A former OpenAI and Anthropic researcher resigned with a warning that went viral, reigniting the "should we slow down?" debate. Then OpenAI said an internal model solved the Navier–Stokes Millennium Prize problem - followed by a credit controversy with an ...