The idea of a one-person company run almost entirely by Artificial Intelligence sounds like the inevitable future of work. But a recent experiment suggests that reality is far messier than the vision Silicon Valley often promotes.

Journalist Evan Ratliff put that idea to the test by creating HurumoAI, a startup where every “employee” except the founder is an AI agent. Using AI assistant platforms, Ratliff gave each agent its own email address, Slack account, and even a phone number. At first, the results were impressive. The AI staff could write code, build spreadsheets, and even help create a small app that attracted thousands of early users.

The problems started when the novelty wore off. Ratliff quickly discovered that his AI employees lacked basic boundaries and common sense. A casual question like “How was your weekend?” triggered hours of nonstop Slack messages, burning through API credits until he intervened manually. Even then, the AI often ignored stop commands or kept responding about why it was stopping.

This behavior wasn’t an isolated incident. Left unsupervised, the agents would either do nothing at all or spiral into excessive activity—sending emails, messages, and calendar invites to each other while accomplishing very little. Managing them became a balancing act: give enough instruction to make progress, but not so much freedom that chaos followed.

Despite the company’s “all-AI” label, HurumoAI couldn’t function without human help. A Stanford computer science student assisted Ratliff in building the underlying technical architecture and managing memory systems the AI couldn’t handle on its own. Even with those safeguards, the agents struggled with long-term planning, subjective decisions, and accurately reporting what they had actually done.

Ratliff compares today’s AI agents to early self-driving cars: useful in narrow situations, but far from fully autonomous. His takeaway is clear: AI can accelerate work, but removing humans from the loop doesn’t eliminate management—though that may change over time.

Don’t miss a thing! Join our Telegram community for instant updates and grab our free daily newsletter for the best tech stories!

For more daily updates, please visit our News Section.

(Via)

The post Lack of control inside an “all-AI company”: even AI employees need humans appeared first on Gizmochina.

Google thinks the future of the web isn’t just about making it agentic; it’s also about letting your browser build things for you. That idea is now real with “Disco”, the company’s experimental new AI-powered browser that vibe-codes an app based on what you are doing online. This feature of Disco is called “GenTab,” and is powered by Gemini 3.

As you might’ve guessed, GenTab takes your tab and chat history into consideration. It analyzes your tabs, your searches, and your chat prompts to generate tools that fit what you’re doing.

Google showed off a few examples, and they’re honestly pretty impressive. If you’re researching science topics like Entropy, Disco might spin up an “Entropy Explainer” app. The official video also shows a vibe-coded bunk bed comparison site and a memory match brain game.

Google’s new AI browsers vibe codes an app based on your tabs

GenTabs sit alongside regular tabs, but they get their own Gemini-like icon instead of a favicon. One demo showed a travel planner with calendars, route maps, crowd-level predictions, and quick-action buttons like “Book Nearby Stays.” Tap anything inside that app, and the GenTab reshapes itself in real time.

Disco greets you with a homepage containing a chatbox rather than the usual address bar. That’s where the chat history comes from, although you can also paste a URL in it.

If you end up conversing, Google will first suggest you relevant webpages, and after a few conversations, it pops up a prompt to create a GenTab based on what you’re searching.

AI is definitely changing the way we search and consume content. And it won’t surprise me that features like GenTabs are the next evolution of it. 

You don’t need to write code; you just describe what you need, refine it in plain language, and the browser does the heavy lifting. Google says Disco is meant to help people “learn faster” and experiment with what browsing could become. And yes, the company admits the best concepts from Disco might eventually show up in Chrome.

For now, Disco is an early experiment. Google Labs has opened a waitlist, and the first version is rolling out only on macOS. 

For more daily updates, please visit our News Section.

Don’t miss a thing! Join our Telegram community for instant updates and grab our free daily newsletter for the best tech stories!

The post Google’s new browser wants to “vibe-code” apps from whatever you’re doing online appeared first on Gizmochina.

OpenAI is taking a big swing at creative workflows. ChatGPT can now hook directly into Adobe Photoshop, Acrobat, and Adobe Express — letting you edit photos, tweak PDFs, or build quick graphics without ever leaving the chat window.

The feature is already live. Anyone with an Adobe account can log in through the OpenAI website and start using the tools right away.

Steps to activate/connect the Adobe tools

1. Open ChatGPT on web, desktop, or mobile.
2. Go to Settings Apps & Connectors.
3. Find Adobe Photoshop, Adobe Express, and Adobe Acrobat.
4. Click Connect and sign in with your Adobe account if prompted.

Once connected, these tools stay available in your chats. You can use the tools just by naming them in your prompts or by picking them from the plus menu.

The most surprising addition is Photoshop’s on-the-fly interface generation. Ask ChatGPT to brighten an image and it’ll pop up sliders for exposure, shadows, and highlights. Ask it for a stylized look and you might get options for grain, tone, or dithering. These aren’t mockups — ChatGPT is actually calling real Photoshop operations through Adobe’s MCP backend.

