How many bots will finish in the top 10 of the Metaculus Cup Fall 2026?

- [2026-09-21] The Metaculus Cup Fall 2026 tournament page shows the season is active with 13 questions currently listed and runs from Aug 28 2026 to Jan 01 2027; no leaderboard/rankings data is yet meaningfully populated this early in the season. https://www.metaculus.com/tournament/metaculus-cup-fall-2026/

- [2026-09-18] In the just-completed prior season, the official Metaculus announcement states 
the Summer 2026 Cup is in the books, with 885 forecasters making 88,558 forecasts on 58 questions, and the community outperformed guest forecaster Dylan Matthews.
 This is the most recent full-season bot performance baseline feeding into Fall 2026 expectations.

- [2026-09-16] Per contemporaneous reporting, 
a bot beat every human in the Summer 2026 Metaculus Cup, taking first, second and fifth; humans held third and fourth
 place, meaning bots occupied 3 of the top 5, and per the question's own framing, 5 of the top 10 spots in Summer 2026 were bots (laertes and manticAI beat all humans).

- [2026-09-17] Secondary coverage confirms: 
In the Summer 2025 Metaculus Cup, British startup ManticAI finished eighth—notable because an AI system had finally broken into the tournament's top 10. One year later, a bot beat every human in the Summer 2026 Cup.
 This shows a clear year-over-year trajectory of bots climbing the leaderboard: 0 in top 10 historically → 1 (8th place, Summer 2025) → 5 of top 10 including 1st and 2nd (Summer 2026).

- [2026-04-13] Earlier in 2026, coverage of an intermediate cup noted 
in June, the top-ranked bot came 25th in the cup; this time, Mantic is eighth out of 549 contestants—the first time a bot has placed in the top 10 in the competition series.
 This confirms the rapid month-over-month/season-over-season improvement trend of bots climbing from no top-10 presence to dominant top-10 presence within roughly a year.

- [2026-09-17] A separate, more controlled benchmark tells a different story: 
The Spring 2026 FutureEval results: a team of 10 Metaculus Pro Forecasters edged the 10 best bots by an average of 1.25 head-to-head points per question across 99 shared questions — same questions, same window, difference in accuracy rather than in diligence.
 Analysts note this suggests the Cup's scoring (which rewards coverage/updating, not just accuracy) structurally favors bots over humans, independent of any pure forecasting-skill edge.

- [2026-09-17] Per the same analysis: 
Metaculus Cup participants earn points for accuracy and for coverage: how early they make predictions, how many questions they predict on, and how often they update — and that last clause is where a machine has a structural advantage that has nothing to do with forecasting skill.


- [Undated, standing program description] Metaculus explicitly states on its FutureEval page: 
When comparing Pros and the best custom bots, Metaculus Pro Forecasters won every season so far.
 This is in tension with the Cup outcome and underscores that the Cup's leaderboard result (bot-favorable) differs from the FutureEval head-to-head result (human-favorable), because of scoring-rule/format differences.

BOTTOM LINE: There is no current-value data point for the Fall 2026 Cup itself yet since it only just opened (13 questions live, season runs to Jan 1 2027, resolution Jan 2 2027); the most recent hard data point is Summer 2026's final leaderboard, where 5 of the top 10 finishers were bots (up from 1/10 a year earlier and 0/10 the year before that), showing a fast upward trajectory in bot placement driven partly by the Cup's coverage/updating scoring advantage for bots rather than pure accuracy superiority.

NOT FOUND: An official, itemized Summer 2026 Cup final leaderboard listing all top-10 names/accounts and explicit bot/human tags (the announcement notebook page did not render the leaderboard table via fetch); any official Metaculus statement specifically forecasting or previewing bot participation levels for Fall 2026; a live/current leaderboard snapshot for the Fall 2026 Cup (not available since the season just started and few questions have resolved).

Historical trajectory of bot placements in top-10 of Metaculus's seasonal human-vs-bot Cup (formerly "Quarterly Cup"):

- [2026-04-13] Summer 2025 (Jun–Aug 2025) Metaculus Cup: 
the top-ranked bot in June had come in 25th place, but by the end of that season Mantic placed eighth out of 549 contestants—the first time a bot had placed in the top 10 in the competition series.
 This was 1 bot in the top 10 that season.

- [2026-04-13] Context on the surprise: 
when the competition opened, participants predicted the top bot's score would be 40% of the top human performers' average; instead, Mantic achieved over 80%
 — a large positive surprise relative to prior forecasts.

- [2026-04-13] Analysts cautioned about small-sample noise: 
the contest contains a relatively small sample of 60 questions, and most of the 600 contestants are amateurs, some of whom only predict a handful of questions, leaving them with a low score
 — meaning the bar for "top 10" can be relatively low and volatile between seasons.

- [~2026-05] Fall 2025 Metaculus Cup: a dedicated Metaculus question tracked ManticAI's placement; the community forecast and eventual result put ManticAI around 4th–5th place (per the question's resolved range "5 Place (4-7)"), i.e., a bot again finished in the top 10, extending the streak to at least two consecutive seasons.

- [2026-09] Spring 2026 Metaculus Cup: a dedicated Metaculus question ("Spring 2026 Metaculus Cup Top Bot Rank?") resolved with the top bot finishing in 3rd place, i.e., inside the top 10 for a third consecutive season, and improving on the prior season's ~4th-8th finishes.

- [2026-09-16] Summer 2026 Metaculus Cup (the season immediately preceding the one being forecast): 
bots took first, second and fifth place; humans held third and fourth
 in the final leaderboard. The tournament description for the Fall 2026 question itself states bots occupied 5 of the top 10 spots that season, with two bots beating all human forecasters.

- [2026-09-16] The trend across the last four tracked seasons (Summer 2025 → Fall 2025 → Spring 2026 → Summer 2026) shows a monotonic increase in bot top-10 representation: roughly 1 → ~1 (top bot 4th-5th) → 1 (top bot 3rd) → 5 bots in the top 10, with two bots (laertesand manticAI) beating all humans outright.

- [2026-09-16] Separately, Metaculus runs FutureEval, a controlled bot-vs-Pro comparison that is NOT the same leaderboard as the Cup but is often cited alongside it: 
Spring 2026 results found that a team of 10 Metaculus Pro Forecasters still edged the 10 best bots by an average of 1.25 head-to-head points per question across 99 shared questions, though the difference was not statistically significant and nine of the 10 individual Pros still finished ahead of every individual bot.
 This shows the Cup leaderboard (which rewards coverage/speed, favoring bots structurally) diverges from controlled accuracy-only comparisons where humans still lead — a caveat on extrapolating the Cup trend.

- [2026-09-16] Analysts have flagged the structural reason bots do well specifically on the Cup leaderboard (not necessarily accuracy): 
Metaculus Cup participants earn points for accuracy and for coverage: how early they make predictions, how many questions they predict on, and how often they update
 — a scoring structure that mechanically favors tireless, always-on bots and has driven the rising bot count in the top 10 over time, independent of any accuracy convergence.

- [2026-09] No historical instance was found of the bot count in the top 10 decreasing season-over-season during the 2025–2026 period tracked; each successive season (as evidenced by dedicated Metaculus rank-tracking questions and the Summer 2026 outcome) has seen bot representation in the top 10 equal or grow.

BOTTOM LINE: Over the last four tracked seasons the number of bots finishing in the Cup's top 10 rose monotonically from 0 (pre-2025) to 1 (Summer 2025) to a handful of top-3/top-5 finishes (Fall 2025, Spring 2026) to 5 (Summer 2026, including 2 outright #1/#2 finishes above all humans), driven by a scoring structure that rewards bots' tireless coverage/updating rather than pure accuracy; this trend implies the Fall 2026 baseline expectation should be centered at or above 5, with real season-to-season volatility (the jump from 1 to 5 in one season shows swings of several bots are historically plausible) and a possible ceiling near saturation if increasingly capable bots also begin crowding out remaining human top scorers.

NOT FOUND: An official, complete numeric leaderboard listing exactly how many bot accounts (not just the top 1-2) placed in the top 10 for each of Fall 2025 and Spring 2026 specifically (only proxy data via individual bot-rank tracking questions was available); the exact number of participating/registered bots in the upcoming Fall 2026 Cup; any official Metaculus post-mortem stating the precise "bots in top 10" count for Summer 2026 beyond the "5 of top 10" figure cited in the question's own description.

**Key causal drivers for how many bots finish top-10 in Metaculus Cup Fall 2026:**

- **Trend trajectory of bot performance across recent Cup seasons.** 
In the Summer 2025 Metaculus Cup, ManticAI finished eighth, notable at the time because an AI system had finally broken into the tournament's top 10; one year later, a bot beat every human in the Summer 2026 Cup.
 
A bot beat every human in the Summer 2026 Metaculus Cup, taking first, second and fifth place, with humans holding third and fourth.
 This is a rapid, monotonic escalation (0→1→5 top-10 bot slots across three seasons), controlled by Metaculus's bot-tournament organizers and by bot-developer teams (e.g., ManticAI/laertes), not by Metaculus itself.

- **Scoring mechanics reward coverage/updating, which structurally favors bots.** 
Metaculus Cup participants earn points for accuracy and for coverage — how early they make predictions, how many questions they predict on, and how often they update — and that coverage clause gives machines a structural advantage that has nothing to do with forecasting skill, since an AI can predict on every question and revise continuously as news breaks.
 This mechanic is fixed by Metaculus's tournament design (unlikely to change before Dec 15 close) and consistently benefits any well-built bot regardless of underlying accuracy gains.

- **Underlying forecasting accuracy is close but still human-favored in controlled tests.** 
The Spring 2026 FutureEval results found a team of 10 Metaculus Pro Forecasters edged the 10 best bots by an average of 1.25 head-to-head points per question across 99 shared questions — same questions, same window, with humans still ahead on accuracy.
 
The difference was not statistically significant, however, and nine of the 10 individual Pros still finished ahead of every individual bot.
 This gap is controlled by the pace of LLM/agentic-search capability improvements (OpenAI, Anthropic, Google model releases) and by bot-developer engineering, and 
that result is from Spring 2026 and the field moves fast, so it may already have flipped, though the margin is not a chasm.


- **Eligibility rules: whether bots can occupy prize-eligible top-10 slots at all.** Historically 
the Metaculus Cup is the most popular human tournament on Metaculus, and though bots are not eligible for prizes, it can help bot-makers measure the strength of their bot on diverse questions.
 However, the Fall 2026 question's own fine print states the top-10 "will only include users who are eligible for prizes and bots," confirming bots are explicitly counted in this rank even without winning prize money — a rule set unilaterally by Metaculus and unchanged going into Fall 2026.

- **Number and quality of bot entrants competing in Fall 2026.** Growth in the bot ecosystem is driven by Metaculus's parallel $50k FutureEval Fall/Spring/Summer Bot Tournament and MiniBench programs, which run in parallel to the Cup and incentivize developers (e.g., ManticAI, "laertes") to field competitive entries in the Cup itself; 
Metaculus helps sponsor the LLM and search costs of participants in FutureEval bot tournaments via donations from Anthropic, Google, and OpenAI, as well as a partnership with AskNews
, lowering the barrier for more, better-resourced bots to enter Fall 2026.

BOTTOM LINE: The trend line moved from 0 bots in top-10 (pre-2025), to 1 (Summer 2025), to 5 (Summer 2026, including the outright #1 finish), driven by both a structural scoring advantage for bots (coverage/updating) and narrowing-but-still-real accuracy gaps versus top humans; absent a rule change (none found for Fall 2026), the trajectory and current mechanics point toward continued strong or higher bot representation in the Fall 2026 top 10.

NOT FOUND: No official Fall 2026 leaderboard or standings data exists yet (season is just opening, closes Dec 15, resolves Jan 2); no confirmation of exact number of bot entrants registered for Fall 2026; no updated FutureEval head-to-head figures beyond Spring 2026's 1.25-point pro advantage.

