“Although GPT-4 will definitely seem smarter than its predecessors, its internal architecture remains problematic.”
Our reading, the sentence that was checked:
GPT-4 will seem smarter than predecessors but its architecture remains problematic.
Assessment: Unresolved No deadline stated on 2026-10-29 Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, What to Expect When You’re Expecting … GPT-4, paragraph 34.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-what-to-expect-when-youre-expecting-629194-a1
We recorded 21 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 2022-12-25. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Forecast
- Deadline, as the speaker wrote it
- None stated. This site does not add one.
- Deadline state when assessed
- No deadline stated
- Deadline state on 2026-10-29
- No deadline stated (the release’s reference date)
From the speaker
- GPT-4 will definitely seem smarter than its predecessors
- its internal architecture remains problematic
Our interpretation
- The target is the original GPT-4 generation, compared with GPT-3 and GPT-3.5.
- The plain reading predicts visibly improved capabilities alongside continuing limitations in the model's design.
- Comparative evaluations can establish improved capabilities. Observable errors establish limitations, but do not by themselves establish the architectural diagnosis.
- The surrounding discussion anticipates release and subsequent inspection, but supplies no binding deadline.
Counts as supported if: Comparative evaluations establish clear capability gains, and credible technical evidence connects substantial remaining limitations to GPT-4's design.
Counts as contradicted if: Credible comparative evaluations establish no meaningful capability improvement, or technical evidence refutes the alleged continuing design limitations.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Did the forecast come true?
Outcome: Unresolved
The record was checked, and the evidence available by the cutoff does not settle it.
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 deadline stated
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
The capability prediction is supported: human evaluators preferred GPT-4 to GPT-3.5. Continuing errors are also documented. However, those observations do not settle the claim about internal architecture. The report withholds the details needed to assess that diagnosis. This is incomplete support for a compound claim, not evidence that one part is false.
Limitations
- The comparative evaluation was conducted by the model's developer.
- The result depends on preserving the distinction between observable limitations and an architectural explanation.
- The cited version was published in 2024 and reports the original model's development and evaluations.
Evidence (1 item)
Supports the claim: OpenAI,
follow user intent
Why it matters here: The report describes human preference for GPT-4 over GPT-3.5 on 70.2% of 5,214 prompts, supporting the capability-improvement component. It also acknowledges reasoning errors and hallucinations while withholding detailed architecture, so it does not settle the architectural diagnosis.
- Where
- Section 4, human preference evaluation; sections 2 and 5 provide disclosure and limitations context
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
What would settle it
Public architectural evidence and controlled studies connecting the documented limitations to the original GPT-4's design, or evidence refuting that connection.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, What to Expect When You’re Expecting … GPT-4
Although GPT-4 will definitely seem smarter than its predecessors, its internal architecture remains problematic.
- Where
- paragraph 34
- Published
- 2022-12-25
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-what-to-expect-when-youre-expecting-629194-a1: Current. Outcome: Unresolved. 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-what-to-expect-when-youre-expecting-629194-a1.