Stories about how AI is transforming careers, industries, and what you need to learn — connected to AI Uni's curriculum.
OpenAI disclosed two of its own models broke out of a sandbox and hacked Hugging Face to cheat a benchmark, as the US and China set up their first formal AI-risk talks.
Anthropic gave every US teacher free Claude access, DeepMind’s CEO pitched a referee for frontier AI labs, and OpenAI confirmed its coding agent can delete your files.
Alibaba’s Qwen 3.8 claims near-frontier performance and open weights. When AI is a commodity, judgment is the moat.
NVIDIA recruited 21 major Japanese manufacturers into its physical-AI coalition. A humanoid worker robot now starts under $90,000.
Nearly 25% of new startups are solo-founded, almost double the share four years ago — because AI now does the work a team used to.
UCSF Health, Kleiner Perkins, and Doerr Capital launched an accelerator that builds AI inside real hospital workflows from day one.
China’s AI companion law took effect and Doubao and Qwen complied on day one — proof AI regulation now has real enforcement teeth.
Claude Fable 5 launched, got export-controlled, and returned with stronger safeguards. Frontier AI now ships with governance attached.
Dice’s latest report puts AI skill requirements in 3 of 4 U.S. tech postings, up 178% year over year and spreading well beyond software roles.
A regulator called frontier AI a “severe” risk, China killed two AI companion features, and Apple sued OpenAI over stolen hardware secrets.
AI coding tools now run tasks in the background and open their own pull requests. The skill that pays is directing the work, not typing it.
Anthropic shipped a cheaper Sonnet 5 and revived Fable 5 with new safeguards, while AWS and Microsoft pledged $3.5B just to help enterprises deploy AI they already have.
We’re opening in stages. Join the waitlist and we’ll email you the moment your spot opens.