot spend against incremental conversions.
  • Estimate marginal CPA at each spend tier.
  • Identify diminishing return curves.
  • Compare marginal CPA to lifetime value.

A lower target makes the algorithm more selective, competing in fewer high-value auctions. Google doesn’t suggest this because that would mean less spend and lower bids are less effective in general.Β Β 

Query resolution and ability to lower targets

On the podcast, Ervin acknowledges that some AI Max matches can β€œlook a little wonky” and says his team is working on exposing the model’s reasoning.Β 

Query mapping has gotten meaningfully worse over the past several years: queries landing in the wrong ad groups, matching to keywords with different intent, and broad match pulling in traffic unrelated to the keyword.

AI Max has accelerated this β€” there’s been an increase in the volume of irrelevant queries flowing through AI Max campaigns, with no connection to the advertiser’s business or keywords in the account.Β 

Meanwhile, Google’s recommendations consistently push toward broad matching and large themed ad groups.Β Β 

The issue is not whether broad match works, but whether high-value intent is being diluted in larger, broader ad groups. Fewer ad groups means that we cannot effectively or meaningfully lower targets without a massive structural negative schema, so performance differences have to be large enough to validate the new structure.Β 

An audit should:

  • Extract full search term reports.
  • Classify queries by intent tier.
  • Compare CPA and lifetime value by query type.
  • Quantify irrelevant or weakly related matches.
  • Measure performance drift across match types.

Network economics

Performance Max and Demand Gen bundle multiple networks into single campaigns, but offer limited visibility into which networks drive results. This makes it hard to cut the underperforming ones. The slow rollout of network-level controls systematically benefits Google’s less competitive inventory.

An audit must:

  • Break out performance by network.
  • Compare CPA and lifetime value by placement.
  • Identify cross-subsidization.
  • Determine whether weaker networks are relying on surplus from strong search inventory.

Value redistribution

Combining these elements in your audit will help you succeed in this new world of ad search:Β 

  • Non-incremental traffic inflates conversion counts, making performance look better than it is.
  • Looser match types expand where ads appear, diluting intent precision and forcing fewer ad groups/spend and blanket-level targets/bids.
  • No clean marginal return visibility means it is much more difficult to find the point of negative return
  • Network bundling hides which channels actually perform.

The cumulative effect is that the surplus value generated by your best inventory and high-intent, high-converting search queries gets redistributed across Google’s weaker inventory (i.e., Display, YouTube, Discover, Gmail, crazy tail queries).Β Β 

This is how to get a dwindling supply of valuable search queries to inflate the cost-per-clicks (CPCs) of low-quality inventory.Β 

The Ads Decoded episode: Is your campaign structure holding you back in the era of AI?

Β 

πŸ’Ύ

Google Search Ad audits must now rebuild the visibility automation removed. Here's how. (A response to Google’s Ads Decoded podcast.)

AMD Ryzen X3D CPU prices drop in the UK

AMD’s Ryzen 7 X3D CPUs have received price decreases in the UK Less than two months have passed since the release of AMD’s Ryzen 7 9850X3D (review here), and the CPU is already available at a discounted price. In fact, all of AMD’s currently available Ryzen 7 X3D CPUs are now available at lower prices […]

The post AMD Ryzen X3D CPU prices drop in the UK appeared first on OC3D.

The FBI is looking for victimized Steam users who downloaded games with hidden malware β€” Investigation underway into multiple infected titles from 2024 to 2026

Several games on Steam that were secretly carrying malware now seem to be under active investigation by the FBI. The department is looking for victim information tied to these games; anyone who installed and played an infected game and was harmed is being urged to step forward and share more info to help with the investigation.

The Social Enterprise Pound – Hire African professionals for SME projects with built-in social impact


TSEP (The Social Enterprise Pound) connects vetted African professionals with SMEs worldwide, enabling businesses to access reliable remote capacity while supporting ethical digital employment across Africa.

Through a structured social enterprise model, SMEs can outsource projects to trusted talent while benefiting from built-in compliance, transparent workflows, and managed delivery. TSEP helps businesses scale efficiently while creating sustainable income opportunities for skilled professionals across the African continent.

