Scores by kind of claim
Each kind of claim is scored on its own, out of 100: the share that held up, with the range it likely falls in. Fewer resolved claims give a wider range.
-
Factual claims
Statements about the present or the past, checked against the record: did they hold up?
79 72 85 -
Predictions
Statements about events outside the person's control, checked once they fell due: did they come true?
80 72 88 -
Promises
Commitments to the person's or organisation's own future action, with something to deliver: was it delivered?
Insufficient record: 1 promise resolved, 5 needed for a score.
All types combined: 79 (likely 74 to 84, high confidence), from 102 claims: 64 factual claims, 38 predictions. It mixes kinds of claim in different proportions for different people, so compare people kind by kind.
Gary Marcus's claims
52 claims
“Neurosymbolic AI will become much more prominent.”
Gary Marcus, Substack author 1 January 2025 25 AI Predictions for 2025, from Marcus on AI (Marcus on AI)
Neurosymbolic AI will become much more prominent in 2025.
Mostly true True in substance, wrong in a detail.
What would count as true or false
Judge as of 31 Dec 2025. True: during 2025 neurosymbolic AI became much more prominent than in 2024, as shown by several of the following: a sharp rise in research papers, workshops or conference tracks; major labs or companies publicly adopting or launching systems they explicitly describe as neurosymbolic (e.g. combining LLMs with symbolic solvers, formal verification or code-execution reasoning); a notable increase in funding or startups; and markedly more media and industry attention. Mostly true: prominence rose clearly, but not 'much' in all of these respects (e.g. strong research growth but limited commercial use). Mixed: some measures rose noticeably while others stayed flat, so whether the rise was large is debatable. Mostly false: only a small increase, or attention limited to niche communities, with mainstream AI discourse still dominated by pure scaling of neural models. False: no notable increase, or a decline, in neurosymbolic AI's prominence in 2025. Systems count only if they are credibly described (by their makers or by independent analysts) as combining neural and symbolic methods.
Evidence
- Neuro-symbolic AI - Wikipedia Wikipedia
“Neuro-symbolic AI gained wider industrial adoption and public visibility in 2025 to address hallucination in large language models (LLMs); for example, Amazon applied it in its Vulcan warehouse robots and Rufus AI shopping assistant to enhance accuracy and decision-making.”
- The Quiet Return of an Old Idea | Daice Labs Daice Labs · 1 January 2026
“In August 2025, Amazon Web Services launched Automated Reasoning, designed to flag incorrect AI responses with claimed accuracy of up to 99 percent. Its cvc5 verifier reportedly performs roughly a billion automated reasoning checks each day across AWS cloud services.”
- The Quiet Return of an Old Idea | Daice Labs Daice Labs · 1 January 2026
“Startups are commercializing the approach across every regulated industry: Imandra for financial code verification, Elemental Cognition for enterprise reasoning (founded by David Ferrucci, who built Watson for Jeopardy!), Kognitos with a $25 million Series B in June 2025.”
- The Quiet Return of an Old Idea | Daice Labs Daice Labs · 1 January 2026
“Gartner placed neurosymbolic AI on its 2025 Hype Cycle as a composite emerging technology with the assessment that it can "augment and automate decision making with less risk of unintended consequences."”
Reasoning
E1 states that neuro-symbolic AI gained wider industrial adoption and public visibility in 2025, citing Amazon's use of it in its Vulcan warehouse robots and its Rufus shopping assistant. E2–E4 report three more 2025 developments: AWS launched an Automated Reasoning product in August, Kognitos raised a $25M Series B, and Gartner put neurosymbolic AI on its 2025 Hype Cycle. Together these meet several criteria: companies adopting such systems, new funding, and more industry attention. However, the evidence gives no figures on research papers, workshops or media coverage. E2's own title calls it a "quiet return", so it does not show that prominence rose sharply on every measure.
Cite this claim
Gary Marcus, “Neurosymbolic AI will become much more prominent in 2025.” (said 1 January 2025). Verdict: mostly true. Reliability API, claim c_dc0c60cfeea3, method 0.1.0+c876030c. https://epistacklabs.com/claims/c_dc0c60cfeea3
Claims without a verdict
Is a verdict here wrong, or do you speak for Gary Marcus? Reply on the claim's own page, with the form at its foot (how replies work). Verdicts are made automatically; how verdicts and scores are made.