“And LLMs themselves have already resorted to using lots of external tools like calculators and python interpreters; it is already clear that they are more of a partial solution than the full solution people imagine.”
Our reading, the sentence that was checked:
LLMs have already resorted to using lots of external tools such as calculators and Python interpreters, and it is already clear that they are more of a partial solution than the full solution people imagine.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Three years on, ChatGPT still isn't what it was cracked up to be – and it probably never will be, paragraph 38.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-three-years-on-chatgpt-still-isnt-90ddbb-a1
We recorded 30 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 2025-11-29. 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
- already resorted to using lots of external tools
- calculators and python interpreters
- more of a partial solution than the full solution people imagine
Our interpretation
- LLMs using tools refers to deployed systems in which a language model requests external computation or services.
- Calculators and Python interpreters are examples; the claim does not require a separate calculator product in every system.
- Partial solution is read as a present architectural claim: the language model supplies part of a broader system's capabilities.
- Tool adoption alone cannot establish that tools are permanently necessary or that a standalone model could never attain general intelligence.
- Already makes this a cumulative observation within the post's three-year retrospective.
Counts as supported if: Dated deployment documentation establishes that LLM systems integrate multiple external capabilities, including executable computation, with the model acting as a component of the resulting system.
Counts as contradicted if: The cited capabilities were generated entirely by the language model without external execution or services, or documentation establishes the claimed complete standalone capability under a defined task scope.
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
Provider announcements establish that LLM systems used external Python execution, web search, and other services before the post. The model was therefore one component of a broader working system. This supports the present architectural reading of partial solution. It does not establish the stronger proposition that standalone models can never become sufficient.
Limitations
- Tool adoption does not by itself prove necessity, failure, or a permanent capability ceiling.
- The conclusion about a partial solution depends on reading it as a claim about deployed systems.
Evidence (2 items)
Supports the claim: OpenAI,
working Python interpreter
Why it matters here: ChatGPT's interpreter performed computation in an external execution environment. This directly supports tool integration and distinguishes executing Python from merely generating Python text.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Supports the claim: Anthropic,
web search and citations
Why it matters here: The announcement makes code execution, external server connections, file handling, and other capabilities available around Claude. It corroborates the use of multiple external components across providers.
- Publisher
- Anthropic
- 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, Three years on, ChatGPT still isn't what it was cracked up to be – and it probably never will be
And LLMs themselves have already resorted to using lots of external tools like calculators and python interpreters; it is already clear that they are more of a partial solution than the full solution people imagine.
- Where
- paragraph 38
- Published
- 2025-11-29
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-three-years-on-chatgpt-still-isnt-90ddbb-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-three-years-on-chatgpt-still-isnt-90ddbb-a1.