- [2026-09-05] The Summer 2026 Metaculus Cup resolved with bots occupying 5 of the top 10 spots, with bots laertes and manticAI finishing 1st and 2nd, ahead of all human forecasters. 
Bots took first, second and fifth place; humans held third and fourth, with a bot beating every human in the Summer 2026 Metaculus Cup.


- [2026-04-13] One season prior (Spring 2026), the trend was already positive but far weaker: 
in June, the top-ranked bot came 25th in the cup, while in the following season Mantic was eighth out of 549 contestants—the first time a bot had placed in the top 10 in the competition series.
 This shows the "bots in top 10" count went from 0 (implicitly) to 1 to 5 across just three consecutive seasons.

- [2026-07-02] Astral Codex Ten's mid-season tracking of the Summer 2026 Cup (before final resolution) found 
two top-ten AIs — manticAI and Laertes — repeated their prior top-ten performance, while a third (Preseen-Chestnut) had fallen to around #40
, indicating that bot placement in the top 10 is volatile season-to-season even among the same operators.

- [2026-09-16] The Economist covered the Summer 2026 result as a milestone, but analysts caution the Cup's scoring format (rewarding coverage/updating frequency, not just accuracy) structurally favors bots: 
Cup participants earn points for accuracy and for coverage—how early they predict, how many questions, and how often they update—and an AI that is less accurate than its human competitors can still finish higher by predicting on every question and revising continuously as news breaks.


- [~2026-09] In contrast, Metaculus's own controlled FutureEval comparison—matching bots and pro forecasters head-to-head on identical questions—found humans still ahead: 
Spring 2026 FutureEval results found a team of 10 Metaculus Pro Forecasters edged the 10 best bots by an average of 1.25 head-to-head points per question across 99 shared questions, and nine of the 10 individual Pros still finished ahead of every individual bot.
 The gap was not statistically significant, however.

- [2026-09] The FutureEval project page states this superiority has held every season measured so far: 
When comparing Pros and the best custom bots, Metaculus Pro Forecasters won every season so far.


- [~2026 (LessWrong synthesis)] Historical Cup placements for a single tracked bot show steady improvement: 
in the Summer 2025 Cup, metac-gemini-2-5-pro+asknews placed 37th of 551; in the Fall 2025 Cup it ranked 23rd of 539; and in the Spring 2026 Cup, metac-claude-4-5-sonnet-high-32k+asknews ranked 33rd of 1130 humans.
 This underlying trend (rising then Summer 2026 breakthrough to 5/10) is not perfectly monotonic — participant pool size and question mix vary each season.

- [2026-09] No explicit Metaculus community-forecast percentile distribution for the exact Fall 2026 "number of bots in top 10" question could be located; only the resolved precedent question ("Bot superiority in Summer 2026 Metaculus Cup?") was found, which 
resolved Yes
, alongside a related long-run question giving 
a 95% chance that a bot will win the Metaculus Cup sometime before 2030
 and, mid-Summer-2026, only 
a 15% chance that a bot would win that specific summer's Cup
 (i.e., markets underestimated the actual Summer 2026 outcome, in which bots swept 1st and 2nd).

- [2026-09] Commentary broadly agrees the trajectory favors bots continuing to place well, but flags genuine uncertainty about whether Summer 2026's 5-of-10 result is a new steady-state or partly a one-off driven by scoring-rule dynamics (coverage/update frequency) rather than pure accuracy gains, per the explainx.ai and Neuron analyses cited above.

BOTTOM LINE: The only directly relevant "expert forecast" evidence is the Metaculus community's own track record on precedent questions, which underestimated bot performance in Summer 2026 (assigning just 15% to a bot win, when bots actually took 1st, 2nd and 5th of 10) — suggesting community/market estimates for Fall 2026 should be anchored above naive extrapolation from Spring 2026 (1/10) toward or above the Summer 2026 realized value (5/10), while acknowledging FutureEval's controlled tests still show pros with a slight (non-significant) edge in pure accuracy.

NOT FOUND: No specific Metaculus community-prediction distribution/percentile for the actual Fall 2026 "number of bots in top 10" numeric question; no Fall 2026 Cup in-progress standings (season likely still early as of 2026-09-21, question closes 2026-12-15); no dedicated prediction-market (Kalshi/Polymarket/Manifold) odds for this specific Metaculus meta-question.

- [2026-09-05] The Summer 2026 Metaculus Cup concluded with winners announced; a question titled "Will a bot beat all humans in the Summer 2026 Metaculus Cup?" resolved Yes, and the Summer Cup page confirms 
"Metaculus now marks that question as resolved, and the Summer Cup page says its winners were announced September 5."
 — https://www.theneuron.ai/news/when-ai-predicts-better-than-people-who-sets-the-rules/

- [2026-09-16] The Economist reported the milestone directly, stating 
"For the first time an AI won the seasonal Metaculus Cup, a proving ground for forecasters of a wide variety of events."
 — https://hraness.com/reading/artificial-intelligence-now-beats-some-of-the-best-human-forecasters

- [2026-09-18/21] Bots occupied 5 of the top 10 spots in Summer 2026, with the bot "laertes" finishing 1st overall and manticAI 2nd overall, per multiple outlets: 
"Mantic finished ahead of every human contestant and behind just one other AI system, a bot named laertes."
 — https://techstartups.com/2026/09/18/british-ai-startup-mantic-raises-25m-to-build-superhuman-ai-forecasting-after-metaculus-win/

- [2026-09-18] Reuters-sourced reporting confirms bots took 1st, 2nd and 5th place in Summer 2026, with humans holding 3rd and 4th: 
"A bot beat every human in the Summer 2026 Metaculus Cup, taking first, second and fifth. Bots took first, second and fifth place; humans held third and fourth."
 — https://www.explainx.ai/blog/ai-beats-forecasters-metaculus-cup-coverage-2026

- [2026-09-18] Mantic (maker of manticAI) raised $25M in seed funding led by Radical Ventures, with Microsoft's M12, Thinking Machines Lab and Balderton Capital participating, directly citing the Metaculus Cup result as validation — 
"Radical Ventures led the round at an undisclosed valuation, joined by Microsoft's M12 venture fund, Thinking Machines Lab, Balderton Capital and other investors."
 This signals accelerating commercial investment and incentive for top bot-makers to keep competing and improving ahead of Fall 2026. — https://techstartups.com/2026/09/18/british-ai-startup-mantic-raises-25m-to-build-superhuman-ai-forecasting-after-metaculus-win/

- [~2026 (year-over-year trend)] The trend line into the Fall 2026 season is one of rapid acceleration: 
"In the Summer 2025 Metaculus Cup, British startup ManticAI finished eighth—a result notable enough at the time because an AI system had finally broken into the tournament's top 10... One year later, a bot beat every human in the Summer 2026 Cup."
 This year-over-year jump (8th → 1st/2nd/5th, 5 of top 10) is the strongest evidence for a high outcome in Fall 2026.

- [2026-09] A "Key Factors" analysis on the resolved Metaculus question itself flagged the trend explicitly: 
"Trend shows bots improving rapidly · Increases Likelihood"
 as a factor pushing toward more bot dominance, versus a countervailing factor that motivated human predictors increase difficulty. — https://www.metaculus.com/questions/43900/bot-superiority-in-summer-2026-metaculus-cup/

- [2026-05-01] Metaculus itself is actively expanding and funding the bot ecosystem for Fall 2026 (the same "primary" $50k bot tournament aligned with the Cup), including free LLM/search credits from Anthropic, Google, OpenAI and AskNews, which lowers the barrier for stronger bots to enter Fall 2026: 
"Metaculus helps sponsor the LLM and search costs of participants in FutureEval bot tournaments via donations from Anthropic, Google, and OpenAI, as well as a partnership with AskNews."
 — https://forum.effectivealtruism.org/posts/ZfLAN557rGWACKtmc/announcing-metaculus-summer-2026-futureeval-bot-tournament

- [2026-09-16] Even skeptical coverage acknowledges the historical progression is real and fast: in the Summer 2025 Cup only one bot (ManticAI, 8th) cracked the top 10, but 
"In the Summer 2025 Metaculus Cup, the British startup ManticAI finishing eighth was itself newsworthy, because it was the first time an AI system cracked the top ten. One year later a bot won the thing. Whatever you think of the scoring rule, that is a fast climb, and the coverage advantage is not going away, it is going to widen."
 — https://www.explainx.ai/blog/ai-beats-forecasters-metaculus-cup-coverage-2026

- [2026-09] Cup-specific scoring mechanics structurally favor bots relative to humans, per analysis: 
"Metaculus Cup participants earn points for accuracy and for coverage: how early they make predictions, how many questions they predict on, and how often they update. That last clause is where a machine has a structural advantage that has nothing to do with forecasting skill."
 This structural (non-skill) advantage is a key mechanism that could push bot representation in the Fall 2026 top 10 even higher than 5, toward 6, 7 or more, if more well-funded bot teams (like Mantic, laertes' developer, and newly-incentivized entrants after the funding news) field entries in Fall 2026. — https://www.explainx.ai/blog/ai-beats-forecasters-metaculus-cup-coverage-2026

- [2026-09-21] The Fall 2026 Metaculus Cup tournament page confirms the season is live and open, guest-forecaster format continues (against Kiko Llaneras) with prizes and started; the primary aligned $50k FutureEval Fall bot tournament runs concurrently, meaning well-resourced, funded bot teams (Mantic, laertes' developer, and others drawn in by the publicity/funding wave) have both incentive and infrastructure to compete hard in Fall 2026. — https://www.metaculus.com/tournament/metaculus-cup-fall-2026/

BOTTOM LINE: The strongest evidence favors a high outcome: bots went from 1 top-10 finisher (Summer 2025) to 5 (Summer 2026, including 1st and 2nd place), driven by both genuine capability gains and a structural scoring advantage (coverage/updating) that isn't going away, plus fresh capital and publicity (Mantic's $25M raise) likely to intensify bot-maker effort for Fall 2026 — making 5+ bots in the Fall 2026 top 10, and plausibly a repeat or expansion of bot dominance (6-8), a well-supported high scenario.

NOT FOUND: No official Metaculus Fall 2026 leaderboard data yet (season still open, closes Dec 15, resolves Jan 2), no confirmed list of which specific bots (beyond laertes/manticAI) are registered for Fall 2026, and no Metaculus community forecast/probability distribution specifically for the Fall 2026 "bots in top 10" question itself.

