AI News
AI news that matters. Updated daily.
No stories match your filters.
Anthropic Overtakes OpenAI as the World's Most Valuable AI Startup
Anthropic has surpassed OpenAI to claim the title of the world's most valuable AI startup — a milestone that would have seemed far-fetched two years ago, when OpenAI's ChatGPT was still the only name most people knew in consumer AI.
A $6,400 Local LLM Server vs. API Bills: Real Numbers, Real Accounting
$6,400. That's what one developer spent building a local LLM server - a computer dedicated to running AI models without sending data or dollars to any API provider. They then ran the actual accounting on whether it was worth it, and the methodology they used is more important than the final number.
What AI Actually Does for People Who Use It Every Day
Three years of AI headlines have been heavy on capability announcements and light on "here's what I actually use it for every Tuesday." The gap between AI as a concept and AI as a daily work tool is worth closing. So: what do practitioners - the people running it daily, not demoing it at conferences - actually find useful?
NVIDIA Releases Quantized Qwen3 35B MoE in FP4 Format for Local Inference
NVIDIA published a quantized build of Qwen3-35B-A3B on Hugging Face under the nvidia/Qwen3.6-35B-A3B-NVFP4 repository, targeting developers who run large language models locally on NVIDIA hardware.
GitHub Copilot's Token-Based Billing Has Developers Looking for the Exit
Flat-rate AI coding subscriptions may be going the way of the unlimited data plan. GitHubb Copilot is shifting to token-based billing - charging based on how much the underlying model actually processes rather than a fixed monthly fee - and developers are not happy about it.
Meta Reportedly Building an AI Pendant After Smart Glasses Success
Meta's Ray-Ban smart glasses sold better than most analysts expected. Now, according to a TechCrunch report, the company is working on another wearable: an AI pendant.
Google's Gemini Spark Works Well - and That Makes Its Existence Confusing
What happens when a major platform launches an AI product that genuinely works, but nobody can explain why it exists separately from the other AI products it already makes?
The AI Spending Hangover: Companies Start Rationing Tools as Costs Mount
The AI spending hangover is here. After two years of aggressive adoption, companies are installing usage caps, requiring budget approvals for AI subscriptions, and in some cases pulling tools entirely - according to a Wall Street Journal report.
G7 Nations Agree on Shared Definitions for Open-Source AI and Open Weights
The G7 nations have agreed on shared language to distinguish "open-source AI" from "open weights AI" - a definitional clarification that matters far more than it sounds.
Claude's Pushback Habit Is Splitting Its User Base
There's a debate running through Claude's user community right now: the AI pushes back too much, and a vocal group wants it to stop.
Free vs. Paid Transcription Tools: Do the Upgrades Actually Pay Off?
Free transcription software is genuinely good now. Apple's built-in dictation is free. Google Docs has voice typing on every account. And since OpenAI released Whisper - their speech recognition model - as open-source in 2022, dozens of apps have built on top of it, many at no cost. So why are people paying $15 to $30 per month for tools like Wispr Flow?
Kimi AI Blocks Gaokao Questions, Tells Students to Study Instead
China's Kimi AI chatbot, built by startup Moonshot AI, is turning away academic questions during Gaokao season - and it isn't polite about it. Users report that instead of answering, the model tells them to go study.
Claude's UltraCode Burned 1.7M Tokens With No Output - Anthropic Offers No Refunds
1.7 million tokens. That's how many Claude credits one user burned in minutes while testing UltraCode, Anthropic's newly launched multi-agent mode inside Claude Code - with no output to show for it and no refund available.
Mystery Company Ran Up $500M Claude Bill in One Month After Skipping Usage Limits
$500 million. That's what one unnamed company burned through on Claude AI in a single month - not because of a billing glitch or a vendor error, but because nobody bothered to set spending limits on their employee licenses.
OpenAI Built a Data Center in a Terminally Ill Child's Home. The Onion Predicted It First.
The Onion published a satirical video mocking OpenAI's data center expansion by imagining the company building one inside the home of a terminally ill child. As of late May 2026, the joke is no longer a joke.