Governance and control

Rohin Shah warns a short AI pause could reduce safety time

Rohin Shah supports enforceable global slowing, but identifies a countervailing mechanism: hardware, algorithms, and investment could keep advancing during a training pause while safety researchers lose access to intermediate models. A poorly designed pause could therefore reduce capabilities-adjusted safety time without proportionally delaying dangerous systems.

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 DOOM20confidence 62/100

Why it moved the index

The conditional mechanism identifies how a prominent governance intervention could reduce model access for safety work while leaving other capability drivers intact, directly increasing the difficulty of controlling frontier progress; it is reasoned analysis rather than an observed outcome.

AUDIT TRAIL

Assessment history

  1. R1
    Toward 20 · confidence 62

    Historical people-backfill item from a dated full transcript, with a distinct pause-design mechanism and verified exact GPT-4 context.

    13 Aug 2026