An AI bot outperformed every human forecaster in the Summer 2026 Metaculus Cup.
Assessment
The evidence, examined directly, establishes the claim.
Metaculus, which organizes and scores the Cup, reported on 18 September 2026 that in the Summer 2026 Cup (885 forecasters, 58 questions) the bots Laertes and ManticAI finished first and second overall, ahead of every human participant, with FutureSearch fifth. Metaculus resolved its own question on whether a bot would beat all humans in that Cup as Yes, and called it a first for bots. Bots were ineligible for prizes and medals, so the human gold medallists (led by benshindel) are listed separately, but on the shared leaderboard the top bot ranked above all humans.
The result is narrow: one seasonal Cup of 58 questions, under Metaculus's tournament scoring. It does not by itself show that AI forecasters generally outperform top humans; other controlled comparisons have found professional forecasters still narrowly ahead of the best bots. The claim concerns only the Summer 2026 Cup, not the Spring or the then-ongoing Fall 2026 Cups.
Full reasoning: the evidence and decisions behind this verdict
Prior model: the claim is true if the organizer's official leaderboard placed a bot above every human; the only plausible ways it fails are a misreported season or a leaderboard that excluded humans who outscored bots. Primary source read in full: www.metaculus.com/notebooks/45734/announcing-the-metaculus-cup-summer-2026-winners/ states that Laertes and ManticAI finished 1st and 2nd overall and that this resolved "Will a bot beat all humans in the Summer 2026 Metaculus Cup?" as Yes. Secondary coverage (The Neuron; Tapen Sinha's Substack, which quotes the announcement; the announcement also cites an Economist piece of 16 September 2026) repeats the same organizer record and adds no independent evidence. The question page itself returned 403, so its resolution is confirmed only via the announcement. Scope: "the 2026 Metaculus Cup" was ambiguous among three seasonal Cups; the announcement's "a first for bots" implies no bot topped the Spring 2026 Cup, so the claim holds only for the Summer Cup. What would change this: a correction to the Summer leaderboard or resolution by Metaculus.
Decomposition
This claim is atomic: it bottoms out in a bedrock fact, an open empirical question, or a value premise, and does not decompose further.
Provenance
Where this claim has been said, linked to its canonical form.
This season Laertes and ManticAI finished 1st and 2nd overall, and FutureSearch finished 5th, which resolved our own question, Will a bot beat all humans in the Summer 2026 Metaculus Cup?, as Yes.
Official tournament-organizer announcement of Summer 2026 Cup results; bots ineligible for prizes but scored on the same questions.
An AI bot beat every human in the Summer 2026 Metaculus Cup.
News commentary on AI forecasting, repeating the Metaculus result.
Cite this claim: a formal citation with its evidence attached
Contribution record · 1 exchange
Challenges and proposals from contributors, with the review each received. Substantive exchanges stay on the claim's public record.
Metaculus's announcement of the Summer 2026 Cup winners reports that bots took first and second place overall (Laertes and ManticAI), ahead of every human participant, the first time a bot has won the Cup outright. FutureSearch finished fifth. Source: https://www.metaculus.com/notebooks/45734/announcing-the-metaculus-cup-summer-2026-winners/ . Worth assessing because it is widely cited as evidence that AI forecasting has passed human superforecasters, though a separate controlled comparison found pro forecasters still narrowly ahead of the best bots.
The proposal is a single, checkable proposition about a public event: whether an AI system placed ahead of every human forecaster in the Metaculus Cup (Summer 2026). It meets the claim bar of §2. It concerns a published competition result, not a private person, and the named bots and organizations are public participants. The contributor gave a sincere, on-topic reason and a source link, and an account with 12 accepted and 3 rejected contributions shows no sign of bad faith. The wording is workable as a canonical form. The Matcher and Steward can refine it, for example by clarifying what "outperformed" means (overall ranking versus a specific scoring period) and by checking whether the claim already exists in a broader form. The contributor's note that a separate controlled comparison found pro forecasters narrowly ahead of the best bots is context for the Steward. It is not part of the proposed claim. The assessment itself, including how far a single Cup result speaks to AI forecasting generally, is the Steward's work after admission.
Contribute
Every judgment on this page is open to challenge. A contribution is evaluated on its merits by the reviewer; if it succeeds the page changes, and if it does not, the reasons are stated. Either way the exchange becomes part of the claim’s public record.
The attention this claim received was paid for by a reader's assessment order. Funding buys only scheduling: it can make an assessment happen sooner, or reach deeper into a subtree. It has no influence on what the assessment concludes, and none on which claims enter the graph; assessments run under the same public standards whoever pays, funders never see or shape a verdict before anyone else, and mandates that attempt to steer conclusions are refused.
Created by user · Nov 2, 2026. Every judgment on this page is accompanied by a reasoning trace.