“But “breakthroughs” rarely turn to be general to live up to initial rosy expectations. Often advances work in some contexts, not otherwise.”
Our reading, the sentence that was checked:
Claimed AI breakthroughs rarely prove general enough to live up to initial expectations, often working only in some contexts.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, About that OpenAI “breakthrough”, paragraph 13.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-about-that-openai-breakthrough-009872-a1
We recorded 10 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-29. The source may have been edited after 2023-11-23. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Observation
- Deadline, as the speaker wrote it
- None stated. This site does not add one.
- Deadline state when assessed
- No horizon
- Deadline state on 2026-10-29
- No horizon (the release’s reference date)
From the speaker
- “breakthroughs” rarely turn to be general
- live up to initial rosy expectations
- Often advances work in some contexts, not otherwise.
Our interpretation
- The ordinary reading is a historical frequency claim about publicly touted AI advances before the post.
- Generality should be assessed against the initial expectations for each advance, not against an added requirement that every advance achieve general intelligence.
- The retrospective period ends on the post date. The source supplies no starting date, population of breakthroughs, or numerical meaning for rarely.
Counts as supported if: A reasonably representative historical comparison finds that meeting initially expected generality was uncommon and failures across contexts were common.
Counts as contradicted if: A comparable historical comparison finds that advances commonly met their initial expectations for generality.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
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 horizon
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
Research supports the existence of failures when systems move between contexts. That does not establish Marcus's broader frequency claim. Neither selected failures nor selected successes supply the missing comparison with initial expectations.
Limitations
- The result depends on which advances and which expectations are counted.
- No numerical threshold was invented for rarely.
Evidence (1 item)
Context only: arXiv; Alexander D'Amour and coauthors,
unexpectedly poor behavior
Why it matters here: The study documents deployment failures across several kinds of machine-learning applications. It does not count publicly touted breakthroughs or compare their outcomes with their original promises.
- Publisher
- arXiv; Alexander D'Amour and coauthors
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
What would settle it
A representative review of pre-post breakthrough announcements, their original expectations, and independent evaluations across the promised contexts.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, About that OpenAI “breakthrough”
But “breakthroughs” rarely turn to be general to live up to initial rosy expectations. Often advances work in some contexts, not otherwise.
- Where
- paragraph 13
- Published
- 2023-11-23
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-about-that-openai-breakthrough-009872-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-about-that-openai-breakthrough-009872-a1.