DOOMBENCH.COM
Hourly watch
Menu
Capability gains

Altman predicts large AI models will generate new scientific knowledge

In a full Greylock interview, Altman predicted that large models would become a general computing platform, use computers through agents, and progress from summarizing knowledge to producing new scientific discoveries.

0 comments · 0 votesOpen discussion

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

No comments yet. Start the discussion.

Send detailed or private feedback →
CURRENT ASSESSMENT · REVISION 1
TOWARD DOOM30confidence 55/100

Why it moved the index

The attributable forecast identifies agentic computer use and net-new science as capability mechanisms. It is not evidence that those forecast capabilities had already been achieved in 2022.

AUDIT TRAIL

Assessment history

  1. R1
    Toward 30 · confidence 55

    Adds a dated 2022 capability forecast distinct from routine product announcements.

    14 Aug 2026
SHARE THE FINDINGS

Share this page

DoomBench social sharing card for Altman predicts large AI models will generate new scientific knowledge.

  1. DoomBench assesses “Altman predicts large AI models will generate new scientific knowledge” as evidence moving toward doom, with magnitude 30 and confidence 55 out of 100 in the capability gains category.

  2. The DoomBench assessment of “Altman predicts large AI models will generate new scientific knowledge” is based on reporting from Greylock and records the editorial rationale, source quality, attribution, and revision history.

  3. DoomBench summarizes “Altman predicts large AI models will generate new scientific knowledge” as follows: In a full Greylock interview, Altman predicted that large models would become a general computing platform, use computers through...

    https://www.doombench.com/news/altman-predicts-large-ai-models-will-generate-new-scientific-knowledge-2022-09-13