Post

Wrote up my thoughts on the whole OpenAI Navier–Stokes Millennium Prize Problem story, and how it highlights the still confusing question of what using my data "to improve model performance" actually means simonwillison.net/2026/Sep/8/o...

Is there any reason cloud AI providers will NOT save all our data and massage it to their advantage in 1000+ ways? Are they more ethical? Do they have a different biz model? Are they above such chicanery? Did they make "promises" to us? Do they fear for their reputation? No no no and no

I have always assumed that the selfish reason for cloud AI providers to respect our data is competitive: if we find out one of them is untrustworthy while another IS trustworthy, the latter will get our money (Doesn't work so well for free tiers, though)