“But there’s another side to Altman, too, the part that was recently forced to acknowledge that ChatGPT, for all its fame, really isn’t anything like AGI yet:”
Our reading, the sentence that was checked:
ChatGPT, despite its fame, is not anything like AGI yet.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Is Microsoft about to get the deal of the century? Or is Sam Altman unloading OpenAI at just the right time?, paragraph 20.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-is-microsoft-about-to-get-the-deal-0c09b1-a3
We recorded 6 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 2023-01-11. 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
- ChatGPT, for all its fame
- really isn’t anything like AGI yet
- recently forced to acknowledge
Our interpretation
- The ordinary capability claim is that contemporary ChatGPT lacked the broadly dependable, flexible competence associated with human-level general intelligence.
- Anything like is emphatic language, not a literal assertion that ChatGPT shared no capabilities with general intelligence.
- The context anchors the observation to Altman's December 11, 2022 assessment through the January 11, 2023 post. This is the contemporary product period, not a claim about later ChatGPT versions.
- Forced to acknowledge is rhetorical framing. The normalized proposition does not require proving coercion or Altman's private beliefs.
Counts as supported if: Evidence about the contemporary ChatGPT system demonstrates substantial failures of reasoning, robustness, and factual reliability across tasks inconsistent with broadly dependable general intelligence.
Counts as contradicted if: Evidence about that same system demonstrates broadly dependable human-level general competence across varied tasks, sufficient to overcome the asserted limitations.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
Outcome: Supported
The dated evidence matches what the claim committed to, under the criteria frozen before grading.
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
The evaluation identifies the tested system as the December 15, 2022 ChatGPT version [e3]. It reports 63.41% average accuracy across ten reasoning categories and hallucination problems, supporting the ordinary reading that this historical version lacked broadly dependable general intelligence [e2]. It also reports performance exceeding fully fine-tuned task-specific models on four tasks; these strengths do not establish broadly dependable general competence [e3].
Limitations
- The evidence supplies neither an agreed AGI threshold nor a comprehensive comparison with human competence.
- The packet flags inconsistent version-year references elsewhere in the paper, which limit confidence in its historical attribution.
- The findings concern the identified historical version and do not establish later ChatGPT capabilities or a timeline to AGI.
- The evidence does not establish that Altman was forced to acknowledge limitations.
Evidence (3 items)
Supports the claim: arXiv, Yejin Bang and colleagues,
hallucination problems
Why it matters here: The paper identifies the December 15, 2022 ChatGPT version and reports 63.41% average accuracy across ten reasoning categories, inconsistent reasoning, and hallucinations. These results support the ordinary capability reading; the event date identifies the historical version rather than the later publication.
- Publisher
- arXiv, Yejin Bang and colleagues
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Context only: OpenAI,
incorrect or nonsensical answers.
Why it matters here: OpenAI acknowledged factual unreliability at launch. This disclosure precedes the target period and is contextual rather than decisive.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Context only: arXiv, Yejin Bang and colleagues,
4 different tasks.
Why it matters here: The same evaluation documents meaningful breadth and strong performance on some tasks. This limits a literal reading of anything like AGI, but does not demonstrate broadly dependable general intelligence.
- Publisher
- arXiv, Yejin Bang and colleagues
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Is Microsoft about to get the deal of the century? Or is Sam Altman unloading OpenAI at just the right time?
But there’s another side to Altman, too, the part that was recently forced to acknowledge that ChatGPT, for all its fame, really isn’t anything like AGI yet:
- Where
- paragraph 20
- Published
- 2023-01-11
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-is-microsoft-about-to-get-the-deal-0c09b1-a3: Current. Outcome: Supported. 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-is-microsoft-about-to-get-the-deal-0c09b1-a3.
HeldBy this site
The final-text check failed: citations_correct.
About assessment marcus-is-microsoft-about-to-get-the-deal-0c09b1-a2, which was not published.