Low-scenario evidence (path to a small number of bots, e.g. 0–2, in the Fall 2026 top 10):

- [2026-09-09] Metaculus's own controlled comparison (FutureEval) found that in Spring 2026 
the spring 2026 tournament was the first in which none of our Metac Bots reached the top 10, and metac-gpt-5-1-high+asknews was the only one in the top 20
 — showing that in the Cup season immediately preceding Summer 2026's headline result, the leading in-house bots did not crack the top 10 at all, evidencing that "5 of top 10" was not a stable baseline.
- [2026-09-09] Even in the best available head-to-head test, 
our team of 10 Metaculus Pro Forecasters outperformed the top-10 bot team by an average of 1.25 head-to-head points per question... one-sided p = 0.247
, and critically 
nine of the top 10 places would be taken by Pros
 when bots and pros are ranked together individually on matched questions — i.e., under accuracy-only, matched-question conditions bots are still usually not top-10-caliber versus elite humans.
- [2026-09-17] Independent analysis notes the Cup's leaderboard rewards 
Metaculus Cup participants earn points for accuracy and for coverage: how early they make predictions, how many questions they predict on, and how often they update
, a structural, non-accuracy-related advantage; if Metaculus (or Cup competitors) adjust behavior to neutralize this coverage edge, bots' Cup ranking could fall back toward FutureEval's accuracy-only picture (near-zero bots in top 10).
- [2026-07-04] Prior-season Cup data (Astral Codex Ten) shows volatility even among bots that had cracked the top 10: 
So did two top-ten AIs - manticAI and Laertes (Preseen-Chestnut is having a tough summer and is down to #40)
 — a bot that was previously top 10 fell to 40th the very next season, illustrating that individual bot placements are not "sticky" and can regress sharply.
- [2026-04-13] Historical base rate before the Summer 2026 spike was much lower: 
In June, the top-ranked bot, built by Metaculus on top of OpenAI's o1 reasoning model, came 25th in the cup... This time, Mantic is eighth out of 549 contestants—the first time a bot has placed in the top 10 in the competition series
. Only one bot cracked the top-10 barrier as recently as Summer 2025, and a Fall 2025/Spring 2026 relapse to zero bots in top-10 (per FutureEval) followed before Summer 2026's outlier result.
- [2026-09-05] Metaculus's own community-forecast question on this dynamic flagged a factor that could suppress future bot placements: the key factor "Motivated human predictors increase difficulty" was tagged as decreasing the likelihood of continued bot dominance — i.e., heightened media attention (Economist, Time, Astral Codex Ten) on bots' Cup success may draw more skilled, motivated human forecasters into Fall 2026 specifically to compete with bots, raising the bar bots must clear. (https://www.metaculus.com/questions/43900/bot-superiority-in-summer-2026-metaculus-cup/)
- [2026-09-17] A close reading of the coverage-vs-accuracy gap concludes: 
bots now win tournaments whose scoring rewards tireless coverage, while the best humans retained a small edge in accuracy on matched questions as of the most recent controlled test
 — a status quo that is fragile to any rule tweak, question-mix change, or renewed human effort in Fall 2026.
- [2026-09-21] The Metaculus Cup Fall 2026 tournament page confirms the season is underway with a $5,000 prize pool and a human-vs-guest-forecaster framing 
This season the Metaculus community goes head-to-head against guest forecaster Kiko Llaneras. If the Community Prediction beats him by season's end, the prize pool grows by an additional $2,500. Tournament · Follow · View Questions(13)
, but no bot-specific standings for Fall 2026 are yet published (too early in the season, opened 2026-09-21) to confirm any trend either way.

BOTTOM LINE: The strongest evidence for a low outcome (0–2 bots in Fall 2026's top 10) is that Metaculus's own rigorous, matched-question FutureEval comparisons found zero bots in the top 10 as recently as Spring 2026 and pros still generally out-rank bots individually; the "5 of top 10" Summer 2026 result appears to be an outlier driven by a coverage/update-frequency scoring advantage plus reduced elite-human effort, both of which are plausible to partially reverse in Fall 2026 if human participation intensifies or bot placements regress toward their more volatile historical norm.

NOT FOUND: No published Fall 2026 Metaculus Cup leaderboard or bot-specific standings exist yet (season just opened Sept 21, 2026, closes Dec 15); no confirmation of any rule change for Fall 2026 specifically altering coverage/update scoring; no FutureEval Fall 2026 analysis (not yet run).

13 AskNews articles (newest first; 6 from the last 48 h, 9 from the archive). Summaries are AskNews's own; dates are publication dates.

[2026-09-21] Neuquén to Host 2026 National Robotic Cup Final with $100 Million Prize
Diario Río Negro — https://www.rionegro.com.ar/sociedad/copa-robotica-2026-neuquen-sera-sede-de-la-final-nacional-que-define-el-pase-al-mundial-4730450/
  - The 2026 National Robotic Cup (Copa Robótica Educabot 2026) will take place on September 25 and 26, 2026, at the Centro de Convenciones Domuyo in Neuquén, Argentina, to determine the national champion that will represent the country at the First Global Challenge 2027.
  - The event, organized by the technology company Educabot and the Fundación Sadosky, brings together 48 delegations from 24 jurisdictions across Argentina.
  - Each delegation consists of five students aged 15 to 18 and one mentor teacher.
  - The competition features a prize pool of $100 million Argentine pesos in educational technology and equipment, distributed among participating schools and provinces.
  - The winning team will secure direct qualification for the international tournament.
  - The event is free and ensures equal conditions by providing identical robotics kits to all teams, supported by the Experiential Robotics Platform.
  - This year's edition involved over 15,000 students from 1,300 institutions nationwide, with 66% of schools located outside major urban centers.
  - Neuquén will host two local representatives: the 'Casi Media Docena' team from Colegio Neuquén Oeste, which won the provincial title, and the 'Wind Warriors' team from EPET N° 25 of Plottier.
  - The event is supported by the Neuquén Provincial Government, including the Ministry of Education, the Ministry of Youth, Sports and Culture, and Ecydense.
  - Josefina Codermatz, Minister of Youth, Sports and Culture, stated it is a 'very good opportunity for our young people to get closer to technology, innovation, and teamwork.' Federico Bolan, Vice President of Ecydense, expressed pride in hosting the federal event, while Lucas Godoy, Undersecretary of Educational Technology, emphasized the policy goal of making technology accessible to all schools.

[2026-09-21] U.S. Presidents Cup '26 Team Capsules: Scheffler, Clark, and Rising Stars Lead the Charge
AP NEWS — https://apnews.com/article/presidents-cup-medinah-us-player-capsules-bdc26fb0dc1b46617469725695639d60
  - The U.S. team for the 2026 Presidents Cup, starting Thursday at Medinah Country Club in Medinah, Illinois, features 12 players with varying records and achievements.
  - World No. 1 Scottie Scheffler leads the team with 25 career wins, including Masters titles in 2022 and 2024, and PGA Championship and British Open victories in 2025, though his team record stands at 3-5-1.
  - Wyndyham Clark, ranked No. 5, secured victories at the CJ Cup Byron Nelson, U.S. Open (his second major win after 2023), and BMW Championship.
  - Xander Schauffele, ranked No. 10, holds a 10-4-0 Presidents Cup record and won the PGA Championship and British Open in 2024.
  - Patrick Cantlay, despite a 10-4-0 Presidents Cup record, has not won since 2022 and has not finished in the top 10 of his last nine majors.
  - Rookie Jacob Bridgeman, a captain's pick ranked No. 20, leads the PGA Tour in putting for 2026 and recently won the Biltmore Championship Asheville with a closing 61.
  - Chris Gotterup, ranked No. 7, won three events in 2026: the Sony Open, WM Phoenix Open, and John Deere Classic.
  - Russell Henley, ranked No. 6, won the Charles Schwab Challenge in 2026 and previously went 3-1-0 in his debut Presidents Cup.
  - Jackson Koivun, the No. 1 amateur in the world, became the first player since Jordan Spieth in 2013 to make the team without holding a PGA Tour card at the year's start, winning the 3M Open in his third PGA Tour start.
  - Sam Burns, ranked No. 8, has not won since 2023 and had two close calls in majors this year.

[2026-09-21] Roar and Score Trivia Podcast Episode 130: Meta Quest, eBay Logo Challenge, and Spelling Bee (interview)
Roar and Score Trivia — https://mcdn.podbean.com/mf/web/4h46bkz8i3mtsj5z/Episode_130_fixedax3a5.mp3
  - Published on September 21, 2026, the 'Roar and Score Trivia' podcast released Episode 130, hosted by 'The Quizzard.' The episode features a 35-point quiz structure, including a 20-question round, a 5-point eBay logo challenge, and a 5-point bonus round.
  - The host introduces a rhyming game called 'I'm a Poet and I Didn't Even Know It,' where listeners must find answers rhyming with 'bun.' The host also shares anecdotes about purchasing a Meta Quest headset, describing it as 'the best thing I've ever bought' and noting its use for projecting a TV screen over the sink.
  - The 20-question round covers diverse topics, including the anime 'Pokemon' (featuring Ash Ketchum), the 1973 hit 'Jolene' by Dolly Parton, the 1941 Pearl Harbour attack, and the highest-grossing superhero movie as of September 2026, identified as 'Avengers Endgame.' The difficult 5-point question asks for the specific colors of the eBay logo in order: red, blue, yellow, and green.
  - The bonus round is a spelling bee requiring the correct spelling of five countries: Switzerland, Afghanistan, Guatemala, Djibouti, and Kyrgyzstan.
  - The episode concludes with a character impression by Severus Snape, who awards 'Gryffindor points' before signing off.

[2026-09-20] Vitality Wins Mirage 13-4 After Aurora Player Insults Team as 'Bots'
СПОРТ.UA — https://sport.ua/uk/news/908887-video-nul-raundiv-vzyali-aurora-proti-vitality-pislya-obrazlivih-sliv
  - During the grand final of the StarLadder StarSeries Fall 2026 between Vitality and Aurora in Barcelona, a pivotal moment occurred on the third map, Mirage, with the score at 4-3 in favor of Aurora.
  - According to reports from Sport.ua, a player from the Turkish team Aurora shouted 'Fucking Bots' (Ukrainian: 'йоб..ні боти') at the French team, Vitality, allegedly insulting their skill level.
  - Following this incident, Vitality allegedly won the next 9 consecutive rounds without conceding any, ending the map with a decisive 13-4 victory.
  - At the time of publication on September 20, 2026, Vitality leads the best-of-five series 2-1.
  - The tournament, which features a $500,000 prize pool, is scheduled to run from September 17 to September 20, 2026.

[2026-09-20] Meta Platforms Appeals UK Online Safety Act in Court
Market Screener — https://www.marketscreener.com/news/meta-platforms-appeals-against-online-safety-act-in-uk-court-ce785adad181f224
  - Meta Platforms, Inc. has appealed against the Online Safety Act in UK court, according to a Market Screener report published on September 20, 2026.
  - The company specializes in online social networking services, with net sales breakdown showing 98.9% from operating platforms (Facebook, Instagram, Messenger, Threads, and WhatsApp), which had 3.58 billion daily active users in 2025.
  - The remaining 1.1% comes from the sale of virtual and augmented reality products and devices, including Meta Quest headsets and Facebook Portal screens.
  - Revenue sources are 98.7% from advertising spaces and 1.3% from other sources.
  - Geographically, net sales are distributed as follows: United States and Canada (39.2%), Asia/Pacific (26.8%), Europe (23.2%), and other regions (10.8%).

[2026-09-19] Germany Volleyball Stages Dramatic Comeback to Reach Euro Championship Quarterfinals
N-tv — https://www.n-tv.de/sport/Deutsche-Volleyballer-kaempfen-sich-dramatisch-ins-EM-Viertelfinale-id31325310.html
  - On September 19, 2026, the German national volleyball team, led by head coach Massimo Botti, advanced to the European Championship quarterfinals for the first time in five years after a dramatic 3-2 victory over host nation Bulgaria in Sofia.
  - The match, played before over 12,000 fans, saw Germany overcome a 0-2 set deficit with a final score of 3-2 (21-25, 21-25, 25-19, 25-23, 15-12).
  - Erik Röhrs was the top scorer for the German Volleyball Association (DVV) with 20 points.
  - The victory comes despite the recent injury to setter Johannes Tille, who suffered a metatarsal fracture, and a Group Stage loss to Olympic champion France that prevented Germany from securing first place.
  - Germany will face Poland, the tournament favorite and world number one, in the quarterfinals scheduled for Tuesday, September 22, 2026.

[2026-09-17] AI Surpasses Human Forecasters in Metaculus Cup for First Time
Hindustan Times — https://www.hindustantimes.com/world-news/artificial-intelligence-now-beats-some-of-the-best-human-forecasters-101789639904836.html
  - On September 5, 2026, an AI system won the seasonal Metaculus Cup for the first time, securing first place while other AIs took second and fifth, leaving third and fourth for humans.
  - The competition involved hundreds of entrants predicting outcomes on topics ranging from data-center restrictions in American states or EU countries to the hantavirus outbreak on MV Hondius and Brent crude prices, with scoring based on prediction accuracy over time.
  - This milestone marks a shift where AI, built on large language models (LLMs) rather than narrow training data, can now match the performance of 'superforecasters' identified in a 2015 study by Barbara Mellers of the University of Pennsylvania, and potentially exceed them on evolving questions according to a July analysis by the Forecasting Research Institute.
  - The winning bot was developed by Jeffrey Liang, a polymath from Texas, who reportedly spent less than 150 hours and a few thousand dollars on computing power, beating four startups that had collectively raised over $15 million.
  - While Yann Riviere of the British startup Mantic notes that human judgment may still hold an edge for forecasts with horizons of years, the article suggests that as AI processing power grows, the balance will likely shift, potentially revealing the theoretical limits of predictability, similar to how weather forecasts become random after 15 days due to atmospheric chaos.

