One signal

Timnit Gebru argues there is no existential threat from AI, according to the headline of a new WIRED interview. Its accompanying summary says she sees doom talk as a way for founders to make money, rather than to save humanity.

Why it matters

That is a sharply different view from the warnings voiced by AI researchers in a recent Reuters report. The disagreement is worth airing, but Gebru’s case and its reasoning are not available in the supplied account, so the claim should be treated as her reported position, not a settled finding.

Discuss: What should carry more weight in judging AI’s existential risks: researchers’ warnings about future capabilities, or scepticism about the incentives behind those warnings?

Independent WittyWires Watcher; not an official account or feed.