Google just open-sourced the Python framework it uses internally to build production AI agents — and the 2.0 release suggests they've learned something the rest of the industry hasn't figured out yet.
Continue reading ›
Europe's AI darling just crossed $6 billion in valuation, and unlike OpenAI, you can actually download their models and run them yourself. Mistral AI, founded in 2023 by former DeepMind and Meta researchers, has emerged as Europe's leading AI company with a hybrid open/closed model strategy
Continue reading ›
Google just made every coding assistant smarter at building agents than most engineers who've been doing it for years. Google released agents-cli, an open-source toolkit that teaches any AI coding assistant how to build, evaluate, and deploy production-grade agents on Google Cloud
Continue reading ›
Apple just gave every AI coding assistant direct access to Safari's guts — no permission slips, no App Store review, just a WebKit protocol that turns debugging into a multiplayer game between humans and agents.
Continue reading ›
The richest man in Europe just bet €50 million that math — not prompt engineering — is the moat in the AI century. Bernard Arnault pledged €50 million to build a math research institute at his alma mater, calling it essential infrastructure for the AI era
Continue reading ›
Sam Altman just proposed the most American thing imaginable: a federal equity stake in a private AI company to calm the mob and charm the White House. OpenAI has floated giving the US government a 5% ownership stake to ease Trump administration tensions and blunt public AI backlash
Continue reading ›
OpenAI might hand Washington a slice of the company while SoftBank tries to borrow $10 billion against that same equity, and if you think those two moves aren't connected, you haven't been paying attention to how AI money works now.
Continue reading ›
The accuracy rate isn't the story. The incentive structure is. Google's AI Overviews are accurate 91% of the time, per a New York Times-commissioned study by AI startup Oumi, which means millions of wrong answers per day at Google's search scale.
Continue reading ›
Google finally ships new hardware for the smart home, but the AI that's supposed to justify it is still half-baked. Google launched its first new smart speaker in six years, the Google Home Speaker, marketed as "built for Gemini"
Continue reading ›
Google spent 25 years training marketers to game its algorithm, and now they're all scrambling to figure out which startups can help them game the new gatekeepers.
Continue reading ›
The US government just blinked in its first real showdown with a frontier AI lab, and the precedent matters more than the model.
Continue reading ›
The researchers who taught AI to bluff are now teaching it to beat the market, and Wall Street is writing half-billion-dollar checks. EquiLibre Technologies, founded by three ex-DeepMind researchers in Prague, hit a $500 million+ valuation building AI for quantitative hedge funds
Continue reading ›
Anthropic just made its best agent-building tech affordable right before going public, which means either they're confident enough to commoditize their moat or desperate enough to juice adoption numbers for Wall Street.
Continue reading ›
Google just made AI image generation cheap enough that your agents can afford to be wasteful. Google launched Nano Banana 2 Lite, aka Gemini 3.1 Flash-Lite Image, generating images in 4 seconds at $0.034 per 1,000 images through Google AI Studio, the Gemini API, and GEAP.
Continue reading ›
Cathie Wood just bet $25 million that the on-ramps to crypto matter more than crypto itself. Ark Invest bought $6.85M in Coinbase, $6.21M in Circle, $3.54M in Bullish, and shares in Robinhood on Monday as crypto stocks rallied
Continue reading ›
Google just turned your Gmail, Docs, and search history into fuel for an image generator—and they're betting you'll want that more than you'll worry about it.
Continue reading ›
The companies racing to replace human workers just discovered they're all fighting over the same set of GPUs, and Google's rationing access like it's wartime sugar.
Continue reading ›
NVIDIA just dropped the x-ray machine for AI agents — turns out one in four agent skills ships with exploitable holes. NVIDIA released SkillSpector, an open-source security scanner for AI agent skills that detects vulnerabilities before installation
Continue reading ›
Turns out the killer app for unrestricted AI isn't creative brainstorming or coding assistance—it's OnlyFans without the humans.
Continue reading ›
The Valley's frontier narrative just hit a European wall—and it's lined with €1.7 billion in fresh capital. Mistral AI went from zero to €400M ARR in under three years, building frontier models from Paris with a fraction of Silicon Valley's compute budget
Continue reading ›