“That is not going to magically change next year, at a time when AI still can’t reliably solving river-crossing problems.”
Our reading, the sentence that was checked:
In 2026, AI will still not be choosing its own research goals in place of human scientists.
Assessment: Unresolved Not yet due on 2026-10-29 Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Nobel Prizes and The AI Hype Hall of Fame, paragraph 16.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-nobel-prizes-and-the-ai-hype-hall-612ee6-a1
We recorded 7 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 2025-03-09. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Forecast
- Deadline, as the speaker wrote it
- “next year”
- Deadline state when assessed
- Not yet due
- Deadline state on 2026-10-29
- Not yet due (the release’s reference date)
From the speaker
- humans scientists are still choosing the research goals
- That is not going to magically change next year
- conceive of and execute Nobel-caliber work on their own
Our interpretation
- Next year means calendar year 2026, ending December 31.
- The assessed proposition concerns AI taking over the choice of scientific research goals from human scientists. The river-crossing remark is supporting rhetoric, not a condition.
- The ordinary contextual reading concerns substantive research direction, rather than merely generating hypotheses or selecting experiments within an assigned objective.
- Repeated, documented independent goal selection would establish the change. Marcus supplies no numerical threshold or adoption requirement; these criteria do not require replacement of all scientists or an actual Nobel Prize.
Counts as supported if: Evidence covering 2026 shows that leading AI-for-science systems continue to depend on human scientists to choose substantive research objectives, despite automating hypotheses, experiments, or other research steps.
Counts as contradicted if: Documented research records show AI repeatedly choosing and pursuing substantive scientific objectives in place of human scientists, with enough detail to distinguish independent research direction from work within a human-assigned objective.
Prospective: these criteria were frozen before the deadline, while the outcome was not yet 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
- Not yet due
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
The deadline is December 31, 2026, after the evaluation date. Co-Scientist's published description retains scientist-provided objectives. Other records show automated idea generation, AI-generated project topics, and autonomous execution. Those are meaningful counterevidence to a broad claim that AI cannot choose research topics. They do not conclusively establish the contextual claim's stronger change: AI taking over substantive research direction from human scientists. No verified record reviewed meets the early-resolution rule. Continued human direction before the deadline also cannot establish support.
Limitations
- The distinction between research topics and research goals materially affects the assessment.
- The evidence covers selected systems, not every AI research deployment.
- The August results are a preprint, and the October system description could not be verified.
Evidence (3 items)
Context only: Nature,
Conditioned on their research objectives
Why it matters here: The published system description retains scientists' objectives while automating hypothesis generation. This is consistent with Marcus's forecast for this system, but cannot establish the field's position at year-end.
- Publisher
- Nature
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
Context only: Nature,
produce diverse ideas and automatically test
Why it matters here: The paper reports idea generation and testing in both constrained and open-ended settings. This challenges a blanket reading that AI cannot originate research ideas, but the accessible record does not establish repeated replacement of human research-direction choices.
- Publisher
- Nature
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
Context only: arXiv,
only the research topic as a natural-language directive.
Why it matters here: The methods describe 50 topics generated using Gemini under a human-defined AI-research brief, followed by autonomous project execution. This is material evidence of AI topic selection, although treating it as replacement of substantive human research direction depends on the unresolved meaning of goals.
- Publisher
- arXiv
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
What would settle it
Research records through December 31, 2026 showing who selected substantive objectives, with initial prompts and intervention logs. Repeated independent AI selection and pursuit would contradict the forecast; broad evidence of continued human goal selection through year-end would support it.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Nobel Prizes and The AI Hype Hall of Fame
That is not going to magically change next year, at a time when AI still can’t reliably solving river-crossing problems.
- Where
- paragraph 16
- Published
- 2025-03-09
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-nobel-prizes-and-the-ai-hype-hall-612ee6-a1: 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 marcus-nobel-prizes-and-the-ai-hype-hall-612ee6-a1.