“We can be absolutely sure that the task length “time horizon” for AI is not going to keep doubling until “time horizons will be 580 times the age of the universe” as Lisan al-Ghaib joked.”
Our reading, the sentence that was checked:
We can be absolutely sure that the AI task-length 'time horizon' will not keep doubling until time horizons reach 580 times the age of the universe.
Assessment: Unresolved No deadline stated on 2026-10-29 Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, Misplaced panic over AI progress, paragraph 37.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-misplaced-panic-over-ai-progress-c1af8f-a1
We recorded 14 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 2026-05-10. 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)
Conditions
- absolutely sure (qualifier, stated by the speaker)
- until time horizons are 580 times the age of the universe (scope, stated by the speaker)
From the speaker
- task length “time horizon”
- not going to keep doubling
- 580 times the age of the universe
- absolutely sure
Our interpretation
- The target is the METR-style task-completion horizon discussed in the post, measured in human task-completion time.
- The astronomical quantity is a capability threshold, not a calendar deadline.
- Absolutely sure expresses confidence, not an antecedent condition.
- The prediction is that the doubling process breaks before reaching the stated threshold. A temporary noisy estimate or exhausted benchmark does not establish that break.
Counts as supported if: A decisive capability ceiling or lasting structural break prevents the comparable task-horizon doubling process from reaching 580 times the age of the universe.
Counts as contradicted if: Validated, comparable successive task-horizon doublings reach at least 580 times the age of the universe without the predicted break.
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 claim gives a capability threshold but no deadline. The searched records establish neither continued doublings to that threshold nor a decisive ceiling below it. Measurement saturation cannot be treated as evidence that capability growth has ended. The emphatic wording does not make the forecast due by October 28.
Limitations
- The threshold lies far beyond the capabilities measured in the cited report.
- No arbitrary calendar deadline or number of missed doublings has been added.
Evidence (1 item)
Context only: METR,
over two full-time-equivalent days
Why it matters here: The report covers the assessment window ending in March and describes uncertainty caused by benchmark saturation. That is a measurement limitation, not the lasting capability break required to support this forecast.
- Publisher
- METR
- Kind of source
- Independent measurement
- Describes events of
- Public from
What would settle it
Validated continued doublings reaching the stated threshold, or a decisive dated finding that the relevant capability process cannot reach it.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, Misplaced panic over AI progress
We can be absolutely sure that the task length “time horizon” for AI is not going to keep doubling until “time horizons will be 580 times the age of the universe” as Lisan al-Ghaib joked.
- Where
- paragraph 37
- Published
- 2026-05-10
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-misplaced-panic-over-ai-progress-c1af8f-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-misplaced-panic-over-ai-progress-c1af8f-a1.