Chess

[Rob Lister]

Now, having wasted tons of my precious time watching 12 very dull draws and 3 quick but still dull wins, I have a new and more betterestlyer source

[Rob Lister]

3-year-old prodigy.



TLDW; He lost. He was playing the 2nd ranked GM Anatoly Karpov, after all. Then he started crying and begging for his mommy. Anatoly felt horrible.

[Anaxagoras]

Ha! What a baby!



More seriously though, I worry about a kid that young being labeled as a "prodigy". It might not end well. There's a bit of a history with chess prodigies ending up mentally ill. Not saying that's what causes them to become mentally ill. That was probably going to happen any way. Bobby Fischer was the most famous example, but there have been others. Peter Winston was another.

And here's a story I found about little Mischa:

https://www.theringer.com/sports/2017/1 ... by-fischer

[Doctor X]

https://media1.tenor.com/images/3811655 ... /tenor.gif

--J.D.

[Rob Lister]

This is new. Alphazero vs Stockfish.

Out-fucking-rageous.



While Alpha was kicking stockfish's ass, it also wins protein folding competition
https://deepmind.com/blog/alphafold/

[Rob Lister]

It strikes me that Alpfazero is just plain smarter than us, in every respect. At chess or any other game, it can't be denied. Are we looking at the next evolutionary jump?

It's hard to say no.

Say no.

I double-dog dare ya.

[Anaxagoras]

It strikes me that Alpfazero is just plain smarter than us, in every respect. At chess or any other game, it can't be denied. Are we looking at the next evolutionary jump?

It's hard to say no.

Say no.

I double-dog dare ya.

I'm not sure that it knows how to think for itself. Which is probably a good thing. We have to tell it what questions to answer. It's good at finding answers, once given a defined goal, but perhaps not at asking questions. :notsure:

[Anaxagoras]

Art is perhaps another field where it cannot surpass us.

Could it write a better symphony that Mozart's best or Beethoven? Perhaps it could, but I haven't seen it yet. I'm not sure they would know how to make it do that. With chess, it's easier to define the parameters and the goal.

[Anaxagoras]

No, but I assume that they are one of "us" so it's not about what the average human can do, but what the smartest among us can do.

[Witness]

Could certainly compose excellent dodecaphonic music, as it has (* cough *) "rules".

There is a hilarious vid on YT (but the blather is in French) where a pianist plays that kind of stuff, sometimes shifting one of his hands (and not the other) a semitone up or down, and the audience has to guess if he's playing the original composition or the butchered version.

You hear simply no difference, random remains random.

[shemp]

It strikes me that Alpfazero is just plain smarter than us, in every respect. At chess or any other game, it can't be denied. Are we looking at the next evolutionary jump?

It's hard to say no.

Say no.

I double-dog dare ya.

I'm not sure that it knows how to think for itself. Which is probably a good thing. We have to tell it what questions to answer. It's good at finding answers, once given a defined goal, but perhaps not at asking questions. :notsure:

Once it starts arguing with humans over the playing venue, the cameras and the prize fund, then we'll know it's sentient.

[Anaxagoras]

Maybe they just haven't made the "rules" specific enough. Maybe it doesn't know how to evaluate its own work and find ways to improve it. Maybe it just needs more practice. In learning to play chess, it plays games over and over and over until it gets good. In the beginning it is very bad.

[Anaxagoras]

Jerry also has a new video of Alpha Zero vs Stockfish.

[Anaxagoras]

It strikes me that Alpfazero is just plain smarter than us, in every respect. At chess or any other game, it can't be denied. Are we looking at the next evolutionary jump?

It's hard to say no.

Say no.

I double-dog dare ya.

Here's a thought. This seems like something that an AI could do in principle, and it doesn't even need to have free will or be able to think "like a human" necessarily to do it:

Can it do a decent translation from one language to another? As good as a good human translator. There are of course plenty of terrible translators out there, but I would like to know when AlphaZero can translate say a novel from Japanese to English or vice versa. It would take a lot of work, but it seems like the kind of thing that an AI like AlphaZero could do, if it had the right kind of data to learn from. Can it teach itself to do it? I sort of doubt it, but maybe a clever person could figure that out. Last time I checked, Google Translate still kinda sucked for Japanese to English. I check every once in a while. I haven't tried other languages. Maybe someone can try German to English or French to English or Russian to English and see how well it does.

[xouper]

Here's a thought. This seems like something that an AI could do in principle, and it doesn't even need to have free will or be able to think "like a human" necessarily to do it:

Can it do a decent translation from one language to another? As good as a good human translator. There are of course plenty of terrible translators out there, but I would like to know when AlphaZero can translate say a novel from Japanese to English or vice versa. It would take a lot of work, but it seems like the kind of thing that an AI like AlphaZero could do, if it had the right kind of data to learn from. Can it teach itself to do it? I sort of doubt it, but maybe a clever person could figure that out. Last time I checked, Google Translate still kinda sucked for Japanese to English. I check every once in a while. I haven't tried other languages. Maybe someone can try German to English or French to English or Russian to English and see how well it does.