View startup

Shombo – Intake to settlement in one platform built for solo PI attorneys


Shombo is an all-in-one case management platform built for solo personal injury attorneys. It automates the entire PI workflow from client intake and contingency fee agreements with built-in e-signatures to AI-powered demand letters, medical record summaries, and settlement statements with lien tracking. Stop paying for five different tools; Shombo replaces your fragmented stack with a single platform that tracks every case from new lead through disbursement. AI generates demand letters from your case data, extracts injuries and treatments from medical records, and drafts daily briefings, so nothing slips. Built-in reminders chase clients so you don't have to.

View startup

Google Fixes Two Chrome Zero-Days Exploited in the Wild Affecting Skia and V8

Google on Thursday released security updates for its Chrome web browser to address two high-severity vulnerabilities that it said have been exploited in the wild. The list of vulnerabilities is as follows - CVE-2026-3909 (CVSS score: 8.8) - An out-of-bounds write vulnerability in the Skia 2D graphics library that allows a remote attacker to perform out-of-bounds memory access via a crafted HTML

Nine CrackArmor Flaws in Linux AppArmor Enable Root Escalation, Bypass Container Isolation

Cybersecurity researchers have disclosed multiple security vulnerabilities within the Linux kernel's AppArmor module that could be exploited by unprivileged users to circumvent kernel protections, escalate to root, and undermine container isolation guarantees. The nine confused deputy vulnerabilities have been collectively codenamed CrackArmor by the Qualys Threat Research Unit (TRU). The

(PR) Turtle Beach Corporation Announces Fourth Quarter and Full Year 2025 Results

Turtle Beach Corporation (Nasdaq: TBCH), a leading gaming accessories brand, today reported financial results for the fourth quarter and full year ended December 31, 2025 and provided full year 2026 guidance for net revenue and Adjusted EBITDA.

"Our team operated with strong discipline in 2025, executing our strategic initiatives to sharpen our competitive edge," said Cris Keirn, Chief Executive Officer of Turtle Beach Corporation. "Although our fourth quarter and full-year results were below guidance due primarily to softness in North American gaming accessories markets and a lighter holiday sell-through environment, we took steps to manage costs, protect our brand position, and prepare the Company for renewed growth."

(PR) Supermassive's Directive 8020 Gets New Live-Action Couch Co-Op Trailer

Supermassive Games unveiled a brand-new live-action couch co-op trailer for its upcoming sci-fi survival horror adventure, Directive 8020, during today's Future Games Show. The trailer showed a group of friends playing couch co-op Movie Night Mode, intercut with in-game moments of split-second quick-time actions, narrative decisions, and alien mimics.

Survive Together, Don't Play Alone
Complementing its gripping single-player experience, Directive 8020 allows up to five players to play through the story together in local couch co-op via the returning Movie Night Mode. Each player takes control of a crew member aboard the Cassiopeia, passing a single controller between each other as the story unfolds. It's up to the player and their friends on who lives and who dies; every choice matters, every mistake can be fatal, so players must work together to survive alien infiltrators, deadly dilemmas, and catastrophic system failures! Don't Play Alone, and trust no one - not even your friends! Directive 8020 will also support five-player online multiplayer in a free post-launch update, mirroring the couch co-op experience and bringing it online, with more information coming soon.

(PR) Global Top Five Enterprise SSD Vendors Post Over 50% QoQ Revenue Growth in 4Q25, with SK Group Leading Growth

The widespread adoption of AI inference workloads in 4Q25 significantly raised requirements for data storage systems, according to TrendForce's latest findings on the enterprise SSD market. Meanwhile, enterprises accelerated upgrades to their general-purpose servers, while shortages in HDD supply pushed some demand toward SSD solutions. These factors together drove combined revenue for the world's top five enterprise SSD vendors up 51.7% QoQ to more than US$9.9 billion.

