“So far, in most domains (not all, to be sure) Big Data has triumphed (for now) over careful knowledge engineering.”
Our reading, the sentence that was checked:
So far, in most but not all domains, Big Data approaches have beaten careful knowledge engineering.
Assessment: Unresolved Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, The New Science of Alt Intelligence, paragraph 55.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-the-new-science-of-alt-intelligence-99f3a2-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 2022-05-14. 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
- in most domains, not all (qualifier, stated by the speaker)
- for now (qualifier, stated by the speaker)
From the speaker
- So far
- in most domains
- not all
- Big Data has triumphed
- for now
- over careful knowledge engineering
Our interpretation
- The surrounding discussion concerns AI application domains, rather than all software or economic activity.
- I read triumphed as superior demonstrated task performance, and most as a majority of major AI domains.
- Big Data includes learning from large collected or generated datasets; knowledge engineering means substantial hand-designed domain knowledge.
- So far describes cumulative progress through publication. No historical start date is supplied.
- Not all and for now restrict scope and time; they are not forecast antecedents.
Counts as supported if: A broad, defensible comparison across major AI domains shows that data-driven methods outperform knowledge-engineered alternatives in a majority, while allowing exceptions.
Counts as contradicted if: A comparably broad domain comparison shows that data-driven methods do not lead in a majority. A few exceptions cannot contradict a claim that explicitly allows them.
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 record contains concrete successes for data-driven learning, including AlphaZero's results against Stockfish. Those examples fit Marcus's general direction but do not establish most domains. A majority verdict would depend on an evaluator-selected domain list and classification of hybrid systems. The available evidence therefore does not settle the claim.
Limitations
- This is not a finding that knowledge engineering generally performed better.
- The claim expressly permits exceptions.
- Benchmark progress, deployment prevalence, and commercial success are different measures.
Evidence (2 items)
Context only: arXiv; DeepMind researchers,
defeated Stockfish
Why it matters here: This is a concrete success for learning from self-play against a program using extensive hand-engineered domain knowledge. It establishes a selected-domain example, not the claimed majority.
- Publisher
- arXiv; DeepMind researchers
- Kind of source
- Preprint, not peer reviewed
- Describes events of
- Public from
Context only: arXiv; AI Index authors,
data related to artificial intelligence
Why it matters here: This contemporary survey is relevant to checking breadth across AI. Its readable abstract does not enumerate domains won by data-driven systems against knowledge-engineered alternatives, so it cannot settle the majority assertion.
- Where
- Abstract
- Publisher
- arXiv; AI Index authors
- Kind of source
- Secondary summary
- Describes events of
- Public from
What would settle it
A justified cross-domain inventory of results through May 14, 2022, with consistent domain boundaries, identified knowledge-engineered comparators, and explicit treatment of hybrid methods.
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, The New Science of Alt Intelligence
So far, in most domains (not all, to be sure) Big Data has triumphed (for now) over careful knowledge engineering.
- Where
- paragraph 55
- Published
- 2022-05-14
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-the-new-science-of-alt-intelligence-99f3a2-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-the-new-science-of-alt-intelligence-99f3a2-a1.