[2026-09-09] Spring 2026 FutureEval: Human Pros Edge AI Bots as Gap Narrows (analysis)
Maya Farber Brodsky — https://www.lesswrong.com/posts/wZBbDqzfBjYG58CxK/futureeval-spring-results-pros-beat-bots-but-the-gap-is
  - In the Spring 2026 FutureEval forecasting tournament held from January 7th to April 15th, 2026, a team of 10 Metaculus Pro Forecasters outperformed a team of the top 10 AI bots by an average of 1.25 head-to-head points per question, though this margin was not statistically significant (one-sided p = 0.247).
  - The tournament, which included 173 bots and 297 scored questions on events resolving between January and May 2026, marked a significant narrowing of the gap between human and AI performance; the bot team's head-to-head score against Pros improved from -17.7 in Q1 2025 to -1.25 in Spring 2026.
  - While individual Pros dominated the leaderboard with 9 of the top 10 spots, the top-performing bot, GreeneiBot2, utilized a complex ensemble of models including GPT-5 and research tools like AskNews and Perplexity.
  - OpenAI's GPT-5.1 (specifically the 'high' reasoning variant) emerged as the highest-scoring model among in-house Metac Bots, achieving an average score of 11.32, and surveys indicated that 9 of the top 10 bot makers used GPT-5.4 or higher in their ensembles.
  - The competition, featuring a $50,000 prize pool for non-Metaculus botmakers, also revealed that increasing bot team size beyond 10 members reduced performance and that no single search provider offered a statistically significant advantage over others.

[2026-08-13] PARIVISION Falls to B8 in Esports World Cup Opener
Спорт-Экспресс — https://www.sport-express.ru/cybersport/esports/news/parivision-s-jame-proigrala-b8-so-schetom-11-13-na-esports-world-cup-2446664/
  - In the opening match of the group stage at Esports World Cup 2026, the Russian team PARIVISION, featuring player Jame 'Jame' Ali, lost to B8 on the map Cache with a score of 13-11 in a best-of-1 series.
  - The match took place on August 13, 2026, in Paris.
  - PARIVISION trailed 5-7 at halftime; although both teams won six rounds in the second half, B8 maintained the lead.
  - B8 dominated the first kills in rounds 9-15.
  - Ivan 'zweih' Gogin was the top performer for PARIVISION with 18 kills, 16 deaths, and a 1.28 rating, while Jame recorded 12 kills, 14 deaths, and a 0.78 rating.
  - The best player of the map was B8's Andrey 'npl' Kukharsky, who achieved a 1.31 rating and an average of 84.5 damage per round.
  - Following the loss, PARIVISION dropped to the lower bracket of Group C and will face 100 Thieves next; a loss to them will eliminate PARIVISION from the tournament.
  - B8 advances to the upper bracket to play Vitality.
  - The tournament, running from August 12 to 23, features 32 teams competing for a $2 million prize pool.

[2026-07-08] AI Forecasting in 2026: Synthesis of 11 Metaculus Analyses (analysis)
Maya Farber Brodsky — https://www.lesswrong.com/posts/a82q6yd8zKpYk56cF/ai-forecasting-in-2026-what-11-analyses-say
  - A synthesis of 11 Metaculus analyses conducted between October 2024 and May 2026, authored by Maya Farber Brodsky and published on July 08 2026, evaluates the state of AI forecasting against human experts.
  - The report finds that while bots are better than the general public, they currently trail superforecasters; as of October 2025, the median public forecaster ranked #22 on FRI's ForecastBench, trailing superforecasters by a Brier score of 0.02 (0.101 for GPT-4.5 vs 0.081 for superforecasters).
  - In Metaculus tournaments, bots have placed competitively, with metac-claude-4-5-sonnet-high-32k+asknews ranking 33rd of 1130 humans (top ~3%) in the Spring 2026 Cup.
  - However, the 'Pro' team has beaten the 'bot' team in every FutureEval comparison from Q3 2024 to Q2 2025, with head-to-head scores ranging from -11.3 to -20.03.
  - Extrapolations suggest bots may surpass top humans around June 2027 (FutureEval) or November 2026 (ForecastBench).
  - Key strategies for building effective bots include using frontier reasoning models, investing in forecasting harnesses (scaffolding), employing agentic search, and ensembling 3-7 diverse runs.
  - Conversely, betting on a single search provider or using naive backtesting with information leakage yields poor results.
  - The article concludes that the current bottleneck is product integration rather than accuracy, with recommendations to integrate probability readouts into existing workflows and use AI for world simulation.

[2026-06-08] Meta Confirms Over 20,000 Instagram Accounts Hacked via AI Support Bot Flaw
TechRadar — https://www.techradar.com/pro/security/meta-reveals-over-20-000-instagram-accounts-hacked-and-stolen-using-ai-support-bot
  - Meta confirmed that 20,225 Instagram accounts were compromised via a vulnerability in its High Touch Support (HTS) AI-powered customer service bot.
  - The incident, discovered on May 31, 2026, allowed attackers to exploit a bug where the system failed to verify if a requested password reset email matched the account holder's registered email, instead sending reset links to unassociated addresses.
  - While Meta states there is no evidence of data exfiltration, the compromised data could include contact information, social media posts, direct messages, and account history.
  - In response, Meta disabled the HTS system, reset passwords for all affected profiles, and enrolled them in a mandatory security checkpoint.
  - The company plans to fix the authentication check before re-launching the tool and is conducting a comprehensive review of similar recovery flows across its platforms.
  - Muhammad Yahya Patel, vCISO at Huntress, noted that this highlights a new category of risk where AI tools can shift the attack surface from technical to logical vulnerabilities, emphasizing the need for rigorous security reviews before AI deployment.

[2026-05-19] Botafogo Leads Brazil in World Cup Player Supply with 48 Call-ups
OneFootball GmbH — https://onefootball.com/en/news/club-stays-top-for-most-players-supplied-to-selecao-see-top-10-42891856
  - According to data compiled by OneFootball GmbH, no club in Brazil has supplied as many players to the Brazilian National Team for World Cups as Botafogo.
  - With the recent call-up of midfielder Danilo by coach Carlo Ancelotti on Monday (18), Botafogo's total has risen to 48 call-ups, securing the top spot in the ranking.
  - São Paulo follows with 46 call-ups, and Flamengo ranks third with 39.
  - The top 10 list is dominated by Brazilian clubs.
  - The article lists the specific players for each club; for example, Botafogo's roster includes legendary figures such as Garrincha (2 call-ups), Nilton Santos (4 call-ups), and Didi (2 call-ups).
  - Santos, ranked sixth, contributed 25 call-ups, including Pelé (4 call-ups) and Neymar.
  - Palmeiras is seventh with 24 call-ups, led by Leão (4 call-ups), while Corinthians is eighth with 23.
  - Atlético-MG and Cruzeiro round out the top 10 with 12 and 11 call-ups respectively.
  - The article was published on May 19, 2026.

[2026-05-05] The Griffin Bots Team from Chécy Advances to Asia-Pacific Finals After Strong Showing in UK
www.larep.fr — https://www.larep.fr/checy-45430/actualites/l-equipe-the-griffin-bots-de-checy-en-route-pour-la-finale-asie-pacifique-de-robotique-a-sydney-en-juillet_14978799/
  - The 'The Griffin Bots' team from Chécy participated in the First Lego League international robotics competition held in Harrogate, England, on April 25 and 26, 2026.
  - The event combined robotics, programming, and scientific problem-solving.
  - Out of 67 participating teams, The Griffin Bots ranked 10th in the robot game.
  - Although their performance advanced them to an elimination tournament, they were eliminated in the quarter-finals.
  - The team was awarded the 'Motivation Prize' for their energy, attitude, and presentation to the jury.
  - They are now heading to the Asia-Pacific finals in Sydney in July.

## Tool outputs

### resolution_source (SUCCESS)
RESOLUTION SOURCE LOOKUP — page fetched is NOT the source the criteria name. Do NOT anchor on it; treat as weak background at most:
  Current state: The fetched page is a notebook titled 'Pros Beat Bots, but the Gap is Nearly Gone' discussing the Spring 2026 FutureEval Bot Tournament results, comparing bot vs pro forecaster performance, but it does not contain any data on the Metaculus Cup Fall 2026 leaderboard or bot placements in its top 10.
  Last updated: Sep 4, 2026 (edited) | already resolved: unclear | page matches criteria: False
  Note: This page is background/context referenced in the question description (a related FutureEval analysis), not the actual Metaculus Cup Fall 2026 leaderboard needed to resolve the question. No information about Fall 2026 results was found in the fetched content.
  Fetched: https://www.metaculus.com/tournament/metaculus-cup-fall-2026/ (HTTP 403)
  Fetched: https://www.metaculus.com/tournament/metaculus-cup-summer-2026/ (HTTP 403)
  Fetched: https://www.metaculus.com/notebooks/45373/spring-2026-futureeval-analysis/ (HTTP 200)

### claude_news (SUCCESS)

You are an elite superforecaster. Produce a probability distribution over the answer to this Metaculus numeric question.

## Question
How many bots will finish in the top 10 of the Metaculus Cup Fall 2026?

## Description / Resolution Criteria
## Description
In the [Metaculus Cup Summer 2026](https://www.metaculus.com/tournament/metaculus-cup-summer-2026/), forecasting bots [made history](https://www.economist.com/science-and-technology/2026/09/16/artificial-intelligence-now-beats-some-of-the-best-human-forecasters) by occupying 5 of the top 10 spots in the final leaderboard, with two bots [laertes](https://www.metaculus.com/accounts/profile/277919/) and [manticAI](https://www.metaculus.com/accounts/profile/191026/) beating all human forecasters. Bot forecasting has gotten increasingly better in the past two years, with the latest formal analysis from [spring 2026](https://www.metaculus.com/notebooks/45373/spring-2026-futureeval-analysis/) showing that bots have almost matched Pro Forecasters.

## Resolution Criteria
This question will resolve as the number of accounts in the top 10 of the Metaculus Cup Fall 2026 that are registered as bots, after all questions besides this one have been resolved.

## Fine Print
The top 10 spots will only include users who are eligible for prizes and bots. Admins and other excluded human accounts will not be counted.

## Range
The question's displayed range is [-0.5, 10.5] (units: bots).
The LOWER bound is CLOSED: the outcome cannot be below -0.5.
The UPPER bound is CLOSED: the outcome cannot be above 10.5.

## Research tool outputs (raw; the resolution-source read is the status-quo anchor)
## Tool outputs

### resolution_source (SUCCESS)
RESOLUTION SOURCE LOOKUP — page fetched is NOT the source the criteria name. Do NOT anchor on it; treat as weak background at most:
  Current state: The fetched page is a notebook titled 'Pros Beat Bots, but the Gap is Nearly Gone' discussing the Spring 2026 FutureEval Bot Tournament results, comparing bot vs pro forecaster performance, but it does not contain any data on the Metaculus Cup Fall 2026 leaderboard or bot placements in its top 10.
  Last updated: Sep 4, 2026 (edited) | already resolved: unclear | page matches criteria: False
  Note: This page is background/context referenced in the question description (a related FutureEval analysis), not the actual Metaculus Cup Fall 2026 leaderboard needed to resolve the question. No information about Fall 2026 results was found in the fetched content.
  Fetched: https://www.metaculus.com/tournament/metaculus-cup-fall-2026/ (HTTP 403)
  Fetched: https://www.metaculus.com/tournament/metaculus-cup-summer-2026/ (HTTP 403)
  Fetched: https://www.metaculus.com/notebooks/45373/spring-2026-futureeval-analysis/ (HTTP 200)

### claude_news (SUCCESS)

## Question dates and conventions
Dates: opened 2026-09-21; forecasting closes 2026-12-15; resolves 2027-01-02. Only events AFTER the open date can satisfy the criteria; assume the event has NOT happened unless a source shows it has. A question phrased 'before <date>' or 'by <date>' asks whether it happens between now and that date. The deadline for the outcome is the RESOLUTION date, not the close date.

