“The first one collapsed; this one has more staying power.”
Our reading, the sentence that was checked:
The current revival of neural networks has more staying power than the earlier one, which collapsed.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Happy Groundhog Day, The AI Edition, paragraph 2.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-happy-groundhog-day-the-ai-edition-d1ccc9-a1
We recorded 12 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 2023-02-03. 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
- compared with the earlier neural network rebirth (qualifier, stated by the speaker)
From the speaker
- 30 years ago
- in the midst of a rebirth
- The first one collapsed
- this one has more staying power
Our interpretation
- The comparison concerns the connectionist revival of the 1980s and early 1990s and the deep-learning revival observable by February 3, 2023.
- The ordinary reading of staying power is sustained research activity and established practical adoption, rather than a specified number of years.
- Collapsed means a substantial retreat in prominence and enthusiasm, not the disappearance of neural-network research.
- The target period spans both waves because this is a historical comparison. Survival through 2026 is not an additional requirement.
Counts as supported if: Historical evidence establishes a substantial retreat of the earlier revival, while evidence available by the post establishes sustained research and substantially broader practical adoption during the current revival.
Counts as contradicted if: Comparable historical records show no substantial retreat of the earlier revival, or show that the current revival had already suffered a comparable retreat without stronger practical foundations.
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 earlier revival substantially lost prominence, while the current revival had become embedded in widely used products and remained commercially active immediately before the post. That supports the ordinary historical comparison. The earlier wave did not disappear: important research and applications continued. The result therefore depends on reading collapsed as a retreat in prominence.
Limitations
- This does not establish that the current revival could never collapse.
- The comparison is qualitative, and the historical account expressly documents continuity during the earlier retreat.
Evidence (3 items)
Supports the claim: MIT Press; Ian Goodfellow, Yoshua Bengio and Aaron Courville,
until 2007.
Why it matters here: The account dates the retreat from the mid-1990s through 2007. It also records continued technical successes and institutional support, so it supports declining prominence rather than literal extinction.
- Publisher
- MIT Press; Ian Goodfellow, Yoshua Bengio and Aaron Courville
- Kind of source
- Secondary summary
- Describes events of
- Public from
Supports the claim: Association for Computing Machinery,
billions of people.
Why it matters here: ACM describes deep learning as embedded in widely used computing applications. This supports practical entrenchment, although the adoption statement is not an independently audited census.
- Publisher
- Association for Computing Machinery
- Kind of source
- Secondary summary
- Describes events of
- Public from
Supports the claim: Microsoft,
one million people
Why it matters here: Microsoft reported substantial actual use of its AI coding product immediately before the post. Together with the earlier adoption record, this supports continuing commercial relevance.
- Publisher
- Microsoft
- 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, Happy Groundhog Day, The AI Edition
The first one collapsed; this one has more staying power.
- Where
- paragraph 2
- Published
- 2023-02-03
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-happy-groundhog-day-the-ai-edition-d1ccc9-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-happy-groundhog-day-the-ai-edition-d1ccc9-a1.