Will a single model achieve superhuman performance on all Atari environments by 2025?
17
Ṁ1kṀ4.4kresolved Jan 6
Resolved
NO1H
6H
1D
1W
1M
ALL
Note: single *model*, not single *algorithm* (the latter has already been done). If I decide the model is actually many environment specific models stuck together (e.g. a hierarchical model that predicts current environment and delegates to a subagent) it will not count (a single model that *learns* a hierarchy without any inductive bias pushing it towards that *would* count, however).
This question is managed and resolved by Manifold.
Market context
Get 
1,000 to start trading!
1,000🏅 Top traders
| # | Trader | Total profit |
|---|---|---|
| 1 | Ṁ272 | |
| 2 | Ṁ224 | |
| 3 | Ṁ62 | |
| 4 | Ṁ46 | |
| 5 | Ṁ28 |
People are also trading
Related questions
Will an AI achieve >85% performance on the FrontierMath benchmark before 2028?
61% chance
Will an AI achieve >80% performance on the FrontierMath benchmark before 2027?
45% chance
Will models be able to do the work of an AI researcher/engineer before 2027?
25% chance
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?
92% chance
Will a human player reach level 255 in NES Tetris by 2027?
57% chance
Will World Model AIs start being used in games in early 2026 and quickly start dominating the Game Development industry?
7% chance
Will an AI achieve >85% performance on the FrontierMath benchmark before 2027?
35% chance
Will a robot demonstrate exceptional speed, versatility, and performance in extreme environments by 2040?
65% chance
When will any model achieve >=human performance on QuALITY?
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?
78% chance