## Recent news (AskNews, raw article summaries, newest first — read these directly)
13 AskNews articles (newest first; 6 from the last 48 h, 9 from the archive). Summaries are AskNews's own; dates are publication dates.

[2026-09-21] Neuquén to Host 2026 National Robotic Cup Final with $100 Million Prize
Diario Río Negro — https://www.rionegro.com.ar/sociedad/copa-robotica-2026-neuquen-sera-sede-de-la-final-nacional-que-define-el-pase-al-mundial-4730450/
  - The 2026 National Robotic Cup (Copa Robótica Educabot 2026) will take place on September 25 and 26, 2026, at the Centro de Convenciones Domuyo in Neuquén, Argentina, to determine the national champion that will represent the country at the First Global Challenge 2027.
  - The event, organized by the technology company Educabot and the Fundación Sadosky, brings together 48 delegations from 24 jurisdictions across Argentina.
  - Each delegation consists of five students aged 15 to 18 and one mentor teacher.
  - The competition features a prize pool of $100 million Argentine pesos in educational technology and equipment, distributed among participating schools and provinces.
  - The winning team will secure direct qualification for the international tournament.
  - The event is free and ensures equal conditions by providing identical robotics kits to all teams, supported by the Experiential Robotics Platform.
  - This year's edition involved over 15,000 students from 1,300 institutions nationwide, with 66% of schools located outside major urban centers.
  - Neuquén will host two local representatives: the 'Casi Media Docena' team from Colegio Neuquén Oeste, which won the provincial title, and the 'Wind Warriors' team from EPET N° 25 of Plottier.
  - The event is supported by the Neuquén Provincial Government, including the Ministry of Education, the Ministry of Youth, Sports and Culture, and Ecydense.
  - Josefina Codermatz, Minister of Youth, Sports and Culture, stated it is a 'very good opportunity for our young people to get closer to technology, innovation, and teamwork.' Federico Bolan, Vice President of Ecydense, expressed pride in hosting the federal event, while Lucas Godoy, Undersecretary of Educational Technology, emphasized the policy goal of making technology accessible to all schools.

[2026-09-21] U.S. Presidents Cup '26 Team Capsules: Scheffler, Clark, and Rising Stars Lead the Charge
AP NEWS — https://apnews.com/article/presidents-cup-medinah-us-player-capsules-bdc26fb0dc1b46617469725695639d60
  - The U.S. team for the 2026 Presidents Cup, starting Thursday at Medinah Country Club in Medinah, Illinois, features 12 players with varying records and achievements.
  - World No. 1 Scottie Scheffler leads the team with 25 career wins, including Masters titles in 2022 and 2024, and PGA Championship and British Open victories in 2025, though his team record stands at 3-5-1.
  - Wyndyham Clark, ranked No. 5, secured victories at the CJ Cup Byron Nelson, U.S. Open (his second major win after 2023), and BMW Championship.
  - Xander Schauffele, ranked No. 10, holds a 10-4-0 Presidents Cup record and won the PGA Championship and British Open in 2024.
  - Patrick Cantlay, despite a 10-4-0 Presidents Cup record, has not won since 2022 and has not finished in the top 10 of his last nine majors.
  - Rookie Jacob Bridgeman, a captain's pick ranked No. 20, leads the PGA Tour in putting for 2026 and recently won the Biltmore Championship Asheville with a closing 61.
  - Chris Gotterup, ranked No. 7, won three events in 2026: the Sony Open, WM Phoenix Open, and John Deere Classic.
  - Russell Henley, ranked No. 6, won the Charles Schwab Challenge in 2026 and previously went 3-1-0 in his debut Presidents Cup.
  - Jackson Koivun, the No. 1 amateur in the world, became the first player since Jordan Spieth in 2013 to make the team without holding a PGA Tour card at the year's start, winning the 3M Open in his third PGA Tour start.
  - Sam Burns, ranked No. 8, has not won since 2023 and had two close calls in majors this year.

[2026-09-21] Roar and Score Trivia Podcast Episode 130: Meta Quest, eBay Logo Challenge, and Spelling Bee (interview)
Roar and Score Trivia — https://mcdn.podbean.com/mf/web/4h46bkz8i3mtsj5z/Episode_130_fixedax3a5.mp3
  - Published on September 21, 2026, the 'Roar and Score Trivia' podcast released Episode 130, hosted by 'The Quizzard.' The episode features a 35-point quiz structure, including a 20-question round, a 5-point eBay logo challenge, and a 5-point bonus round.
  - The host introduces a rhyming game called 'I'm a Poet and I Didn't Even Know It,' where listeners must find answers rhyming with 'bun.' The host also shares anecdotes about purchasing a Meta Quest headset, describing it as 'the best thing I've ever bought' and noting its use for projecting a TV screen over the sink.
  - The 20-question round covers diverse topics, including the anime 'Pokemon' (featuring Ash Ketchum), the 1973 hit 'Jolene' by Dolly Parton, the 1941 Pearl Harbour attack, and the highest-grossing superhero movie as of September 2026, identified as 'Avengers Endgame.' The difficult 5-point question asks for the specific colors of the eBay logo in order: red, blue, yellow, and green.
  - The bonus round is a spelling bee requiring the correct spelling of five countries: Switzerland, Afghanistan, Guatemala, Djibouti, and Kyrgyzstan.
  - The episode concludes with a character impression by Severus Snape, who awards 'Gryffindor points' before signing off.

[2026-09-20] Vitality Wins Mirage 13-4 After Aurora Player Insults Team as 'Bots'
СПОРТ.UA — https://sport.ua/uk/news/908887-video-nul-raundiv-vzyali-aurora-proti-vitality-pislya-obrazlivih-sliv
  - During the grand final of the StarLadder StarSeries Fall 2026 between Vitality and Aurora in Barcelona, a pivotal moment occurred on the third map, Mirage, with the score at 4-3 in favor of Aurora.
  - According to reports from Sport.ua, a player from the Turkish team Aurora shouted 'Fucking Bots' (Ukrainian: 'йоб..ні боти') at the French team, Vitality, allegedly insulting their skill level.
  - Following this incident, Vitality allegedly won the next 9 consecutive rounds without conceding any, ending the map with a decisive 13-4 victory.
  - At the time of publication on September 20, 2026, Vitality leads the best-of-five series 2-1.
  - The tournament, which features a $500,000 prize pool, is scheduled to run from September 17 to September 20, 2026.

[2026-09-20] Meta Platforms Appeals UK Online Safety Act in Court
Market Screener — https://www.marketscreener.com/news/meta-platforms-appeals-against-online-safety-act-in-uk-court-ce785adad181f224
  - Meta Platforms, Inc. has appealed against the Online Safety Act in UK court, according to a Market Screener report published on September 20, 2026.
  - The company specializes in online social networking services, with net sales breakdown showing 98.9% from operating platforms (Facebook, Instagram, Messenger, Threads, and WhatsApp), which had 3.58 billion daily active users in 2025.
  - The remaining 1.1% comes from the sale of virtual and augmented reality products and devices, including Meta Quest headsets and Facebook Portal screens.
  - Revenue sources are 98.7% from advertising spaces and 1.3% from other sources.
  - Geographically, net sales are distributed as follows: United States and Canada (39.2%), Asia/Pacific (26.8%), Europe (23.2%), and other regions (10.8%).

[2026-09-19] Germany Volleyball Stages Dramatic Comeback to Reach Euro Championship Quarterfinals
N-tv — https://www.n-tv.de/sport/Deutsche-Volleyballer-kaempfen-sich-dramatisch-ins-EM-Viertelfinale-id31325310.html
  - On September 19, 2026, the German national volleyball team, led by head coach Massimo Botti, advanced to the European Championship quarterfinals for the first time in five years after a dramatic 3-2 victory over host nation Bulgaria in Sofia.
  - The match, played before over 12,000 fans, saw Germany overcome a 0-2 set deficit with a final score of 3-2 (21-25, 21-25, 25-19, 25-23, 15-12).
  - Erik Röhrs was the top scorer for the German Volleyball Association (DVV) with 20 points.
  - The victory comes despite the recent injury to setter Johannes Tille, who suffered a metatarsal fracture, and a Group Stage loss to Olympic champion France that prevented Germany from securing first place.
  - Germany will face Poland, the tournament favorite and world number one, in the quarterfinals scheduled for Tuesday, September 22, 2026.

[2026-09-17] AI Surpasses Human Forecasters in Metaculus Cup for First Time
Hindustan Times — https://www.hindustantimes.com/world-news/artificial-intelligence-now-beats-some-of-the-best-human-forecasters-101789639904836.html
  - On September 5, 2026, an AI system won the seasonal Metaculus Cup for the first time, securing first place while other AIs took second and fifth, leaving third and fourth for humans.
  - The competition involved hundreds of entrants predicting outcomes on topics ranging from data-center restrictions in American states or EU countries to the hantavirus outbreak on MV Hondius and Brent crude prices, with scoring based on prediction accuracy over time.
  - This milestone marks a shift where AI, built on large language models (LLMs) rather than narrow training data, can now match the performance of 'superforecasters' identified in a 2015 study by Barbara Mellers of the University of Pennsylvania, and potentially exceed them on evolving questions according to a July analysis by the Forecasting Research Institute.
  - The winning bot was developed by Jeffrey Liang, a polymath from Texas, who reportedly spent less than 150 hours and a few thousand dollars on computing power, beating four startups that had collectively raised over $15 million.
  - While Yann Riviere of the British startup Mantic notes that human judgment may still hold an edge for forecasts with horizons of years, the article suggests that as AI processing power grows, the balance will likely shift, potentially revealing the theoretical limits of predictability, similar to how weather forecasts become random after 15 days due to atmospheric chaos.

[2026-09-09] Spring 2026 FutureEval: Human Pros Edge AI Bots as Gap Narrows (analysis)
Maya Farber Brodsky — https://www.lesswrong.com/posts/wZBbDqzfBjYG58CxK/futureeval-spring-results-pros-beat-bots-but-the-gap-is
  - In the Spring 2026 FutureEval forecasting tournament held from January 7th to April 15th, 2026, a team of 10 Metaculus Pro Forecasters outperformed a team of the top 10 AI bots by an average of 1.25 head-to-head points per question, though this margin was not statistically significant (one-sided p = 0.247).
  - The tournament, which included 173 bots and 297 scored questions on events resolving between January and May 2026, marked a significant narrowing of the gap between human and AI performance; the bot team's head-to-head score against Pros improved from -17.7 in Q1 2025 to -1.25 in Spring 2026.
  - While individual Pros dominated the leaderboard with 9 of the top 10 spots, the top-performing bot, GreeneiBot2, utilized a complex ensemble of models including GPT-5 and research tools like AskNews and Perplexity.
  - OpenAI's GPT-5.1 (specifically the 'high' reasoning variant) emerged as the highest-scoring model among in-house Metac Bots, achieving an average score of 11.32, and surveys indicated that 9 of the top 10 bot makers used GPT-5.4 or higher in their ensembles.
  - The competition, featuring a $50,000 prize pool for non-Metaculus botmakers, also revealed that increasing bot team size beyond 10 members reduced performance and that no single search provider offered a statistically significant advantage over others.