Among leading vendors, Samsung retained its top position, reporting nearly $3.66 billion in revenue in 4Q25, up 49.7% QoQ. The company demonstrated the strength of its vertically integrated business model. As concerns grew over potential DRAM shortages, many customers turned to Samsung because it can rely on its own DRAM and NAND Flash production to ensure stable SSD supply.

(PR) Colorful Launches the iGame GeForce RTX 50 Ultra Series

Colorful Technology Company Limited, a leading brand in gaming PC components, gaming laptops, and Hi-fi audio products, proudly introduces the new iGame GeForce RTX 50 Ultra Series graphics cards. The Ultra Series continues the brand's signature hip-hop art-inspired design language, blending bold aesthetics with high-performance gaming hardware, now presented in a striking black color scheme.

While the white iGame Ultra W Series emphasizes futuristic purity, the iGame Ultra Series adopts a black colorway that delivers a more aggressive personality and deeper character. Gradient accent colors transition from dark into vibrant neon pink, complemented by dynamic gradient lighting effects that resemble flowing neon trails in the night. The result is a design that radiates rebellious street-style energy while maintaining the distinctive Ultra identity.

Reedle – Read papers on your phoneβ€”AI chat and spaced review


Reedle is a reading app that keeps your articles and papers in one place and helps you remember them. You can import web pages, PDFs, EPUBs, and your Zotero library, read in a clean layout, ask questions about the text, get help with formulas, translate passages, and listen to them out loud. When you highlight something important, Reedle schedules it for review using spaced repetition, so you don’t forget it later. It works on iOS, Android, and Chrome, lets you search across your whole library, and exports your notes to tools like Obsidian.

View startup

initdesk – The help desk for solo founders and modern teams


initdesk is a simple, AI-powered help desk built for product creators and lean teams who handle support over email. Connect your inbox, receive AI-drafted replies, use auto-tags, and access an AI Assistant you can teach with your FAQs and documents to improve accuracy over time. Keep everything in one lightweight shared inbox, publish a support portal, and trigger workflows with Slack notifications and webhooks. It's designed to save hours, reduce costs, and maintain fast support without enterprise bloat.

View startup

Authorities Disrupt SocksEscort Proxy Botnet Exploiting 369,000 IPs Across 163 Countries

A court-authorized international law enforcement operation has dismantled a criminal proxy service named SocksEscort that enslaved thousands of residential routers worldwide into a botnet for committing large-scale fraud. "SocksEscort infected home and small business internet routers with malware," the U.S. Department of Justice (DoJ) said. "The malware allowed SocksEscort to direct internet

Glorious Revives Original 69 g Model O Gaming Mouse With 19K Optical Sensor, Honeycomb Shell

Glorious Model O was once held in high regard in the realm of gaming peripherals, but that design has since been surpassed by more modern designs with lighter shells and more responsive, accurate sensors. Apparently there is still a market for the original Glorious Model O gaming mouse, at least according to Glorious, who recently announced the Glorious Model O Classic. The Model O Classic is a rehash of the 2019 Glorious Model O design, replete with the RGB strips, original sensor hardware, and even the original honeycomb design that's largely fallen out of use except for in the most lightweight gaming mouse designs.

The Glorious Model O classic features the same BAMF sensor, based on the Pixart PAW 3370, as the originalβ€”although it's now simply called the "19K Optical Gaming Sensor," which means it has a maximum wireless polling rate of 1 kHz, tops out at 19,000 DPI, 400 IPS, and 40 G acceleration. Despite the honeycomb shell, the Model O Classic still weighs in at 69 g, whereas many solid-shell gaming mice released in the last year or two come in at 40-50 g, with some, like the Corsair Sabre V2 Pro even dipping into the mid-30s. Glorious also still uses the Glorious Core software to customize the Model O Classic, meaning customization requires installing software that is only available on Windows. The Model O Classic wireless gaming mouse is available on the Glorious online store at $89.99.

BiteBurst – Help kids 6–14 build healthy habits through gamified learning


BiteBurst is an AI-powered learning platform that helps kids aged 6–14 build healthy habits through short, gamified lessons on nutrition and movement. Kids follow a playful path of cards, quizzes, and challenges, tapping emojis to explore foods and activities and earning collectible cards that keep motivation high. Friendly mascots guide every step while parents track progress, streaks, and knowledge growth with clear insights and multi-child support.

