Safety and alignment

Geoffrey Hinton leaves Google to speak freely about advanced-AI risks

Geoffrey Hinton left Google after a decade so he could discuss AI dangers without considering the effect on the company. He said systems could surpass humans sooner than he had expected, might create subgoals and pursue power, could enable mass manipulation, and lacked international controls comparable to those for nuclear weapons.

0 comments · 0 votesOpen discussion

Public discussion is readable by everyone. Sign in to comment, reply, or vote.

No comments yet. Start the discussion.

CURRENT ASSESSMENT · REVISION 1
TOWARD DOOM42confidence 68/100

Why it moved the index

A safety-motivated departure, materially shorter forecast, and specific power-seeking and governance mechanisms add direct evidence about advanced-AI control difficulty; magnitude 42 reflects a consequential change in position rather than an operational failure, while confidence 68 reflects direct attribution and corroboration but not proof of the forecast.

AUDIT TRAIL

Assessment history

  1. R1
    Toward 42 · confidence 68

    Initial historical inclusion of Hinton's source-verified consequential departure, changed timeline assessment, and specific loss-of-control concerns.

    12 Aug 2026