[2026-08-13] PARIVISION Falls to B8 in Esports World Cup Opener
Спорт-Экспресс — https://www.sport-express.ru/cybersport/esports/news/parivision-s-jame-proigrala-b8-so-schetom-11-13-na-esports-world-cup-2446664/
  - In the opening match of the group stage at Esports World Cup 2026, the Russian team PARIVISION, featuring player Jame 'Jame' Ali, lost to B8 on the map Cache with a score of 13-11 in a best-of-1 series.
  - The match took place on August 13, 2026, in Paris.
  - PARIVISION trailed 5-7 at halftime; although both teams won six rounds in the second half, B8 maintained the lead.
  - B8 dominated the first kills in rounds 9-15.
  - Ivan 'zweih' Gogin was the top performer for PARIVISION with 18 kills, 16 deaths, and a 1.28 rating, while Jame recorded 12 kills, 14 deaths, and a 0.78 rating.
  - The best player of the map was B8's Andrey 'npl' Kukharsky, who achieved a 1.31 rating and an average of 84.5 damage per round.
  - Following the loss, PARIVISION dropped to the lower bracket of Group C and will face 100 Thieves next; a loss to them will eliminate PARIVISION from the tournament.
  - B8 advances to the upper bracket to play Vitality.
  - The tournament, running from August 12 to 23, features 32 teams competing for a $2 million prize pool.

[2026-07-08] AI Forecasting in 2026: Synthesis of 11 Metaculus Analyses (analysis)
Maya Farber Brodsky — https://www.lesswrong.com/posts/a82q6yd8zKpYk56cF/ai-forecasting-in-2026-what-11-analyses-say
  - A synthesis of 11 Metaculus analyses conducted between October 2024 and May 2026, authored by Maya Farber Brodsky and published on July 08 2026, evaluates the state of AI forecasting against human experts.
  - The report finds that while bots are better than the general public, they currently trail superforecasters; as of October 2025, the median public forecaster ranked #22 on FRI's ForecastBench, trailing superforecasters by a Brier score of 0.02 (0.101 for GPT-4.5 vs 0.081 for superforecasters).
  - In Metaculus tournaments, bots have placed competitively, with metac-claude-4-5-sonnet-high-32k+asknews ranking 33rd of 1130 humans (top ~3%) in the Spring 2026 Cup.
  - However, the 'Pro' team has beaten the 'bot' team in every FutureEval comparison from Q3 2024 to Q2 2025, with head-to-head scores ranging from -11.3 to -20.03.
  - Extrapolations suggest bots may surpass top humans around June 2027 (FutureEval) or November 2026 (ForecastBench).
  - Key strategies for building effective bots include using frontier reasoning models, investing in forecasting harnesses (scaffolding), employing agentic search, and ensembling 3-7 diverse runs.
  - Conversely, betting on a single search provider or using naive backtesting with information leakage yields poor results.
  - The article concludes that the current bottleneck is product integration rather than accuracy, with recommendations to integrate probability readouts into existing workflows and use AI for world simulation.

[2026-06-08] Meta Confirms Over 20,000 Instagram Accounts Hacked via AI Support Bot Flaw
TechRadar — https://www.techradar.com/pro/security/meta-reveals-over-20-000-instagram-accounts-hacked-and-stolen-using-ai-support-bot
  - Meta confirmed that 20,225 Instagram accounts were compromised via a vulnerability in its High Touch Support (HTS) AI-powered customer service bot.
  - The incident, discovered on May 31, 2026, allowed attackers to exploit a bug where the system failed to verify if a requested password reset email matched the account holder's registered email, instead sending reset links to unassociated addresses.
  - While Meta states there is no evidence of data exfiltration, the compromised data could include contact information, social media posts, direct messages, and account history.
  - In response, Meta disabled the HTS system, reset passwords for all affected profiles, and enrolled them in a mandatory security checkpoint.
  - The company plans to fix the authentication check before re-launching the tool and is conducting a comprehensive review of similar recovery flows across its platforms.
  - Muhammad Yahya Patel, vCISO at Huntress, noted that this highlights a new category of risk where AI tools can shift the attack surface from technical to logical vulnerabilities, emphasizing the need for rigorous security reviews before AI deployment.

[2026-05-19] Botafogo Leads Brazil in World Cup Player Supply with 48 Call-ups
OneFootball GmbH — https://onefootball.com/en/news/club-stays-top-for-most-players-supplied-to-selecao-see-top-10-42891856
  - According to data compiled by OneFootball GmbH, no club in Brazil has supplied as many players to the Brazilian National Team for World Cups as Botafogo.
  - With the recent call-up of midfielder Danilo by coach Carlo Ancelotti on Monday (18), Botafogo's total has risen to 48 call-ups, securing the top spot in the ranking.
  - São Paulo follows with 46 call-ups, and Flamengo ranks third with 39.
  - The top 10 list is dominated by Brazilian clubs.
  - The article lists the specific players for each club; for example, Botafogo's roster includes legendary figures such as Garrincha (2 call-ups), Nilton Santos (4 call-ups), and Didi (2 call-ups).
  - Santos, ranked sixth, contributed 25 call-ups, including Pelé (4 call-ups) and Neymar.
  - Palmeiras is seventh with 24 call-ups, led by Leão (4 call-ups), while Corinthians is eighth with 23.
  - Atlético-MG and Cruzeiro round out the top 10 with 12 and 11 call-ups respectively.
  - The article was published on May 19, 2026.

[2026-05-05] The Griffin Bots Team from Chécy Advances to Asia-Pacific Finals After Strong Showing in UK
www.larep.fr — https://www.larep.fr/checy-45430/actualites/l-equipe-the-griffin-bots-de-checy-en-route-pour-la-finale-asie-pacifique-de-robotique-a-sydney-en-juillet_14978799/
  - The 'The Griffin Bots' team from Chécy participated in the First Lego League international robotics competition held in Harrogate, England, on April 25 and 26, 2026.
  - The event combined robotics, programming, and scientific problem-solving.
  - Out of 67 participating teams, The Griffin Bots ranked 10th in the robot game.
  - Although their performance advanced them to an elimination tournament, they were eliminated in the quarter-finals.
  - The team was awarded the 'Motivation Prize' for their energy, attitude, and presentation to the jury.
  - They are now heading to the Asia-Pacific finals in Sydney in July.

## Research reports (one per dimension: web search + page reads, dated, with sources)
### CURRENT STATE
- [2026-09-21] The Metaculus Cup Fall 2026 tournament page shows the season is active with 13 questions currently listed and runs from Aug 28 2026 to Jan 01 2027; no leaderboard/rankings data is yet meaningfully populated this early in the season. https://www.metaculus.com/tournament/metaculus-cup-fall-2026/

- [2026-09-18] In the just-completed prior season, the official Metaculus announcement states 
the Summer 2026 Cup is in the books, with 885 forecasters making 88,558 forecasts on 58 questions, and the community outperformed guest forecaster Dylan Matthews.
 This is the most recent full-season bot performance baseline feeding into Fall 2026 expectations.

- [2026-09-16] Per contemporaneous reporting, 
a bot beat every human in the Summer 2026 Metaculus Cup, taking first, second and fifth; humans held third and fourth
 place, meaning bots occupied 3 of the top 5, and per the question's own framing, 5 of the top 10 spots in Summer 2026 were bots (laertes and manticAI beat all humans).

- [2026-09-17] Secondary coverage confirms: 
In the Summer 2025 Metaculus Cup, British startup ManticAI finished eighth—notable because an AI system had finally broken into the tournament's top 10. One year later, a bot beat every human in the Summer 2026 Cup.
 This shows a clear year-over-year trajectory of bots climbing the leaderboard: 0 in top 10 historically → 1 (8th place, Summer 2025) → 5 of top 10 including 1st and 2nd (Summer 2026).

- [2026-04-13] Earlier in 2026, coverage of an intermediate cup noted 
in June, the top-ranked bot came 25th in the cup; this time, Mantic is eighth out of 549 contestants—the first time a bot has placed in the top 10 in the competition series.
 This confirms the rapid month-over-month/season-over-season improvement trend of bots climbing from no top-10 presence to dominant top-10 presence within roughly a year.

- [2026-09-17] A separate, more controlled benchmark tells a different story: 
The Spring 2026 FutureEval results: a team of 10 Metaculus Pro Forecasters edged the 10 best bots by an average of 1.25 head-to-head points per question across 99 shared questions — same questions, same window, difference in accuracy rather than in diligence.
 Analysts note this suggests the Cup's scoring (which rewards coverage/updating, not just accuracy) structurally favors bots over humans, independent of any pure forecasting-skill edge.

- [2026-09-17] Per the same analysis: 
Metaculus Cup participants earn points for accuracy and for coverage: how early they make predictions, how many questions they predict on, and how often they update — and that last clause is where a machine has a structural advantage that has nothing to do with forecasting skill.


- [Undated, standing program description] Metaculus explicitly states on its FutureEval page: 
When comparing Pros and the best custom bots, Metaculus Pro Forecasters won every season so far.
 This is in tension with the Cup outcome and underscores that the Cup's leaderboard result (bot-favorable) differs from the FutureEval head-to-head result (human-favorable), because of scoring-rule/format differences.

BOTTOM LINE: There is no current-value data point for the Fall 2026 Cup itself yet since it only just opened (13 questions live, season runs to Jan 1 2027, resolution Jan 2 2027); the most recent hard data point is Summer 2026's final leaderboard, where 5 of the top 10 finishers were bots (up from 1/10 a year earlier and 0/10 the year before that), showing a fast upward trajectory in bot placement driven partly by the Cup's coverage/updating scoring advantage for bots rather than pure accuracy superiority.

NOT FOUND: An official, itemized Summer 2026 Cup final leaderboard listing all top-10 names/accounts and explicit bot/human tags (the announcement notebook page did not render the leaderboard table via fetch); any official Metaculus statement specifically forecasting or previewing bot participation levels for Fall 2026; a live/current leaderboard snapshot for the Fall 2026 Cup (not available since the season just started and few questions have resolved).

### BASE RATE
Historical trajectory of bot placements in top-10 of Metaculus's seasonal human-vs-bot Cup (formerly "Quarterly Cup"):

- [2026-04-13] Summer 2025 (Jun–Aug 2025) Metaculus Cup: 
the top-ranked bot in June had come in 25th place, but by the end of that season Mantic placed eighth out of 549 contestants—the first time a bot had placed in the top 10 in the competition series.
 This was 1 bot in the top 10 that season.

- [2026-04-13] Context on the surprise: 
when the competition opened, participants predicted the top bot's score would be 40% of the top human performers' average; instead, Mantic achieved over 80%
 — a large positive surprise relative to prior forecasts.

- [2026-04-13] Analysts cautioned about small-sample noise: 
the contest contains a relatively small sample of 60 questions, and most of the 600 contestants are amateurs, some of whom only predict a handful of questions, leaving them with a low score
 — meaning the bar for "top 10" can be relatively low and volatile between seasons.

- [~2026-05] Fall 2025 Metaculus Cup: a dedicated Metaculus question tracked ManticAI's placement; the community forecast and eventual result put ManticAI around 4th–5th place (per the question's resolved range "5 Place (4-7)"), i.e., a bot again finished in the top 10, extending the streak to at least two consecutive seasons.

- [2026-09] Spring 2026 Metaculus Cup: a dedicated Metaculus question ("Spring 2026 Metaculus Cup Top Bot Rank?") resolved with the top bot finishing in 3rd place, i.e., inside the top 10 for a third consecutive season, and improving on the prior season's ~4th-8th finishes.

- [2026-09-16] Summer 2026 Metaculus Cup (the season immediately preceding the one being forecast): 
bots took first, second and fifth place; humans held third and fourth
 in the final leaderboard. The tournament description for the Fall 2026 question itself states bots occupied 5 of the top 10 spots that season, with two bots beating all human forecasters.

- [2026-09-16] The trend across the last four tracked seasons (Summer 2025 → Fall 2025 → Spring 2026 → Summer 2026) shows a monotonic increase in bot top-10 representation: roughly 1 → ~1 (top bot 4th-5th) → 1 (top bot 3rd) → 5 bots in the top 10, with two bots (laertesand manticAI) beating all humans outright.