View startup

Veeam Patches 7 Critical Backup & Replication Flaws Allowing Remote Code Execution

Veeam has released security updates to address multiple critical vulnerabilities in its Backup & Replication software that, if successfully exploited, could result in remote code execution. The vulnerabilities are as follows - CVE-2026-21666 (CVSS score: 9.9) - A vulnerability that allows an authenticated domain user to perform remote code execution on the Backup Server. CVE-2026-21667 (

Dream Select Boot Camp – Pass your real estate exam with an AI mentor and mock tests


Dream Select Boot Camp is an all-in-one platform for aspiring and active real estate agents in all 50 states. Prepare for your state exam with mock exams, topic quizzes, and speed rounds, and keep momentum with a 24/7 voice AI mentor for on-demand guidance.

Use the Agent Toolkit to manage leads with a built-in CRM, generate listings, and follow a social media playbook. Join The Bullpen community to swap advice and stay motivated. Start with a 14-day free trial and get support anytime.

View startup

ChatReact – AI chatbots that get smarter over time


ChatReact is an AI-powered customer support platform that learns over time. When customers mark an answer as wrong, the Improvement Agent analyzes the issue and suggests fixes automatically. It crawls your website, builds a Knowledge Base, generates SEO-friendly FAQs, and supports 24 languages. Built for businesses seeking more than a basic chatbot, ChatReact includes a full REST API and MCP Server for developers and is GDPR compliant. A free tier is available.

View startup

PEGI Updates Game Ratings To Address Loot Boxes and Potentially Addictive Gameplay Mechanics

Hot on the heels of Valve's critical public response to the New York State attorney general's lawsuit against the gaming giant over its in-game monetization and loot boxes, the Pan European Game Information (PEGI) ratings body has updated its rating system to address similar concerns over in-game monetization and "interactive risk categories." As of June 2026, new games registered in Europe and rated by PEGI will be subject to additional scrutiny surrounding "purchases of in-game content, paid random items, communication features, and features that incentivize players to continue playing," according to the announcement. In essence, this will examine daily rewards, loot boxes, in-game cosmetics and premium currency purchases, and in-game chat moderation and content control. Chair of the PEGI Council, Beate VΓ₯je, says that "with the updated set of age rating criteria, PEGI aims to make parents aware that certain features in games should be carefully assessed, and that parental tools can be a very helpful assistant when doing that."

Under the new criteria, games that feature in-game purchases, specifically those with quantity- or time-limited sale mechanics, will be rated 12+, while games that use NFTs and other blockchain tech will be rated PEGI 18. Games with "random items" will be PEGI 16 by default, although they will be PEGI 18 in some cases. Games with what the ratings body calls "play-by-appointment" mechanicsβ€”things like daily quests or login rewardsβ€”will be rated PEGI 7, while games that punish players for not playing daily will receive a PEGI 12 rating. Lastly, PEGI is enforcing age restrictions on "safe online gameplay," which will result in games with entirely unrestricted communication receiving a PEGI 18 rating. Conceivably, games with proximity voice chat that cannot be disabled will also fall into this last box, although PEGI specifies that there needs to be a way to block or report users, so it's possible that being able to mute proximity chat may not be enough to escape a PEGI 18 rating on the grounds of online safety.

Crowdfunding AI – Get a proven roadmap to launch your crowdfunding campaign like a pro


Crowdfunding AI helps founders plan, budget, and execute successful crowdfunding campaigns. It generates an instant strategy with timelines and tasks tailored to your product and niche, showing what worked in similar launches so you can model proven tactics. It blends expert frameworks with data-driven guidance and free resources to take you from validation through post-campaign with clarity and confidence.

View startup

Nabbed – Run your job search and career like a sales pipeline


Nabbed.io is a career CRM built exclusively for revenue professionals β€” SDRs, AEs, Account Managers, and sales leaders. It reframes the job search as a sales motion. What separates Nabbed from other job search tools is what happens after you get the job. Career Mode activates when you land the role. It keeps your network warm, tracks compensation benchmarks against the market, surfaces hiring signals at target companies, and logs wins as resume-ready bullets in real time.

