“Transformers and generative AI are getting a lot of attention, and have considerable utility, but they are nothing like the all purpose tools some people seems to imagine them to be.”
Our reading, the sentence that was checked:
Transformers and generative AI are getting a lot of attention and have considerable utility, but they are nothing like the all-purpose tools some people seem to imagine them to be.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Reports of the birth of AGI are greatly exaggerated, paragraph 22.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-reports-of-the-birth-of-agi-are-greatly-355e77-a1
We recorded 11 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 2023-10-18. 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)
Conditions
- the all-purpose tools some people seem to imagine (qualifier, stated by the speaker)
From the speaker
- Transformers and generative AI
- getting a lot of attention
- have considerable utility
- nothing like the all purpose tools some people seems to imagine
Our interpretation
- The ordinary reading is that these technologies were prominent and useful but could not reliably perform the broad range of practical tasks expected of an all-purpose AI system.
- I assess the current calendar month through publication, October 1–18, 2023. This is an evaluator-selected reporting window for the ongoing state described.
- This does not require proving that transformers were unused in the fields mentioned nearby. Their use in several fields would not itself establish all-purpose reliability.
- The reference to what some people imagine is a comparison, not a conditional antecedent.
Counts as supported if: Contemporaneous records establish substantial attention and practical uses, while evaluations demonstrate major limitations across distinct kinds of tasks.
Counts as contradicted if: Contemporaneous evaluations establish broadly reliable performance across the practical tasks at issue, or strong evidence refutes the claimed prominence or meaningful utility.
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 ordinary reading is supported. Microsoft's disclosure establishes substantial engagement and useful image-generation applications. October evaluations show major limitations in repository-level software work and current factual answering. Transformer-based robotics demonstrates real breadth, but does not establish all-purpose reliability. This finding does not endorse the separate surrounding assertion about how little transformers were used across AI fields.
Limitations
- The result depends on interpreting all-purpose as broadly reliable practical competence.
- Usage and application examples are company-reported.
- Benchmark findings describe particular models and setups; their publication dates identify the contemporary assessments, not exact experimental execution dates.
- Search augmentation and human assistance can improve performance.
Evidence (4 items)
Supports the claim: Microsoft Bing,
over 1 billion images
Why it matters here: The cumulative usage disclosed on this date demonstrates substantial engagement. The same paragraph documents uses including illustrated stories, social-media thumbnails and design inspiration, supporting practical utility without establishing universal reliability.
- Publisher
- Microsoft Bing
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Supports the claim: arXiv; Carlos E. Jimenez and colleagues,
$4.8$% and $1.7$%
Why it matters here: The October assessment reports severe limitations on real repository issue resolution even when relevant material is supplied. It supports the lack of all-purpose competence, while not measuring the usefulness of assisted coding.
- Publisher
- arXiv; Carlos E. Jimenez and colleagues
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Supports the claim: arXiv; Tu Vu and colleagues,
dramatically degraded accuracies
Why it matters here: The October assessment identifies a second practical limitation: factual answers requiring current information. These baseline tests exclude browsing; the paper also reports improvements from search augmentation.
- Publisher
- arXiv; Tu Vu and colleagues
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Context only: arXiv; Open X-Embodiment collaboration,
Transformer-based robotic policies
Why it matters here: The paper reports successful transfer across robots using transformer-based policies. This is material evidence of breadth and qualifies any reading that confines transformers to language, but it does not demonstrate an all-purpose system.
- Publisher
- arXiv; Open X-Embodiment collaboration
- 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, Reports of the birth of AGI are greatly exaggerated
Transformers and generative AI are getting a lot of attention, and have considerable utility, but they are nothing like the all purpose tools some people seems to imagine them to be.
- Where
- paragraph 22
- Published
- 2023-10-18
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-reports-of-the-birth-of-agi-are-greatly-355e77-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.
About assessment marcus-reports-of-the-birth-of-agi-are-greatly-355e77-a1.