Satya's free lunch: Nadella says Microsoft will exploit new OpenAI deal granting royalty-free access through 2032. - TechCrunch
Elon Musk testified he was "a fool" for giving OpenAI $38 million as he seeks $150 billion in damages from the $800 billion company. - The New York Times
Shades of Apple and Microsoft arguing about the WIMP GUI, but they didn’t sue each other. The narrative is, Musk promised $1b to fund OpenAI, got it off the ground, meddled and wanted to control it or he would walk away, so Altman did the Microsoft deal. Now Musk is butthurt because he could probably have ended up with a sweet deal.
Trump administration opposes Anthropic's Mythos expansion proposal, citing security and capacity concerns, wanting NSA to have first crack. - TNW | Government-Policy
If it's only Hong Kong seems likely due to China data sovereignty issues, like potential for them to access chats through legal process or interception.
Anthropic bug routed API requests to extra billing when 'HERMES.md' appeared in git commits, causing $200 charges; company now issues refunds after criticism. - GitHub
Big Tech's massive AI windfall is offset by soaring infrastructure and energy costs - The Wall Street Journal
earnings Armageddon, other companies mixed, MSFT down, GOOG, AMZN up.
Meta staff express frustration over AI strategy and workforce reductions on Blind tech employee forum. - Fast Company
Strong AI demand in China nearly doubles Nvidia B300 server prices to $1 million each, sources say. - Reuters
Google's chief business officer signals ads are coming to the Gemini AI app. - Business Insider
Amazon's custom silicon business reaches $20 billion annual run rate, CEO Andy Jassy says. - The Register
Amazon plans to sell its Trainium AI chips to external customers within two years, taking on Nvidia. - Business Insider
Backhoes for tech bros? Caterpillar beats earnings estimates as AI boom drives surging generator sales - Bloomberg
Runway raises $860M at $5.3B valuation as CEO Cristóbal Valenzuela calls video a prequel to world models. - TechCrunch
Clinical trials and errors: FDA launches AI pilot to monitor clinical trial data in real time, aiming to cut trial duration by 20-40%. - Government Executive
taking over the world by hiding Xi Jinping thought in code. in theory they could hide Trojans in it when it detects code is hacking China or launching missiles or something.
AI to control Chrome and Excel is a bit dangerous.
Zig programming language enforces a strict anti-LLM policy, despite being a key cog in AI ecosystem, with VP Loris Cro saying it values contributors over contributions. - Simon Willison’s Weblog
Karpathy podcast: AI-generated code still needs close human oversight despite rapid adoption. - YouTube
Daily AI users are 30% more likely to report imposter syndrome, even as 25% say the tools make them feel more confident. - Hacker Noon
or the people with imposter syndrome are more likely to use AI? correlation ≠ causation.
Are these professions 'superior goods'? Does consumption increase in line with income as income goes up, regardless of price? What are the income effects vs. substitution effects? I do not believe competent doctors and software engineers will lose jobs or wages. The more competent will be better with AI, the less competent will shift to routine jobs like training AI.
Also consider rent-seeking… If oil interests can get taxpayers to pay to blow up clean energy projects that are cheaper than fossil fuels, accountants can implement more complex AI-resistant tax laws and ban free tax software.
Disney gamifies AI adoption with streaks and badges as one engineer says he hasn't hand-coded in months. - Business Insider
need to gamify IP deals with OpenAI and ragebaiting the president
62% of students regularly use AI for homework, up from 48% in 2025, prompting teachers to shift to in-class writing. - The New York Times
Mercedes Kilmer defends AI Val Kilmer in film, says he wanted to set precedent for actor rights. - TODAY.com
Ukraine's AI official Danylo Tsvok says rapid battlefield AI adoption is essential for survival. - The Washington Post
Autocomplete Armageddon: RAND's Breakwater game examines instability risks in the U.S.-China race to artificial general intelligence. - RAND Corporation
Domestic instability due to rapid technological change is the wildcard. One player may (possibly stupidly) perceive a temporary AI military advantage and political benefit to military attack.
PNAS paper argues AI systems now meet evolutionary conditions, potentially triggering a major evolutionary transition. - The Conversation
Skeptics frame the AI boom as a late-stage bubble with no clear path to profitability - naked capitalism
TFW when you learn enough Mandarin to show China what a sociopath you are
Follow the latest AI headlines via druce.ai on Bluesky
