User talk:Glibjibb/sandbox
Peer Review for "General game playing (GGP)"
This is a really interesting article, and it can definitely benefit from the addition of more details. The main question I have is how well it needs to be able to play the games and how it learns the rules. I.e. can you simply setup the program to be able to "read" the state of the game and judge how it should play? There's a good set of references to start with, I just think a more technical or at least a more detailed explanation of the core concept would help the page.
I guess the only other thing you could add is the uses of results of certain programs. For instance, the article mentions how this type of AI could be used to test games, but why else is it being made?
NaTaHu (talk) 21:54, 25 March 2016 (UTC)NaTaHu
Peer Review[edit]
Major points[edit]
- Most of the sources seem non-academic
- It would be good to expand on the Stanford project. What is its purpose? Based on the 'Other approaches' heading, I expected the Stanford project to be an GGP AI in and of itself.
- Since this is an article about AI, I would be interested in some more detail about how these AI work.
- Images are always a plus!
Minor Points[edit]
- I think there should be a citation after the quote in the GVGP section.
Pickyt (talk) 12:41, 26 March 2016 (UTC)
Peer Review round 2[edit]
Major Points[edit]
- The section on the Stanford Project does not explain the nature of the Stanford Project very well
- The rules/criteria of the described competition are unclear, and only the winners are listed
- The second section is called "Other approaches", but I am not sure what the goal of any approach is
Minor Points[edit]
- I feel like the end of the first paragraph is a little out of place, as it shifs from games to real life very quickly
- The last section should have a different name