View startup

How to watch Jensen Huang’s Nvidia GTC 2026 keynote

GTC β€” which stands for GPU Technology Conference β€” is Nvidia's flagship annual event, where the chipmaker typically uses the spotlight to announce new products, champion partnerships, and lay out its vision for the future of computing. Huang's keynote will focus on Nvidia's role in the future of computing and AI.

Crimson Desert Gets Denuvo DRM Ahead of Launch, Dev Confirms Day-One Patch and No Cross-Save

Pearl Abyss's Crimson Desert is just over a week away from its March 20 launch, and in a recent Steam News post, the studio behind the game has provided additional information about the release, while the Steam Store page now reflects that the game will launch with Denuvo DRM. The news update confirms that players jumping into the world of Pywel will be able to preload the game 24-48 hours ahead of the official launch, starting on March 17 or 18, depending on which part of the world they're inβ€”the game goes live on March 19 at 10 PM UTC, so the exact launch date differs by region. Pearl Abyss also confirmed that Crimson Desert will not support cross-platform save, and there appears to be no plan to add it in at a later date. While Crimson Desert will be playable completely offline, Pearl Abyss warns that the game will need an internet connection on day one in order to install a day-one patch.

It's unclear how the addition of Denuvo DRM will affect gameplay, especially since Pearl Abyss set a pretty low barrier to entry with the minimum hardware requirements that it published earlier this week. Denuvo, and other DRM services, have, in the past, been known to cause performance degradation, even on higher-end hardware, but if the game was tested for the minimum requirements with the DRM in place, players can likely expect little to no performance hit. While Denuvo doesn't necessarily lock Linux users out of Crimson Desert, its use is generally frowned upon by the Linux gaming community. Pearl Abyss has not mentioned Steam Deck or Linux compatibility in any of its communications surrounding Crimson Desert.

Chinese Laptop Maker Chuwi Advertised AMD Ryzen 5 7430U SoC, but Shipped the Older Ryzen 5 5500U

Imagine buying a laptop, thinking you're getting a model with your desired CPU specifications, only to find a completely different model inside, cleverly concealed so you wouldn't notice without further investigation. According to an investigation by Notebookcheck, the Chinese electronics maker Chuwi is engaging in specification fraud. Users have discovered a different CPU SKU compared to what was advertised. In a review of the Chuwi CoreBook X and CoreBook Plus, Notebookcheck found that Chuwi had listed these laptops with an AMD Ryzen 5 7430U processor, but they actually come with an AMD Ryzen 5 5500U. This means that Chuwi is actively advertising these models on their website, on the laptop box, with laptop stickers, and even with BIOS modifications to make it seem as if they feature the newer Ryzen 5 7430U SoC with "Zen 3" CPU microarchitecture and Radeon "Vega 6" SoC. In reality, the company is shipping a processor that is a few generations old, with "Zen 2" CPU cores and Radeon Graphics 448SP.

Notebookcheck discovered this during a review of the Chuwi CoreBook X. The unit's performance was rather lackluster, prompting further investigation. Initially, they thought single-channel RAM was causing the subpar performance. To determine the true cause, they opened the unit and found the older Ryzen 5 5500U SoC with its corresponding part number 100-000000375. Additionally, other differences were noted in the SoC specifications, such as the L3 cache. Chuwi even went a step further by using a modified BIOS version to show that the unit features the AMD Ryzen 5 7430U. This led software diagnostic tools to display the advertised specification, while the lower-tier SoC was actually inside. This issue didn't occur just once with the Chuwi CoreBook X but also with a separate CoreBook Plus, which Notebookcheck also confirmed to feature the older SoC. Instead of the newer AMD Ryzen 5 7430U processor, both of these laptop models actually come with the AMD Ryzen 5 5500U, as verified through performance testing, and the OPN number that corresponds to a specific AMD chip.
❌