Will AI image generating models score >= 90% on Winoground by June 1, 2025?
30
1kṀ2732
resolved Jun 2
Resolved
NO

Based on the recent discussions between Gary Marcus and @ScottAlexander, this market resolves as follows:

IF any AI image generating model (such as DALL-E, Imagen, Stable Diffusion, etc) scores 90% or more on the Winoground image composition test, this market resolves YES.

ELSE this market resolves NO.

Relevant Links:
Scott's post: https://astralcodexten.substack.com/p/i-won-my-three-year-ai-progress-bet
Gary's post: https://garymarcus.substack.com/p/did-googleai-just-snooker-one-of
Winoground: https://arxiv.org/pdf/2204.03162v2.pdf

AI
ACX
Get
Ṁ1,000
to start trading!

🏅 Top traders

#NameTotal profit
1Ṁ538
2Ṁ305
3Ṁ216
4Ṁ116
5Ṁ99

People are also trading

Will an AI achieve >85% performance on the FrontierMath benchmark before 2028?
72% chance
Will an AI image generation model successfully generate a proper chess board by July 31, 2025
11% chance
Will an AI score over 80% on FrontierMath Benchmark in 2025
17% chance
Will any AI model score above 95% on GRAB by the end of 2025?
40% chance
Will any AI model score >80% on Epoch's Frontier Math Benchmark in 2025?
15% chance
Will AI pass the Winograd schema challenge by the end of 2025?
86% chance
In what year will AI achieve a score of 95% or higher on the SWE-bench Verified benchmark?
10/7/28
Will any AI model achieve > 40% on Frontier Math before 2026?
70% chance
Will an AI model outperform 95% of Manifold users on accuracy before 2026?
46% chance
AI generates better alt text for 90% of test images by June 30, 2026?
65% chance
Sort by:

Looks like Manifold was too bullish!

bought Ṁ600 NO

@SneakySly Resolves NO

https://paperswithcode.com/sota/visual-reasoning-on-winoground

Best results seem to be around 75 %

This market's resolution seems pretty dependent on whether someone runs the benchmark with current frontier models

I like this question, thank you!

@vluzko =)

People are also trading

Will an AI achieve >85% performance on the FrontierMath benchmark before 2028?
+4% 1d72% chance
Will an AI image generation model successfully generate a proper chess board by July 31, 2025
11% chance
Will an AI score over 80% on FrontierMath Benchmark in 2025
17% chance
Will any AI model score above 95% on GRAB by the end of 2025?
40% chance
Will any AI model score >80% on Epoch's Frontier Math Benchmark in 2025?
15% chance
Will AI pass the Winograd schema challenge by the end of 2025?
86% chance
In what year will AI achieve a score of 95% or higher on the SWE-bench Verified benchmark?
10/7/28
Will any AI model achieve > 40% on Frontier Math before 2026?
70% chance
Will an AI model outperform 95% of Manifold users on accuracy before 2026?
46% chance
AI generates better alt text for 90% of test images by June 30, 2026?
65% chance
© Manifold Markets, Inc.•Terms•Privacy