Before 2030, will an AI system be able to solve compositional problems of arbitrary depth?
2
100Ṁ352029
35%
chance
1H
6H
1D
1W
1M
ALL
Given a problem that requires repeated application of specific rules (such as multiplying two large numbers or solving towers of Hanoi), will any AI system be able to generalize to arbitrarily large compositional depth? A new paper by Apple shows that they cannot currently do this.

It still counts even if the system shows some performance drop off with increased depth. What I'm looking for here is performance that does not drop of exponentially past a certain depth.
It does not count if the AI system is doing this by writing code.
Update 2025-06-07 (PST) (AI summary of creator comment): - The AI system is allowed to use a scratch pad or external memory.
This question is managed and resolved by Manifold.
Get
1,000 to start trading!
Sort by:
People are also trading
Related questions
Will we have AI that can explain chess moves logic in human language by 2030?
94% chance
Will AI be able to create a video with this much detail by 2030?
92% chance
When will AIs be good at solving complex problems? (read description)
Meta-Learning Compositionality (MLC) in state of the art AI models by Oct. 2025?
19% chance
Will AI achieve superintelligence in music composition by 2026?
19% chance
On Dec 31, 2025, will a widely available AI model be able to write a sophisticated 2000 line program?
55% chance
Would an AI be capable of constructing a non-brute, ultraweak proof of chess for this board state by 2030?
49% chance
By 2030, will Ai picture generators be able to do basic geometry and projections?
91% chance
Will AI be able to describe the state of a chessboard from an image before 2026?
95% chance
Will AI be able to create art in a "human-like way" within Photoshop, GIMP, or Paint.NET by the end of 2025?
40% chance