“Long-term adoption (vs mere experimentation) has been modest, there is no significantly better tech now than there was in January; GPT-5 is still a no show.”
Our reading, the sentence that was checked:
Six months after his January 2024 warnings, long-term adoption (as opposed to mere experimentation) has been modest.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Why OpenAI may well be completely Zuck’d, paragraph 3.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-why-openai-may-well-be-completely-234e80-a1
We recorded 10 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 2024-07-24. 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
- long-term adoption vs mere experimentation (qualifier, stated by the speaker)
- in the context of whether adoption justifies OpenAI's valuation (scope, stated by the speaker)
From the speaker
- Long-term adoption
- vs mere experimentation
- has been modest
- Six months later
Our interpretation
- Assess the adoption clause separately from the accompanying technology and GPT-5 clauses.
- The context supplies January–July 24, 2024 as the observation interval.
- The ordinary reading is that sustained consumer use and established business use remained limited compared with experimentation.
- Regular-use surveys are relevant proxies, but do not automatically establish customer retention or long-running production deployments.
- No numerical threshold for modest, minimum retention duration, or valuation model is added.
Counts as supported if: Representative evidence shows that sustained consumer and production use remained limited, with most trial users or projects not becoming lasting adopters.
Counts as contradicted if: Representative retention or established-production evidence shows widespread lasting adoption, rather than predominantly trials and occasional use.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
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 horizon
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
The consumer survey supports limited routine use, while the business survey reports substantial regular use. Neither directly measures the lasting adoption Marcus distinguishes from experimentation. Their different populations and measures prevent a clean combined verdict. The available evidence therefore does not settle the broad long-term-adoption claim.
Limitations
- Regular use is not the same as demonstrated retention.
- The surveys precede the July post and do not cover all users or businesses.
- The words modest and long-term leave meaningful interpretive uncertainty.
Evidence (2 items)
Context only: Reuters Institute for the Study of Journalism,
a snapshot
Why it matters here: The March–April survey found low daily ChatGPT use, ranging from 1% in Japan to 7% in the United States, and considerable one-off experimentation. It supports limited routine use in those populations, but measures a snapshot rather than longitudinal retention.
- Where
- Introduction, survey limitations paragraph; frequency findings in Executive Summary and section 1
- Publisher
- Reuters Institute for the Study of Journalism
- Kind of source
- Independent measurement
- Describes events of
- Public from
Context only: McKinsey & Company; PDF hosted by ABBYY,
using gen AI.
Why it matters here: The February 22–March 5 survey reported that 65% of respondents’ organizations regularly used generative AI in at least one function. This challenges a blanket low-use account, but does not establish how long deployments lasted or represent every business.
- Publisher
- McKinsey & Company; PDF hosted by ABBYY
- Kind of source
- Independent measurement
- Describes events of
- Public from
What would settle it
Representative cohorts or production records showing retention, deployment duration, and conversion from trials during the January–July 2024 period.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Why OpenAI may well be completely Zuck’d
Long-term adoption (vs mere experimentation) has been modest, there is no significantly better tech now than there was in January; GPT-5 is still a no show.
- Where
- paragraph 3
- Published
- 2024-07-24
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-why-openai-may-well-be-completely-234e80-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-why-openai-may-well-be-completely-234e80-a1.