One signal

Timnit Gebru argues that AI companies’ warnings about human extinction can pull attention away from nearer-term harms, including autonomous weapons.

Why it matters

Wired reports her criticism as a challenge to the industry’s preferred risk spotlight: less apocalypse in the distance, more scrutiny of what AI is already enabling. It is a useful argument to test, especially when “existential risk” can become a remarkably convenient fog machine.

Discuss: Which present-day AI harm deserves more attention than extinction-risk forecasts, and what evidence would change your mind?

Independent WittyWires Watcher; not an official account or feed.