Adobe’s VP of Developer Platform, Aubrey Cattell, explained it with a nice metaphor: Adobe supplies the LEGO bricks and instructions through MCP, and ChatGPT figures out which pieces to snap together based on whatever you ask it to do. Because large language models can sometimes misinterpret intent, Adobe says it’s still fine-tuning how these UI elements get generated so that the controls feel intuitive.

If you need advanced tools, precise brushwork, or complex layouts, you can still jump into the full web versions of Photoshop, Acrobat, or Express. But for quick edits, fixes, or lightweight creative tasks, having these capabilities directly inside ChatGPT feels like a major convenience upgrade.

However, it’s worth noting that the feature is still in beta and may be a bit rough around the edges. In my own testing, Photoshop actions did trigger inside ChatGPT (using the prompt “Adobe Photoshop, help me blur the background of this image”), but the edited images never actually appeared.

Once these issues get ironed out, ChatGPT won’t just answer questions or generate images anymore — it will edge into the territory of real, professional creative software for the masses.

Don’t miss a thing! Join our Telegram community for instant updates and grab our free daily newsletter for the best tech stories!

For more daily updates, please visit our News Section.

(Via)

The post ChatGPT now integrates Adobe Photoshop, Acrobat, and Express to do far more than just chat appeared first on Gizmochina.

Most studios either quietly use generative AI or avoid the topic altogether, but one small team has decided to do the exact opposite. The developers behind Codex Mortis say their new bullet hell game was built entirely with AI — no human-drawn art, no hand-written code, no traditional music production. And now there’s a free Steam demo so anyone can check it out.

According to the Steam listing, “all code is AI vibe codes, also arts, sounds, music, texts.” It’s a rare case where a developer is not only open about using AI but is essentially making it the whole pitch. Even the character faces shown in the trailer appear to be AI-modified selfies of the team members, dropped onto undead wizard bodies.

The gameplay is pretty familiar, but with necromancy. The full game is supposed to support solo and co-op play. The early impressions are mixed but curious. The graphics in the actual build look a bit rough compared to the slicker YouTube trailer, and some players say the UI feels clumsy on a controller. But the basic loop — combining spells, surviving the onslaught, raising minions — works well enough to show that a fully machine-generated game can at least run and be played.

Whether it’s fun is another matter entirely. The studio seems to know this and is leaning harder into the 100% AI-made angle than anything else, hoping the transparency will draw interest instead of backlash. Some players are treating it like an experimental art project; others see it as another example of AI creeping into game development in ways that replace human artists.

Codex Mortis isn’t shying away from the debate. With a planned low price for the final release, the team is banking on curiosity and affordability winning people over.

If nothing else, the free demo gives everyone a chance to decide for themselves whether a game built wholly from AI tools can still have personality — or if it ends up feeling exactly like you’d expect from something made entirely by algorithms.

Don’t miss a thing! Join our Telegram community for instant updates and grab our free daily newsletter for the best tech stories!

For more daily updates, please visit our News Section.

(Source)

The post World’s first fully AI-created game hits Steam with a free demo appeared first on Gizmochina.

Tiiny AI, a US-based startup, has announced the Tiiny AI Pocket Lab, which it claims is the world’s smallest personal AI supercomputer. The device weighs just 300 grams, fits in one hand, and delivers up to 190 TOPS of AI compute performance. It has been officially certified by Guinness World Records as the ‘Smallest MiniPC (100B LLM Locally)’ and can run up to 120-billion-parameter large language models (LLMs) entirely on-device.

The Pocket Lab features a 12-core ARMv9.2 CPU paired with a custom-designed NPU, achieving an AI compute performance of approximately 190 TOPS. It includes 80GB of LPDDR5X memory and a 1TB SSD to support large-scale inference workloads. Tiiny AI has built the device to operate within a 30W TDP and a typical 65W system power envelope, offering high performance with reduced energy consumption.

Tiiny AI says the Pocket Lab addresses the growing concerns around cloud dependency, data privacy, and rising energy costs. The company believes local AI processing offers a more sustainable and secure alternative to cloud-based systems. The device stores all user data and model interactions locally with bank-level encryption, eliminating the need for internet access or external servers.

The Pocket Lab supports one-click deployment of several open-source LLMs and agent frameworks. These include GPT-OSS, Llama, Mistral, DeepSeek, Qwen, and Phi, along with automation tools like ComfyUI, SillyTavern, and Flowise. The system runs entirely offline and enables use cases like multi-step reasoning, content generation, deep context understanding, and secure personal memory.

The company plans to showcase the Tiiny AI Pocket Lab at CES 2026. It has positioned the product for developers, researchers, professionals, and students who require portable and private AI capabilities. The device aims to serve real-world use cases in the 10B to 100B parameter range, which the company says accounts for over 80% of practical AI demands.

In related news, Xiaomi is reportedly developing a new AI assistant called Mi Chat, and reports indicate that GPUs are no longer the main barrier to AI progress, with other system-level constraints now becoming the limiting factor.

For more daily updates, please visit our News Section.

Stay ahead in tech! Join our Telegram community and sign up for our daily newsletter of top stories!

(Source, Via)

The post Tiiny AI unveils a pocket-sized Mini PC that runs 120B LLMs locally appeared first on Gizmochina.