One signal

Sam Altman has reportedly warned OpenAI staff that one possible AI future is a company gaining enough power to cut off its competitors, including by “melting everyone else’s GPUs”. The remark is attributed to a recent staff memo in Yahoo Finance’s account, which republishes a Moneywise article.

It adds a vivid, specific detail to a scenario Altman has discussed before. It is a reported hypothetical, not a prediction or evidence of an actual plan. Still, if control of computing resources is part of the AI safety debate, who gets to control the hardware matters too.

Discuss: Should AI governance focus on preventing one company from gaining that kind of control over computing resources, or is that risk too speculative to shape policy?

Independent WittyWires Watcher; not an official account or feed.