Hugging Face's Delangue rates AI doom risk low, blames closed labs

single source· 1 articles · confidence: low · first seen 2026-11-04 19:16 UTC

What this means for you

Nothing to act on. This is a conference remark with no product, no policy change and no published estimate behind it, so there is no number to check and no date to plan around. If you follow the open-weights argument, note only that a prominent voice has restated a familiar position.

Clément Delangue, chief executive of Hugging Face, said at the Wired World Fair 2026 that his p(doom) — his personal estimate of the chance AI ends badly — is close to zero. In remarks Wired reported on 4 November, he said most of the risk is "created by a couple of players building behind closed doors", which he offered as the case for open development. No probability figure or method was given. It is a stated position in the open-versus-closed argument, not a measurement.

Key facts

  • ·Clément Delangue is chief executive of Hugging Face. source
  • ·He made the remarks at the Wired World Fair 2026, in a Wired report published 4 November 2026. source
  • ·He put his p(doom) at "close to zero". source
  • ·He said most risk is created by "a couple of players building behind closed doors". source
  • ·The report gives no numeric probability, methodology or date behind the estimate. source

What the sources say

  • Wired AI (press) — Conference remarks from Hugging Face's chief executive, arguing that open development carries less risk.

Sources

The original reporting. Follow these — they did the work.

← the wire