“Big data won’t be enough to solve the problems of robotics”
Our reading, the sentence that was checked:
More data alone will not be enough to solve the problems of robotics.
Assessment: Unresolved No deadline stated on 2026-10-29 Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, What’s wrong with Google’s new robot project, paragraph 36.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-whats-wrong-with-googles-new-robot-525fa1-a1
We recorded 20 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 2022-08-18. 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
- Big data
- won’t be enough
- solve the problems of robotics
Our interpretation
- Read the context as concerning reliable, broadly capable household robots that handle unfamiliar situations safely.
- Read 'enough' as sufficiency of increased training data, rather than a denial that data can improve performance.
- A decisive test must separate gains from more data from gains requiring additional methods.
- No deadline, numeric reliability threshold, or fixed definition of a data-only approach is supplied.
Counts as supported if: A decisive result establishes that increasing data alone cannot deliver the relevant household competence and reliability, under a clearly specified class of approaches.
Counts as contradicted if: Replicated evaluations demonstrate reliable, broad household competence in unfamiliar situations, with controlled evidence showing that scaling training data supplied the sufficient advance.
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. Research and company results show meaningful gains from broader data and newer training approaches, but the cited results do not establish that data scaling alone solves reliable general household robotics. Continuing limitations also cannot prove that it never will. No decisive event meeting the resolution rule was found by October 28, 2026.
Limitations
- The search is not an exhaustive audit of every robotic system.
- Company demonstrations do not provide independent confirmation of broad reliability.
- Any eventual ruling will depend on a defensible interpretation of 'data alone' and 'solve'.
Evidence (2 items)
Context only: arXiv; Fanqi Lin and colleagues,
data scaling in imitation learning.
Why it matters here: The study reports better transfer as training environments and objects become more diverse, including strong results on two tasks. Those bounded successes show the value of scaling but do not establish reliable general household robotics.
- Where
- Abstract
- Publisher
- arXiv; Fanqi Lin and colleagues
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Context only: Physical Intelligence,
a step-change in generalization.
Why it matters here: The company reports broader task and robot transfer, making this material evidence against assuming little progress. Its account also describes diverse prompting and conditioning methods; it does not isolate data quantity or establish the required general household reliability.
- Where
- Opening paragraph
- Publisher
- Physical Intelligence
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- Describes events of
- Public from
What would settle it
Replicated, sustained household evaluations with controlled evidence isolating data scaling as sufficient, or a rigorous applicable limit showing that data scaling alone cannot meet the defined requirements.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, What’s wrong with Google’s new robot project
Big data won’t be enough to solve the problems of robotics
- Where
- paragraph 36
- Published
- 2022-08-18
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-whats-wrong-with-googles-new-robot-525fa1-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-whats-wrong-with-googles-new-robot-525fa1-a1.