Superhuman News

Superhuman mathematical problem solving before 2030, assuming no AGI yet?

Apr 18, 9:11 PMDec 31, 11:59 PM

31.06%chance

109171214

OptionVotes

YES

81796

26531

Will AI be capable of superhuman persuasion well before (>1yr) superhuman general intelligence?

Oct 25, 8:59 PMJan 1, 4:59 AM

64.13%chance

888994

OptionVotes

YES

1087

819

AI model achieves superhuman ELO on Codeforces by June 1st 2027?

Nov 27, 2:04 PMMay 31, 11:59 PM

85.38%chance

323561

OptionVotes

YES

401

128

Will an AI model be capable of superhuman persuasion before 2034.

Dec 31, 3:39 PMDec 9, 11:59 PM

85%chance

372397

OptionVotes

YES

2065

657

Will AI be superhuman at MTG rules by the end of 2030?

Feb 21, 6:12 PMJan 2, 7:59 AM

83.01%chance

161721

OptionVotes

YES

2210

452

Benchmark Gap #3: Once a model achieves superhuman performance on a competitive programming benchmark, will it be less than 2 years before there are "entry level" AI programmers in industry use?

Apr 8, 6:06 PMJan 1, 4:59 AM

78%chance

16943

OptionVotes

YES

1382

859

Superhuman math life-changing for muggles?

Sep 18, 9:23 PMDec 31, 10:14 PM

76.2%chance

6881

OptionVotes

YES

1789

559

Sometime before 2031, AI is going to cross a generality threshold that allows for recursive self-improvement, and then FOOM to superhuman capability. Before that point humanity must solve AI safety or be helpless.

Dec 24, 3:53 AMFeb 1, 7:59 AM

28.65%chance

18812

OptionVotes

YES

1010

919

By EOY 2034, what unexpected views of math will be changed by (superhuman) AI?

Mar 21, 9:42 PMJan 1, 4:59 AM

20717

OptionProbability

At least one of the six left Millennium problems is found with a counter-example

At least one of the six left Millennium problems is undecidable

A unified framework for gravitation and quantum theory is proposed that is not string theory

Neural Networks are explainable and can be trained without gradient descent with better performance.

No AI will be powerful enough to directly (as a necessary condition) lead to a significant discovery in math.

Shinichi Mochizuki is right or almost right about abc conjecture

P vs. NP will be trivial (i.e., provable using simple methods taught in undergrad classes, e.g., with a counter-example)

Once a model achieves superhuman performance on a competitive programming benchmark, will it be less than 5 years before there are "entry level" AI programmers in industry use?

Apr 8, 6:37 PMJan 1, 5:00 AM

91.85%chance

16610

OptionVotes

YES

1169

959

Will LLMs be worse than human level at forecasting when they are superhuman at most things?

Feb 22, 8:18 PMJan 2, 4:59 AM

40.85%chance

15486

OptionVotes

YES

982

957

Will GPT-5 be capable of achieving superhuman performance in at least one exam that is typically taken by humans?

May 8, 5:14 PMJan 1, 2:59 PM

90.55%chance

9339

OptionVotes

YES

465

123

Superhuman News

Prediction markets for Superhuman

Prediction markets for Superhuman

Prediction markets for Superhuman

Superhuman mathematical problem solving before 2030, assuming no AGI yet?

Will AI be capable of superhuman persuasion well before (>1yr) superhuman general intelligence?

AI model achieves superhuman ELO on Codeforces by June 1st 2027?

Will an AI model be capable of superhuman persuasion before 2034.

Will AI be superhuman at MTG rules by the end of 2030?

Benchmark Gap #3: Once a model achieves superhuman performance on a competitive programming benchmark, will it be less than 2 years before there are "entry level" AI programmers in industry use?

Superhuman math life-changing for muggles?

Sometime before 2031, AI is going to cross a generality threshold that allows for recursive self-improvement, and then FOOM to superhuman capability. Before that point humanity must solve AI safety or be helpless.

By EOY 2034, what unexpected views of math will be changed by (superhuman) AI?

Once a model achieves superhuman performance on a competitive programming benchmark, will it be less than 5 years before there are "entry level" AI programmers in industry use?

Will LLMs be worse than human level at forecasting when they are superhuman at most things?

Will GPT-5 be capable of achieving superhuman performance in at least one exam that is typically taken by humans?

Latest stories