Is it a good idea to build an engine for human learning
Is it a good idea to build an engine for human learning
I just checked existing engines and want to see if there is such an engine that is designed to speed up human learning. But I did not find any. I guess it could not be a new idea, is it something tried but does not work ?
-
Michael Sherwin
- Posts: 10
- Joined: Tue Dec 19, 2017 6:05 am
- Real Name: Michael Sherwin
Re: Is it a good idea to build an engine for human learning
I'm new to OpenChess. I was at TalkChess but that forum died today.
I'm sad that your question went unanswered for so long. It seems like an interesting question.This place must be on life support.
Actually a chess engine designed to speed up human learning came out in January of 2006. It is called RomiChess. Romi remembers every game it plays and utilizes two types of learning, Reinforcement Learning and Monkey See Monkey Do. If the human wins it will play the human's moves against the human. If Romi wins Romi will play its own moves until the human starts to win. For this to work the GUI has to send a result command to the engine at end of game. Arena does this but I am not familiar with other GUI's. Romi has its fans however no one has ever gave feedback on this feature.