68
votes
Accepted
Is "Adversarial Policies Beat Professional-Level Go AIs" simply wrong?
One of the authors of the paper here. It might shock you to find that I believe the paper was not "simply wrong". I do however think that we did a poor job explaining our evaluation setting, ...
- 726
20
votes
Is there an equivalent of Deep Blue in the Go world?
AlphaGo
Now there is AlphaGo by Deep Mind, a company recently bought by Google playing currently a match against 9p Lee Sedol. It is the Deep Blue of Go.
EDIT: The final result of the match of five ...
- 1,187
15
votes
Is "Adversarial Policies Beat Professional-Level Go AIs" simply wrong?
Regardless of the validity of this particular strategy, the general technique seems quite valid.
They did not simply discover a bug in KataGo and then write a program to exploit the bug. Rather, they ...
- 251
12
votes
Is there an AI that can build decks?
Don Goodman-Wilson made a program to build decks given a pool of cards. (I guess you could make the pool 60 copies of each card legal for the format).
From his post on Hackernoon:
So I started ...
- 10.4k
12
votes
Accepted
How much of AlphaGo's success is due to novelty?
I would argue that AlphaGo's advantage cannot be significantly attributed to the novelty of its moves.
The original public AlphaGo games were those against Lee Sedol, the second ranked player in the ...
- 78.3k
11
votes
Is "Adversarial Policies Beat Professional-Level Go AIs" simply wrong?
In the Tromp–Taylor rules, the concept of dead stones is left out—all stones on the board at the end of the game are treated as being alive. The KataGo AI was trained primarily using the Tromp–Taylor ...
- 404
10
votes
Accepted
How does AlphaGo calculate?
Here's the AlphaGo team's paper that has all of the details (behind a paywall): http://www.nature.com/nature/journal/v529/n7587/full/nature16961.html
I gave a couple of tech talks about this recently....
- 3,195
9
votes
Accepted
Is there an AI that can build decks?
First, just so you know where I’m coming from, I’m a professional programmer, and I’ve studied AI, but I don’t work with AI professionally, so there are limits to my expertise. I defer to anyone who ...
- 878
8
votes
Accepted
What was the effect of AlphaGo on Go?
I've read a couple of articles discussing the impact of AI on professional Go, the second one being a reply to the first one:
Impact of Go AI on the professional Go world
Impact of Go AI on the ...
- 196
8
votes
Is there a strongly solved game with infinite possible board arrangements?
Yes, with Nim being the best-known example
You ask for a strongly solved game (presumably referring to the term in combinatorial game theory). According to Wikipedia, games that are strongly solved
...
- 8,540
8
votes
Is Risk a computationally difficult game
Disclaimer: I have some programming background but haven't ever actually tried to make a Risk AI, so this isn't 100% definitive.
Although making a "perfect" Risk AI might be, it seems like it wouldn'...
- 667
7
votes
Strongest cribbage bots
I wrote the AI engine for BTO Cribbage, a mobile Cribbage app. I played Cribbage growing up and decided to write my own app after playing the other apps. I have played 10+ other apps and most of them ...
- 209
6
votes
Accepted
By how many points did alphago win against Lee Se-dol?
The DeepMind channel on Youtube has a short review of each game by Michael Redmond 9P.
At the end of the summary for game 1, Michael estimates that black is slightly ahead on the board. When you ...
- 1,041
6
votes
Accepted
Unbeatable Nine Men's Morris AI?
Yes, Gábor E. Gévay et al. have made such an engine.
The program and instructions for downloading the databases necessary for perfect play can be found here.
- 10.4k
5
votes
Is there an AI that can build decks?
You will need a good playing-algorithm.
The most straightforward way is by running a tournament of sampled decks:
Sample decks (randomly / heuristically / using some data)
Play many games using ...
- 3,839
5
votes
Is "Adversarial Policies Beat Professional-Level Go AIs" simply wrong?
I think you're right - this isn't much of a weakness in KataGo. It looks like there's a mismatch between the way KataGo plays, and the scoring method they're using. In Tromp-Taylor scoring, White ...
- 3,195
4
votes
Accepted
What progress has been made in computerized bridge play?
There's a fundamental problem that the rules of bridge are not well-defined for computers. Specifically, the rules require that partnerships communicate their agreements - including implicit ...
- 1,936
3
votes
AI that plays Quoridor?
I made an AI agent playing Quoridor. You can play against it right on the browser here: https://gorisanson.github.io/quoridor-ai/. As you can read on the "about" section on the page, I imitated the ...
- 131
3
votes
What activity is there in Computer Go tournaments?
The Computer Go Server1 is currently a very active test bed for bots. The Computer Go page at Sensei’s Library refers, under Competitions, to various competitions including the Computer Go Server, ...
- 1,328
3
votes
Is this double ko played reasonably by AlphaGo Zero and AlphaGo Lee?
There is a difference between the two semi-stable double ko positions, although a rather small one. C14 reduces the immediate number of liberties of the G13 leg. This means that black will have a free ...
- 533
3
votes
Is Risk a computationally difficult game
Yes and no. Every AI will have strategies that it uses, and some will be better than others. The combination of that and how many turns deep you're willing to execute your AI to decide on a move, and ...
- 10.8k
3
votes
Is there a strongly solved game with infinite possible board arrangements?
A non-Nim example is one usually presented as a puzzle, such as this Puzzling.SE question. The rules of the game are:
Given a symmetrical (normally circular or square) table, two players take turns ...
- 10.7k
3
votes
Accepted
Ranking of Go Software
Obviously, AlphaGo is the top AI now. Besides it, a simple way to find information about Computer Go Ranking would be to search for :
Computer go tournament
Here is a list of Go AI from wikipedia
...
- 465
3
votes
By how many points did alphago win against Lee Se-dol?
You can download the game record (.sgf) file from this link and use an score estimator as in the KGS Goban (free), which will give you W+2.5 (including komi) and by looking at the screen we can see ...
- 131
3
votes
Accepted
Where can I find 'no limits' Texas Hold'em game logs?
Building a Poker Playing Agent based on Game Logs using Supervised Learning
I found this paper on creating a Hold'em AI that refers to two external log files on page 100 of that paper. One no longer ...
- 4,545
2
votes
Accepted
Has Hive already been completely 'solved' for AI?
No, just the total number of placements is far too huge to "solve", and that doesn't even start to take into account movements occurring between placements.
- 8,358
2
votes
Accepted
Changing bot difficulty in Dominion Online
No, there is not currently any difficulty settings for the AI.
I cannot find an official source stating this to quote/link as reference, but you can read through the official FAQ and see some ...
- 72.1k
2
votes
Accepted
Overplay in orthodox fuseki - Leela AI fails miserably?
W1 should be N11, then black dies, I'd claim from memory.
W1 at R13 is probably not a complete disaster in itself, but O12 should then be O11, or maybe P10 should be O10. That would force black to ...
- 6,902
2
votes
How does AlphaGo calculate?
AlphaGo relies heavily on Monte-Carlo Tree Search (MCTS), which is a form of decision making that utilizes random choices as an analog for creativity. Specifically, it allows the algorithm to "think" ...
- 531
2
votes
What is the highest game score theoretically possible in Scrabble using 1 player and the International Scrabble Dictionary?
This is an open problem that hasn't been resolved yet.
That said, the highest known possible score is in the vicinity of 4000 points. See page 69 of Total Scrabble.
- 12.5k
Only top scored, non community-wiki answers of a minimum length are eligible