I will start a new series of blog posts covering some ideas related to General Artificial Intelligence (GAI).
Ever since the appearance of the first algorithms and models that were able to perform tasks seen as a proof of “intelligence” (playing chess, finding cats in a picture, etc.) there was a general expectation that in the next 10 years we will see General AI emerging. Every 10 years the goal posts seem to move and the goal of GAI appears to become more illusory.
In the next series of posts I will try to introduce a number of updates to the classical Reinforcement Learning (RL) theory that, in my opinion, might move us closer to GAI.
This page will be a table of contents for the posts:
- Intro: Classical Reinforcement Learning
- From RL to GRL:
- Redefining the Environment
- States and Observations
- Revisiting the GRL model