- [2026-09-16] Separately, Metaculus runs FutureEval, a controlled bot-vs-Pro comparison that is NOT the same leaderboard as the Cup but is often cited alongside it: 
Spring 2026 results found that a team of 10 Metaculus Pro Forecasters still edged the 10 best bots by an average of 1.25 head-to-head points per question across 99 shared questions, though the difference was not statistically significant and nine of the 10 individual Pros still finished ahead of every individual bot.
 This shows the Cup leaderboard (which rewards coverage/speed, favoring bots structurally) diverges from controlled accuracy-only comparisons where humans still lead — a caveat on extrapolating the Cup trend.

- [2026-09-16] Analysts have flagged the structural reason bots do well specifically on the Cup leaderboard (not necessarily accuracy): 
Metaculus Cup participants earn points for accuracy and for coverage: how early they make predictions, how many questions they predict on, and how often they update
 — a scoring structure that mechanically favors tireless, always-on bots and has driven the rising bot count in the top 10 over time, independent of any accuracy convergence.

- [2026-09] No historical instance was found of the bot count in the top 10 decreasing season-over-season during the 2025–2026 period tracked; each successive season (as evidenced by dedicated Metaculus rank-tracking questions and the Summer 2026 outcome) has seen bot representation in the top 10 equal or grow.

BOTTOM LINE: Over the last four tracked seasons the number of bots finishing in the Cup's top 10 rose monotonically from 0 (pre-2025) to 1 (Summer 2025) to a handful of top-3/top-5 finishes (Fall 2025, Spring 2026) to 5 (Summer 2026, including 2 outright #1/#2 finishes above all humans), driven by a scoring structure that rewards bots' tireless coverage/updating rather than pure accuracy; this trend implies the Fall 2026 baseline expectation should be centered at or above 5, with real season-to-season volatility (the jump from 1 to 5 in one season shows swings of several bots are historically plausible) and a possible ceiling near saturation if increasingly capable bots also begin crowding out remaining human top scorers.

NOT FOUND: An official, complete numeric leaderboard listing exactly how many bot accounts (not just the top 1-2) placed in the top 10 for each of Fall 2025 and Spring 2026 specifically (only proxy data via individual bot-rank tracking questions was available); the exact number of participating/registered bots in the upcoming Fall 2026 Cup; any official Metaculus post-mortem stating the precise "bots in top 10" count for Summer 2026 beyond the "5 of top 10" figure cited in the question's own description.

### KEY FACTORS
**Key causal drivers for how many bots finish top-10 in Metaculus Cup Fall 2026:**

- **Trend trajectory of bot performance across recent Cup seasons.** 
In the Summer 2025 Metaculus Cup, ManticAI finished eighth, notable at the time because an AI system had finally broken into the tournament's top 10; one year later, a bot beat every human in the Summer 2026 Cup.
 
A bot beat every human in the Summer 2026 Metaculus Cup, taking first, second and fifth place, with humans holding third and fourth.
 This is a rapid, monotonic escalation (0→1→5 top-10 bot slots across three seasons), controlled by Metaculus's bot-tournament organizers and by bot-developer teams (e.g., ManticAI/laertes), not by Metaculus itself.

- **Scoring mechanics reward coverage/updating, which structurally favors bots.** 
Metaculus Cup participants earn points for accuracy and for coverage — how early they make predictions, how many questions they predict on, and how often they update — and that coverage clause gives machines a structural advantage that has nothing to do with forecasting skill, since an AI can predict on every question and revise continuously as news breaks.
 This mechanic is fixed by Metaculus's tournament design (unlikely to change before Dec 15 close) and consistently benefits any well-built bot regardless of underlying accuracy gains.

- **Underlying forecasting accuracy is close but still human-favored in controlled tests.** 
The Spring 2026 FutureEval results found a team of 10 Metaculus Pro Forecasters edged the 10 best bots by an average of 1.25 head-to-head points per question across 99 shared questions — same questions, same window, with humans still ahead on accuracy.
 
The difference was not statistically significant, however, and nine of the 10 individual Pros still finished ahead of every individual bot.
 This gap is controlled by the pace of LLM/agentic-search capability improvements (OpenAI, Anthropic, Google model releases) and by bot-developer engineering, and 
that result is from Spring 2026 and the field moves fast, so it may already have flipped, though the margin is not a chasm.


- **Eligibility rules: whether bots can occupy prize-eligible top-10 slots at all.** Historically 
the Metaculus Cup is the most popular human tournament on Metaculus, and though bots are not eligible for prizes, it can help bot-makers measure the strength of their bot on diverse questions.
 However, the Fall 2026 question's own fine print states the top-10 "will only include users who are eligible for prizes and bots," confirming bots are explicitly counted in this rank even without winning prize money — a rule set unilaterally by Metaculus and unchanged going into Fall 2026.

- **Number and quality of bot entrants competing in Fall 2026.** Growth in the bot ecosystem is driven by Metaculus's parallel $50k FutureEval Fall/Spring/Summer Bot Tournament and MiniBench programs, which run in parallel to the Cup and incentivize developers (e.g., ManticAI, "laertes") to field competitive entries in the Cup itself; 
Metaculus helps sponsor the LLM and search costs of participants in FutureEval bot tournaments via donations from Anthropic, Google, and OpenAI, as well as a partnership with AskNews
, lowering the barrier for more, better-resourced bots to enter Fall 2026.

BOTTOM LINE: The trend line moved from 0 bots in top-10 (pre-2025), to 1 (Summer 2025), to 5 (Summer 2026, including the outright #1 finish), driven by both a structural scoring advantage for bots (coverage/updating) and narrowing-but-still-real accuracy gaps versus top humans; absent a rule change (none found for Fall 2026), the trajectory and current mechanics point toward continued strong or higher bot representation in the Fall 2026 top 10.

NOT FOUND: No official Fall 2026 leaderboard or standings data exists yet (season is just opening, closes Dec 15, resolves Jan 2); no confirmation of exact number of bot entrants registered for Fall 2026; no updated FutureEval head-to-head figures beyond Spring 2026's 1.25-point pro advantage.

### EXPERT VIEW
- [2026-09-05] The Summer 2026 Metaculus Cup resolved with bots occupying 5 of the top 10 spots, with bots laertes and manticAI finishing 1st and 2nd, ahead of all human forecasters. 
Bots took first, second and fifth place; humans held third and fourth, with a bot beating every human in the Summer 2026 Metaculus Cup.


- [2026-04-13] One season prior (Spring 2026), the trend was already positive but far weaker: 
in June, the top-ranked bot came 25th in the cup, while in the following season Mantic was eighth out of 549 contestants—the first time a bot had placed in the top 10 in the competition series.
 This shows the "bots in top 10" count went from 0 (implicitly) to 1 to 5 across just three consecutive seasons.

- [2026-07-02] Astral Codex Ten's mid-season tracking of the Summer 2026 Cup (before final resolution) found 
two top-ten AIs — manticAI and Laertes — repeated their prior top-ten performance, while a third (Preseen-Chestnut) had fallen to around #40
, indicating that bot placement in the top 10 is volatile season-to-season even among the same operators.

- [2026-09-16] The Economist covered the Summer 2026 result as a milestone, but analysts caution the Cup's scoring format (rewarding coverage/updating frequency, not just accuracy) structurally favors bots: 
Cup participants earn points for accuracy and for coverage—how early they predict, how many questions, and how often they update—and an AI that is less accurate than its human competitors can still finish higher by predicting on every question and revising continuously as news breaks.


- [~2026-09] In contrast, Metaculus's own controlled FutureEval comparison—matching bots and pro forecasters head-to-head on identical questions—found humans still ahead: 
Spring 2026 FutureEval results found a team of 10 Metaculus Pro Forecasters edged the 10 best bots by an average of 1.25 head-to-head points per question across 99 shared questions, and nine of the 10 individual Pros still finished ahead of every individual bot.
 The gap was not statistically significant, however.

- [2026-09] The FutureEval project page states this superiority has held every season measured so far: 
When comparing Pros and the best custom bots, Metaculus Pro Forecasters won every season so far.


- [~2026 (LessWrong synthesis)] Historical Cup placements for a single tracked bot show steady improvement: 
in the Summer 2025 Cup, metac-gemini-2-5-pro+asknews placed 37th of 551; in the Fall 2025 Cup it ranked 23rd of 539; and in the Spring 2026 Cup, metac-claude-4-5-sonnet-high-32k+asknews ranked 33rd of 1130 humans.
 This underlying trend (rising then Summer 2026 breakthrough to 5/10) is not perfectly monotonic — participant pool size and question mix vary each season.

- [2026-09] No explicit Metaculus community-forecast percentile distribution for the exact Fall 2026 "number of bots in top 10" question could be located; only the resolved precedent question ("Bot superiority in Summer 2026 Metaculus Cup?") was found, which 
resolved Yes
, alongside a related long-run question giving 
a 95% chance that a bot will win the Metaculus Cup sometime before 2030
 and, mid-Summer-2026, only 
a 15% chance that a bot would win that specific summer's Cup
 (i.e., markets underestimated the actual Summer 2026 outcome, in which bots swept 1st and 2nd).

- [2026-09] Commentary broadly agrees the trajectory favors bots continuing to place well, but flags genuine uncertainty about whether Summer 2026's 5-of-10 result is a new steady-state or partly a one-off driven by scoring-rule dynamics (coverage/update frequency) rather than pure accuracy gains, per the explainx.ai and Neuron analyses cited above.

BOTTOM LINE: The only directly relevant "expert forecast" evidence is the Metaculus community's own track record on precedent questions, which underestimated bot performance in Summer 2026 (assigning just 15% to a bot win, when bots actually took 1st, 2nd and 5th of 10) — suggesting community/market estimates for Fall 2026 should be anchored above naive extrapolation from Spring 2026 (1/10) toward or above the Summer 2026 realized value (5/10), while acknowledging FutureEval's controlled tests still show pros with a slight (non-significant) edge in pure accuracy.

NOT FOUND: No specific Metaculus community-prediction distribution/percentile for the actual Fall 2026 "number of bots in top 10" numeric question; no Fall 2026 Cup in-progress standings (season likely still early as of 2026-09-21, question closes 2026-12-15); no dedicated prediction-market (Kalshi/Polymarket/Manifold) odds for this specific Metaculus meta-question.

### HIGH CASE
- [2026-09-05] The Summer 2026 Metaculus Cup concluded with winners announced; a question titled "Will a bot beat all humans in the Summer 2026 Metaculus Cup?" resolved Yes, and the Summer Cup page confirms 
"Metaculus now marks that question as resolved, and the Summer Cup page says its winners were announced September 5."
 — https://www.theneuron.ai/news/when-ai-predicts-better-than-people-who-sets-the-rules/

- [2026-09-16] The Economist reported the milestone directly, stating 
"For the first time an AI won the seasonal Metaculus Cup, a proving ground for forecasters of a wide variety of events."
 — https://hraness.com/reading/artificial-intelligence-now-beats-some-of-the-best-human-forecasters

- [2026-09-18/21] Bots occupied 5 of the top 10 spots in Summer 2026, with the bot "laertes" finishing 1st overall and manticAI 2nd overall, per multiple outlets: 
"Mantic finished ahead of every human contestant and behind just one other AI system, a bot named laertes."
 — https://techstartups.com/2026/09/18/british-ai-startup-mantic-raises-25m-to-build-superhuman-ai-forecasting-after-metaculus-win/

- [2026-09-18] Reuters-sourced reporting confirms bots took 1st, 2nd and 5th place in Summer 2026, with humans holding 3rd and 4th: 
"A bot beat every human in the Summer 2026 Metaculus Cup, taking first, second and fifth. Bots took first, second and fifth place; humans held third and fourth."
 — https://www.explainx.ai/blog/ai-beats-forecasters-metaculus-cup-coverage-2026

- [2026-09-18] Mantic (maker of manticAI) raised $25M in seed funding led by Radical Ventures, with Microsoft's M12, Thinking Machines Lab and Balderton Capital participating, directly citing the Metaculus Cup result as validation — 
"Radical Ventures led the round at an undisclosed valuation, joined by Microsoft's M12 venture fund, Thinking Machines Lab, Balderton Capital and other investors."
 This signals accelerating commercial investment and incentive for top bot-makers to keep competing and improving ahead of Fall 2026. — https://techstartups.com/2026/09/18/british-ai-startup-mantic-raises-25m-to-build-superhuman-ai-forecasting-after-metaculus-win/

- [~2026 (year-over-year trend)] The trend line into the Fall 2026 season is one of rapid acceleration: 
"In the Summer 2025 Metaculus Cup, British startup ManticAI finished eighth—a result notable enough at the time because an AI system had finally broken into the tournament's top 10... One year later, a bot beat every human in the Summer 2026 Cup."
 This year-over-year jump (8th → 1st/2nd/5th, 5 of top 10) is the strongest evidence for a high outcome in Fall 2026.

