“OpenAI conveniently forgot to include this comparison (ARC-AGI-2) in their livestream recital of benchmark progress, which left the livestream looking like marketing rather than science.”
Our reading, the sentence that was checked:
OpenAI left the ARC-AGI-2 comparison out of its GPT-5 livestream benchmarks, which made the livestream look like marketing rather than science.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, GPT-5 hot take, paragraph 6.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-gpt-5-hot-take-5eb0a3-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
- forgot to include this comparison (ARC-AGI-2)
- in their livestream recital of benchmark progress
- looking like marketing rather than science
Our interpretation
- Assess the factual allegation that the August 7 launch presentation omitted the ARC-AGI-2 comparison.
- 'Conveniently forgot' does not independently establish intent or an actual lapse of memory.
- 'Looking like marketing rather than science' is the speaker's characterization, not a separately measurable outcome.
Counts as supported if: A complete presentation record or reliable direct reporting establishes that OpenAI omitted the ARC-AGI-2 comparison from the launch presentation.
Counts as contradicted if: The launch recording, slides, transcript, or reliable direct reporting shows that OpenAI presented the ARC-AGI-2 comparison.
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 factual omission is supported by direct launch-day reporting that OpenAI did not mention ARC Prize during the GPT-5 presentation. This assessment relies on that report because a complete primary transcript and slide record were not accessible. It does not endorse an inference about OpenAI's intent or score Marcus's characterization of the presentation as marketing.
Limitations
- The decisive source is contemporaneous specialist reporting rather than an independently inspected complete recording.
- The outcome applies to the factual omission only.
Evidence (1 item)
Supports the claim: The Decoder,
did not mention the ARC Prize
Why it matters here: The launch-day report discusses the ARC-AGI comparisons and expressly reports their omission from the presentation. It does not establish why OpenAI omitted them.
- Publisher
- The Decoder
- Kind of source
- News report
- 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
OpenAI conveniently forgot to include this comparison (ARC-AGI-2) in their livestream recital of benchmark progress, which left the livestream looking like marketing rather than science.
- Where
- paragraph 6
- Published
- 2025-08-07
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-gpt-5-hot-take-5eb0a3-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.