“GPT-5 is obviously not AGI.”
Our reading, the sentence that was checked:
GPT-5 is not AGI.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, GPT-5 hot take, paragraph 9.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-gpt-5-hot-take-5d0a8a-a1
We recorded 8 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 2025-08-07. 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
- GPT-5
- obviously not AGI
Our interpretation
- The claim concerns the GPT-5 system announced on August 7, 2025.
- The ordinary reading of AGI is broadly human-level general cognitive capability, including reliable adaptation and learning across unfamiliar tasks.
- Strong performance on selected benchmarks or expert-level answers in some domains does not by itself establish AGI.
- Obviously expresses confidence; it adds no separate measurable threshold.
Counts as supported if: Evidence identifies substantial missing general capabilities, supported by evaluations or a concrete acknowledgment from the developer that the launched system falls short of AGI.
Counts as contradicted if: Broad, credible evaluations establish general human-level capability and adaptation across unfamiliar intellectual tasks, rather than isolated benchmark successes.
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
At launch, Altman acknowledged missing capabilities needed for AGI, with continuous learning given as a concrete example. OpenAI also documented substantial performance across occupations, but those results did not establish the broad general capability in the stated reading. The claim is supported under that reading, rather than proved under every possible definition of AGI.
Limitations
- The outcome depends on the stated interpretation of AGI.
- The decisive acknowledgment is reported by WIRED.
- Neither failing to learn continuously nor any single benchmark is a definition-independent AGI test.
Evidence (2 items)
Supports the claim: WIRED,
lacks the ability to learn continuously after deployment.
Why it matters here: WIRED reports that Altman identified missing traits preventing GPT-5 from reaching AGI, including continuous learning. This supports the claim under the stated ordinary reading; it is attributed reporting rather than a primary transcript.
- Publisher
- WIRED
- Kind of source
- News report
- Describes events of
- Public from
Context only: OpenAI,
our best performing model on an internal benchmark
Why it matters here: OpenAI reports expert-comparable or better results in roughly half the cases across more than 40 occupations. This is material evidence of breadth, but does not establish general human-level intelligence or continuous learning.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, GPT-5 hot take
GPT-5 is obviously not AGI.
- Where
- paragraph 9
- Published
- 2025-08-07
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-gpt-5-hot-take-5d0a8a-a1: 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.