Red Queen hypothesis: a new way forward for self-improving AI
Red queen hypothesis – a new way forward for self-improving AI
Researchers at the University of Cambridge propose applying the Red Queen hypothesis to AI, suggesting that self-improving systems must continuously adapt and evolve to keep pace with ever-changing challenges. This new framework could lead to more robust and resilient AI.
The Red Queen hypothesis offers a new way forward for self-improving AI by suggesting that systems must constantly evolve just to maintain their performance.
- robotresearcher
Here’s a paper by Floreano at EPFL from 1997 explicitly on Red Queen dynamics for creating neural networks for intelligent robot control.
There was lots of discussion of these ideas in the 1990s. In those days we trained very small NNs - tens of nodes - by evolving their weights and topologies. A run could take days on a workstation of the time.
This particular paper is about co-evolving predator and prey, where the behavior of each is the ‘evaluation’ of the other.
https://infoscience.epfl.ch/entities/publication/a65d0679-68...
- throwa356262
Have not read the paper yet, but this not sound like GAN applied to agent training?
- richardfey
> "Instead of improving an agent against a fixed test, we let the evaluation evolve alongside the agent"
This quote should have been highlighted earlier in the article.