“The trouble is that so far RAG has been hit and miss.”
Our reading, the sentence that was checked:
So far retrieval-augmented generation has produced inconsistent results.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, No, RAG is probably not going to rescue the current situation, paragraph 7.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-no-rag-is-probably-not-going-to-rescue-38ee57-a1
We recorded 11 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 2024-02-23. 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)
From the speaker
- so far
- RAG
- hit and miss
Our interpretation
- The plain reading is that retrieval-augmented systems had produced a mixture of successful answers and factual or citation failures.
- The phrase does not specify equal numbers of successes and failures or claim that retrieval never improves performance.
- 'So far' describes the accumulated record through February 23, 2024. The target therefore has no asserted starting date.
Counts as supported if: Contemporary evaluations of systems that condition generated answers on retrieved material document both supported outputs and material answer or citation failures.
Counts as contradicted if: Representative contemporary evaluations establish consistently reliable factual answers and citations across the relevant uses, and the alleged failures are shown not to involve retrieval-augmented generation or not to be failures.
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
A primary evaluation of four commercial systems using retrieved content found both supported outputs and substantial citation failures. That directly supports the ordinary meaning of 'hit and miss' in the record available before the post. It does not imply that every unsupported sentence was false or that RAG offered no improvement.
Limitations
- Citation correctness is narrower than overall factual accuracy.
- Results concern particular systems, versions, and queries.
- No particular overall error rate is attributed to RAG as a whole.
Evidence (1 item)
Supports the claim: Liu, Zhang and Liang; Findings of EMNLP 2023, hosted on arXiv,
do not cite comprehensively or correctly.
Why it matters here: The study evaluated Bing Chat, NeevaAI, perplexity.ai, and YouChat using responses collected in February–March 2023; its methods identify their use of retrieved content. Average sentence citation support was 51.5% and citation precision was 74.5%, documenting supported outputs alongside citation failures.
- Publisher
- Liu, Zhang and Liang; Findings of EMNLP 2023, hosted on arXiv
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, No, RAG is probably not going to rescue the current situation
The trouble is that so far RAG has been hit and miss.
- Where
- paragraph 7
- Published
- 2024-02-23
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-no-rag-is-probably-not-going-to-rescue-38ee57-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-no-rag-is-probably-not-going-to-rescue-38ee57-a1.