“QUIP: AI Will Do Science Research, Somehow — No it won't, here's a writeup about why foundation models can't do this.”
Our reading, the sentence that was checked:
AI foundation models will not do scientific research.
Assessment: Unresolved No deadline stated on 2026-10-29 Evidence as of
Where and when Ed Zitron wrote it
Where's Your Ed At, How To Argue With An AI Booster, paragraph 383.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment wyea-how-to-argue-with-an-ai-booster-2883f4-a2
We recorded 35 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-28. The source may have been edited after 2025-08-25. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Forecast
- Deadline, as the speaker wrote it
- None stated. This site does not add one.
- Deadline state when assessed
- No deadline stated
- Deadline state on 2026-10-29
- No deadline stated (the release’s reference date)
From the speaker
- AI Will Do Science Research, Somehow — No it won't
- foundation models can't do this
- they're insufficient
Our interpretation
- The plain reading is that foundation models will not independently conduct substantive scientific investigations, rather than that they cannot assist scientists at all.
- A qualifying example must attribute the core hypothesis, investigation and analysis to a foundation-model system and independently validate the scientific result.
- Human provision of tools, broad objectives, ethical oversight or physical execution is compatible with this reading; humans supplying the substantive scientific solution is not.
- No deadline is stated. A finite period without a qualifying example cannot support the indefinite negative.
Counts as supported if: No finite sequence of failures establishes this unrestricted prediction. Support would require a demonstrated general barrier applying to the claimed class of systems.
Counts as contradicted if: A dated, independently validated scientific investigation establishes that a foundation-model system performed the substantive research under the stated autonomy criteria.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Did the forecast come true?
Outcome: Unresolved
The record was checked, and the evidence available by the cutoff does not settle it.
Scored on a plain reading. The speaker named no measure for this claim, so it was judged on the reading an informed reader would take, stated below under our interpretation. Both reviews had to find that reading fair.
- Evidence as of
- Deadline state when assessed
- No deadline stated
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
The Nature paper published on March 25, 2026 describes automated research and a generated manuscript passing an initial workshop review [e1]. An independent preprint submitted on November 17, 2025 reports one well-supported KOSMOS finding, one uncertain result and one false hypothesis [e2]. Neither inspected record establishes substantive model-led research, novelty and independent validation together under the contract's criteria [e1, e2]. They also establish no general barrier to foundation-model research, leaving the forecast unresolved [e1, e2].
Limitations
- Initial workshop review does not by itself establish independent validation of the scientific result.
- The KOSMOS abstract does not establish the complete research contribution or the extent of human intervention.
- The potentially relevant paper Closing the Empirical Loop could not be inspected.
- The claim gives no deadline; a finite period without a qualifying example cannot establish its indefinite negative.
Evidence (2 items)
Context only: Nature,
automatically test, report on
Why it matters here: This publication documents a foundation-model research workflow and a manuscript passing an initial workshop review. The date identifies publication of that assessment; it is not asserted to be the date of the underlying research or workshop review.
- Publisher
- Nature
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
Context only: arXiv; Humza Nusrat and Omar Nusrat,
three problems
Why it matters here: The independent assessment reports one supported discovery, one uncertain finding and one false hypothesis. It is material counterevidence, but the retrieved record does not establish the complete research contribution, novelty and external acceptance needed for decisive resolution.
- Where
- Abstract
- Publisher
- arXiv; Humza Nusrat and Omar Nusrat
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
What would settle it
A publicly documented investigation with methods, outputs and intervention records establishing that a foundation-model system supplied the substantive hypothesis, investigation and analysis, plus independent validation available by the evaluation date, would contradict the forecast. Support would require a demonstrated general barrier applying to the claimed class of systems.
Where it was said (1)
Said once in the sources this register covers.
: Where's Your Ed At, How To Argue With An AI Booster
QUIP: AI Will Do Science Research, Somehow — No it won't, here's a writeup about why foundation models can't do this.
- Where
- paragraph 383
- Published
- 2025-08-25
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment wyea-how-to-argue-with-an-ai-booster-2883f4-a2: Current. Outcome: Unresolved. Evidence as of 2026-10-28. Public date not recorded.
Recorded events
Assessment publishedBy this site
First assessment published after both reviews and the final-text check passed.
About assessment wyea-how-to-argue-with-an-ai-booster-2883f4-a2.