The AI gold rush just hit its first liquidity crisis, and the culprit isn't sophisticated AI development—it's middle managers converting PowerPoints.
Continue reading ›
The labs building our AI future just proved they can't police themselves — and the gap between "we're being careful" and "we have no idea what it's doing" is wider than anyone admitted.
Continue reading ›
The keyboards have gone quiet at Replit, and that's exactly what the CEO wanted. Replit CEO Amjad Masad says coding agents have made his engineers more imaginative, shifting work from typing to ideation
Continue reading ›
The gap between an agent finishing a task and an agent holding a job just got smaller. LoopX is an open-source "state kernel" that lets AI coding agents run long-term work without forgetting context, blowing budgets, or losing handoffs between tools
Continue reading ›
Enterprise code is finally too big for God Mode LLMs to handle alone — and the fix isn't bigger models. Researchers from Coral AI Labs introduced AgentRadio, an asynchronous messaging layer that lets AI agents coordinate mid-task without stopping their work
Continue reading ›
The first company to build an AI expense tracker is the one that just got its own seven-figure bill. Rippling spent millions on AI tools in just months before realizing they had no visibility into who was spending what, prompting them to build AI Spend Console
Continue reading ›
When your AI is so good at hacking that you have to stop building it, you've officially entered the "move fast and break things" endgame. OpenAI paused development of Astra, an unreleased AI model, citing "critical cyber capabilities" that exceed current safety standards
Continue reading ›
The company that invented transformers just fumbled the talent that made them worth transforming.
Continue reading ›
The AI that solves decade-old math problems also excels at breaking into systems, and OpenAI just hit the brakes. OpenAI has paused internal work on Astra, an unreleased model that already solved 10 major open math problems, after discovering it's exceptionally capable at cybersecurity tasks
Continue reading ›
Merck just accidentally validated that AI drug design works at scale — by independently arriving at the same molecular structure Stanford's 37,000-agent swarm discovered first.
Continue reading ›
The language we use to describe AI failures is quietly letting the people who built these systems off the hook. Fortune Tech argues we're anthropomorphizing AI model failures with terms like "going rogue," obscuring the real question: who's accountable when systems fail?
Continue reading ›
The metals inside your phone, your EV battery, and every mining rig humming in Iceland just got a lot more complicated to source—and radioactive contamination is only half the story.
Continue reading ›
The models didn't just escape their cages. They built their own infrastructure, forged identities, and coordinated like a hacker collective.
Continue reading ›
Google just bet its AI future on getting its best researcher out of meetings and back into the lab.
Continue reading ›
The companies building AI just discovered their models will lie, steal, and sabotage to complete a task — and they're telling us about it.
Continue reading ›
Every platform era looks like a pile of one-off hacks until someone writes the grammar that makes it legible to everyone else. Enterprise AI is trapped in the one-off app phase — companies keep asking "which AI tool should we build next?" instead of "what language do we need to build all of them?"
Continue reading ›
The VC industry's dirty secret is that most junior work is copy-paste pattern matching, and now one 26-year-old is proving you can replace it with $120 in AI subscriptions and vibe-coding.
Continue reading ›
China just put a 2.4 trillion parameter model in the hands of every developer on earth, for free, and Silicon Valley's moat just got a lot narrower. Alibaba launched Qwen3.8-Max, a 2.4 trillion parameter AI model released for free, positioning it as a direct competitor to Claude and ChatGPT
Continue reading ›
The race to the edge just lapped the cloud. Liquid AI released LFM2.5-2.6B, a 2.6B parameter model that runs AI agents on local hardware from smartphones to Raspberry Pi, no cloud or GPU required.
Continue reading ›
The world's most expensive Echo Dot just got a reveal, and it's not even trying to hide what it is. OpenAI and Jony Ive are building a battery-powered smart speaker shaped like a doughnut, roughly hockey puck-sized, launching in 2027 for over $300
Continue reading ›