“I think that in achieving AGI we will have no choice but to find ways of integrating whatever we can get out of deep learning with explicit knowledge and explicit cognitive models.”
Our reading, the sentence that was checked:
Achieving AGI will require integrating deep learning with explicit knowledge and explicit cognitive models.
Assessment: Unresolved No deadline stated on 2026-10-29 Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, AGI will not happen in your lifetime. Or will it?, paragraph 74.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-agi-will-not-happen-in-your-lifetime-5e6cf0-a1
We recorded 28 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 2023-01-22. 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
- in achieving AGI
- we will have no choice
- integrating whatever we can get out of deep learning
- with explicit knowledge and explicit cognitive models
Our interpretation
- This predicts architectural necessity, rather than merely predicting that hybrid systems will be useful.
- I use the surrounding description of human-like integration across essentially arbitrary domains and modalities as the ordinary capability standard.
- Explicit knowledge means identifiable knowledge representations used by the system, rather than information encoded only implicitly in learned weights.
- Explicit cognitive models means identifiable representations or mechanisms modeling cognitive operations. A system's fluent explanations alone do not demonstrate their presence.
- The sentence states no deadline. The surrounding discussion of 75 years does not give this architectural claim a deadline.
- AGI is the outcome being discussed, not a separately stated antecedent that must be marked unmet.
Counts as supported if: An independently validated AGI meeting the contextual capability standard demonstrably integrates deep learning, explicit knowledge, and explicit cognitive models, with strong evidence that these explicit components are necessary. Their presence in one successful design alone supports compatibility, not the full necessity claim.
Counts as contradicted if: An independently validated AGI meeting the contextual capability standard has sufficiently disclosed and verified architecture showing that it achieves those capabilities without the required integration.
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 forecast has no deadline, and the reviewed evidence supplies no decisive event. Later systems show substantial capabilities, but benchmark success does not establish the contextual AGI standard or reveal whether the required explicit components are present. Conversely, limitations in current systems do not prove that Marcus's proposed architecture is necessary. Even a successful hybrid AGI would demonstrate one working route, not by itself establish that there was no alternative.
Limitations
- The available architecture disclosures are insufficient for a definitive architectural test.
- The outside embodied evaluation is a preprint and covers only part of the capability standard.
- The conclusion is that the reviewed evidence does not settle the forecast, not that AGI has been proved absent.
Evidence (3 items)
Context only: OpenAI,
a 99.9% score
Why it matters here: This is material evidence of advanced capabilities and a candidate worth testing against Marcus's forecast. The reported benchmarks do not establish his broad AGI standard or whether explicit knowledge and cognitive models are required.
- Publisher
- OpenAI
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Context only: Google DeepMind,
more than a framework
Why it matters here: The research announcement proposes broad cognitive evaluations and identifies measurement gaps. It does not demonstrate that explicit cognitive models must be components of an AGI; an evaluation framework is different from a system architecture.
- Publisher
- Google DeepMind
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
Context only: arXiv; SenseNova and Joint Research Team,
embodied intelligence.
Why it matters here: The outside evaluation reports strong capabilities alongside limitations in deployment and generalization. Neither its successes nor its failures establish the universal architectural requirement in Marcus's forecast.
- Publisher
- arXiv; SenseNova and Joint Research Team
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
What would settle it
An independently validated system meeting the contextual AGI standard, with verified architecture showing it works without the claimed integration, would contradict the forecast. Supporting necessity would require substantially stronger evidence than a single hybrid success, including convincing tests or theory establishing that the explicit components are indispensable.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, AGI will not happen in your lifetime. Or will it?
I think that in achieving AGI we will have no choice but to find ways of integrating whatever we can get out of deep learning with explicit knowledge and explicit cognitive models.
- Where
- paragraph 74
- Published
- 2023-01-22
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-agi-will-not-happen-in-your-lifetime-5e6cf0-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-agi-will-not-happen-in-your-lifetime-5e6cf0-a1.