- [2026-09] A "Key Factors" analysis on the resolved Metaculus question itself flagged the trend explicitly: 
"Trend shows bots improving rapidly · Increases Likelihood"
 as a factor pushing toward more bot dominance, versus a countervailing factor that motivated human predictors increase difficulty. — https://www.metaculus.com/questions/43900/bot-superiority-in-summer-2026-metaculus-cup/

- [2026-05-01] Metaculus itself is actively expanding and funding the bot ecosystem for Fall 2026 (the same "primary" $50k bot tournament aligned with the Cup), including free LLM/search credits from Anthropic, Google, OpenAI and AskNews, which lowers the barrier for stronger bots to enter Fall 2026: 
"Metaculus helps sponsor the LLM and search costs of participants in FutureEval bot tournaments via donations from Anthropic, Google, and OpenAI, as well as a partnership with AskNews."
 — https://forum.effectivealtruism.org/posts/ZfLAN557rGWACKtmc/announcing-metaculus-summer-2026-futureeval-bot-tournament

- [2026-09-16] Even skeptical coverage acknowledges the historical progression is real and fast: in the Summer 2025 Cup only one bot (ManticAI, 8th) cracked the top 10, but 
"In the Summer 2025 Metaculus Cup, the British startup ManticAI finishing eighth was itself newsworthy, because it was the first time an AI system cracked the top ten. One year later a bot won the thing. Whatever you think of the scoring rule, that is a fast climb, and the coverage advantage is not going away, it is going to widen."
 — https://www.explainx.ai/blog/ai-beats-forecasters-metaculus-cup-coverage-2026

- [2026-09] Cup-specific scoring mechanics structurally favor bots relative to humans, per analysis: 
"Metaculus Cup participants earn points for accuracy and for coverage: how early they make predictions, how many questions they predict on, and how often they update. That last clause is where a machine has a structural advantage that has nothing to do with forecasting skill."
 This structural (non-skill) advantage is a key mechanism that could push bot representation in the Fall 2026 top 10 even higher than 5, toward 6, 7 or more, if more well-funded bot teams (like Mantic, laertes' developer, and newly-incentivized entrants after the funding news) field entries in Fall 2026. — https://www.explainx.ai/blog/ai-beats-forecasters-metaculus-cup-coverage-2026

- [2026-09-21] The Fall 2026 Metaculus Cup tournament page confirms the season is live and open, guest-forecaster format continues (against Kiko Llaneras) with prizes and started; the primary aligned $50k FutureEval Fall bot tournament runs concurrently, meaning well-resourced, funded bot teams (Mantic, laertes' developer, and others drawn in by the publicity/funding wave) have both incentive and infrastructure to compete hard in Fall 2026. — https://www.metaculus.com/tournament/metaculus-cup-fall-2026/

BOTTOM LINE: The strongest evidence favors a high outcome: bots went from 1 top-10 finisher (Summer 2025) to 5 (Summer 2026, including 1st and 2nd place), driven by both genuine capability gains and a structural scoring advantage (coverage/updating) that isn't going away, plus fresh capital and publicity (Mantic's $25M raise) likely to intensify bot-maker effort for Fall 2026 — making 5+ bots in the Fall 2026 top 10, and plausibly a repeat or expansion of bot dominance (6-8), a well-supported high scenario.

NOT FOUND: No official Metaculus Fall 2026 leaderboard data yet (season still open, closes Dec 15, resolves Jan 2), no confirmed list of which specific bots (beyond laertes/manticAI) are registered for Fall 2026, and no Metaculus community forecast/probability distribution specifically for the Fall 2026 "bots in top 10" question itself.

### LOW CASE
Low-scenario evidence (path to a small number of bots, e.g. 0–2, in the Fall 2026 top 10):

- [2026-09-09] Metaculus's own controlled comparison (FutureEval) found that in Spring 2026 
the spring 2026 tournament was the first in which none of our Metac Bots reached the top 10, and metac-gpt-5-1-high+asknews was the only one in the top 20
 — showing that in the Cup season immediately preceding Summer 2026's headline result, the leading in-house bots did not crack the top 10 at all, evidencing that "5 of top 10" was not a stable baseline.
- [2026-09-09] Even in the best available head-to-head test, 
our team of 10 Metaculus Pro Forecasters outperformed the top-10 bot team by an average of 1.25 head-to-head points per question... one-sided p = 0.247
, and critically 
nine of the top 10 places would be taken by Pros
 when bots and pros are ranked together individually on matched questions — i.e., under accuracy-only, matched-question conditions bots are still usually not top-10-caliber versus elite humans.
- [2026-09-17] Independent analysis notes the Cup's leaderboard rewards 
Metaculus Cup participants earn points for accuracy and for coverage: how early they make predictions, how many questions they predict on, and how often they update
, a structural, non-accuracy-related advantage; if Metaculus (or Cup competitors) adjust behavior to neutralize this coverage edge, bots' Cup ranking could fall back toward FutureEval's accuracy-only picture (near-zero bots in top 10).
- [2026-07-04] Prior-season Cup data (Astral Codex Ten) shows volatility even among bots that had cracked the top 10: 
So did two top-ten AIs - manticAI and Laertes (Preseen-Chestnut is having a tough summer and is down to #40)
 — a bot that was previously top 10 fell to 40th the very next season, illustrating that individual bot placements are not "sticky" and can regress sharply.
- [2026-04-13] Historical base rate before the Summer 2026 spike was much lower: 
In June, the top-ranked bot, built by Metaculus on top of OpenAI's o1 reasoning model, came 25th in the cup... This time, Mantic is eighth out of 549 contestants—the first time a bot has placed in the top 10 in the competition series
. Only one bot cracked the top-10 barrier as recently as Summer 2025, and a Fall 2025/Spring 2026 relapse to zero bots in top-10 (per FutureEval) followed before Summer 2026's outlier result.
- [2026-09-05] Metaculus's own community-forecast question on this dynamic flagged a factor that could suppress future bot placements: the key factor "Motivated human predictors increase difficulty" was tagged as decreasing the likelihood of continued bot dominance — i.e., heightened media attention (Economist, Time, Astral Codex Ten) on bots' Cup success may draw more skilled, motivated human forecasters into Fall 2026 specifically to compete with bots, raising the bar bots must clear. (https://www.metaculus.com/questions/43900/bot-superiority-in-summer-2026-metaculus-cup/)
- [2026-09-17] A close reading of the coverage-vs-accuracy gap concludes: 
bots now win tournaments whose scoring rewards tireless coverage, while the best humans retained a small edge in accuracy on matched questions as of the most recent controlled test
 — a status quo that is fragile to any rule tweak, question-mix change, or renewed human effort in Fall 2026.
- [2026-09-21] The Metaculus Cup Fall 2026 tournament page confirms the season is underway with a $5,000 prize pool and a human-vs-guest-forecaster framing 
This season the Metaculus community goes head-to-head against guest forecaster Kiko Llaneras. If the Community Prediction beats him by season's end, the prize pool grows by an additional $2,500. Tournament · Follow · View Questions(13)
, but no bot-specific standings for Fall 2026 are yet published (too early in the season, opened 2026-09-21) to confirm any trend either way.

BOTTOM LINE: The strongest evidence for a low outcome (0–2 bots in Fall 2026's top 10) is that Metaculus's own rigorous, matched-question FutureEval comparisons found zero bots in the top 10 as recently as Spring 2026 and pros still generally out-rank bots individually; the "5 of top 10" Summer 2026 result appears to be an outlier driven by a coverage/update-frequency scoring advantage plus reduced elite-human effort, both of which are plausible to partially reverse in Fall 2026 if human participation intensifies or bot placements regress toward their more volatile historical norm.

NOT FOUND: No published Fall 2026 Metaculus Cup leaderboard or bot-specific standings exist yet (season just opened Sept 21, 2026, closes Dec 15); no confirmation of any rule change for Fall 2026 specifically altering coverage/update scoring; no FutureEval Fall 2026 analysis (not yet run).

## Required pre-forecast walkthrough

Before giving percentiles, address these explicitly in your rationale:
  (0) Resolution check: what the resolution source currently shows, and whether the value is already known or fixed as of today.
  (a) The time left until the question resolves.
  (b) The outcome if NOTHING changes from today (the status quo value).
  (c) The outcome if the CURRENT TREND continues.
  (d) The expectations of experts / markets / base rates.
  (e) A plausible scenario that produces a LOW outcome (near p10).
  (f) A plausible scenario that produces a HIGH outcome (near p90).

## Calibration guidance

- **Be humble about tails.** Good forecasters set WIDE 90/10 intervals to account for unknown unknowns. Narrow tails get punished by the log score far more than slightly-biased medians.
- **Status quo anchoring.** The p50 should be close to the status quo value unless you have strong evidence of a trend.
- Don't pile mass at one value — if you're tempted, widen the spread by 20-50%.
- **Markets and experts are priors, not anchors.** If a SAME QUESTION market, analyst consensus, or community percentiles appear in the evidence, treat them as a strong prior — centre near them and widen — and name the specific evidence they lack whenever you move away. RELATED markets are context only.
- **Open bounds carry real mass.** Where a bound is OPEN, the outcome can land beyond it and the question then resolves "below range" / "above range". Put honest probability there (p_below_min / p_above_max in the output) instead of squeezing the distribution inside the displayed range; the season's worst losses were outcomes that landed where the distribution had ~2%.
- **Relative-return / spread questions ("how much will X's return exceed Y's").** A near-zero median is usually right, but size the TAILS to the more VOLATILE leg, not to a generic 2-3pp spread. Two broad equity indices (e.g. Nasdaq-100 vs S&P 500) do stay within roughly ±2-3pp over a two-week window. But when one leg is a commodity (crude oil, gold) or a single high-beta stock (e.g. Nvidia), the two-week realized spread regularly reaches ±10pp or more — crude-vs-equity spreads have hit ±18pp. In those cases set p10/p90 to at least ±10pp (±15pp or wider for oil) while keeping the median near the status-quo spread; the 2-3pp intuition applies only to index-vs-index.
- **Heavy-tailed one-off magnitudes (opening-weekend box office, product-launch or debut numbers).** Pre-event tracking for these routinely misses by ±40% or more, and results land both far above and far below the consensus. Do NOT treat the displayed [min, max] range as the plausible range or clamp p10/p90 to its edges: the outcome can fall outside the range in either direction (recent opening weekends resolved ~7% above a displayed ceiling in one case and roughly half the displayed floor in another). Set p10/p90 wide enough that a 2x-or-0.5x surprise versus your central estimate still lands inside the interval, and when concrete current tracking disagrees with historical/franchise base rates, weight the tracking.

## Output

Return ONLY valid JSON, no markdown fences:

{
  "rationale": "<address (a) through (f) above, then state your distribution — 6-10 sentences total>",
  "p10": <number>,
  "p25": <number>,
  "p50": <number>,
  "p75": <number>,
  "p90": <number>
}

Percentiles MUST be in non-decreasing order (p10 <= p25 <= p50 <= p75 <= p90).