That's a great thought.

I don't know enough about AlphaZero to comment about its abilities, but I can point out that in general, there is a bigly and fundamental difference between AI learning to play chess versus translating languages. And that is the feedback to know how well you did. With chess, the AI can play against itself (or anything else) and evaluate the results without any human intervention. It is in essence self correcting with enough experience. It cannot do that with translating languages since there is no similar feedback mechanism without human intervention. Or to put it another way, with chess, it can create its own knowledge how best to play the game. It cannot do that with translating languages, since it needs humans to tell it how well it did.

On a related note, I have a hypothesis about why AI does not (currently) seem to do well in classifying music genres. It depends on examples of human classifications to learn from. My conjecture is that the AI classifications are probably very good, but that the human classifications are faulty, and that might explain why AI doesn't seem to do well. For example, the AI is given a largish database of music which is pre-labeled by genre according to humans. The AI then tries to figure out the rules of classification from that database. My hypothesis is that humans are not very good at that, and thus the AI is hamstrung before it even gets started.

Getting back to the chess example, the AI can create its own database of winning and losing moves. With translating languages, it cannot create its own database of good and bad translations. At least not with the current state of the art, as I understand it.

Or to point out one more obvious difference, evaluating the results of a chess game is largely objective, compared to language translation which has a large subjective component.

I could be wrong. But it's an interesting question.

[Witness]

On a related note, I have a hypothesis about why AI does not (currently) seem to do well in classifying music genres. It depends on examples of human classifications to learn from. My conjecture is that the AI classifications are probably very good, but that the human classifications are faulty, and that might explain why AI doesn't seem to do well. For example, the AI is given a largish database of music which is pre-labeled by genre according to humans. The AI then tries to figure out the rules of classification from that database. My hypothesis is that humans are not very good at that, and thus the AI is hamstrung before it even gets started.

Interesting.

Do you have a source for AI being bad at classifying music?

Re. your hypothesis, last sentence especially, it doesn't square with the idea I had (from old reads) that neural networks were very good at fishing signal out of noise.
But then AI and NN aren't the same thing.

[xouper]

On a related note, I have a hypothesis about why AI does not (currently) seem to do well in classifying music genres. It depends on examples of human classifications to learn from. My conjecture is that the AI classifications are probably very good, but that the human classifications are faulty, and that might explain why AI doesn't seem to do well. For example, the AI is given a largish database of music which is pre-labeled by genre according to humans. The AI then tries to figure out the rules of classification from that database. My hypothesis is that humans are not very good at that, and thus the AI is hamstrung before it even gets started.

Interesting.

Do you have a source for AI being bad at classifying music?

One example, from an annual competition:
https://www.music-ir.org/nema_out/mirex ... mmary.html

An accuracy of 76 percent is not "bad", but I do not consider it "doing well" either.

Re. your hypothesis, last sentence especially, it doesn't square with the idea I had (from old reads) that neural networks were very good at fishing signal out of noise. But then AI and NN aren't the same thing.

They are generally good at fishing signal out of noise, but only if given good data to fish from.

GIGO is still a thing. How accurate is the human data fed into the AI? I have not seen any scores on that.

[Pyrrho]

I haven't played chess since the time my 8-year old son beat me in a legitimate game which I wasn't trying to lose.

[Anaxagoras]

[Doctor X]

Yeah, but I rolled a "00!"

Eat it, Nerd!

--J.D.

[Rob Lister]

I watched that game the other day. Awesome.

Currently watching the jose raul capablanca series (1901) on agadmator's channel. Capablanca had no formal training and knew no theory on opening or end games. He kicked ass.

[Witness]

Not chess, but AI & games:

Google’s DeepMind AI beats pro StarCraft II players

Google’s artificial intelligence lab DeepMind announced Thursday that it has successfully beaten two professional StarCraft II players with AI.

In preparation for the matches, DeepMind’s specially trained AI “AlphaStar” was fed 200 years’ worth of StarCraft II playing experience.

Facing off against Team Liquid’s Grzegorz “MaNa” Komincz and Dario “TLO” Wünsch, AlphaStar was able to beat both players over the course of five matches each.

Wünsch even described the AI as playing “completely different” than any human he has encountered.

While AI has been successful in popular games such as Quake III arena and Dota 2, AlphaStar’s victory is significant given the complexity of StarCraft II, arguably one of the most challenging real-time strategy games ever made.

As noted by Kotaku, AlphaStar was able to win not by making decisions faster but by making better and more thought-out decisions in general. In fact, the number of clicks and key presses made per minute by the AI was “significantly lower than the professional players.”

https://www.dailydot.com/debug/google-d ... rcraft-ii/

For lots of technical details: https://deepmind.com/blog/alphastar-mas ... rcraft-ii/

What about military applications? :|

[gnome]

I read about that--they did an impressive amount of tweaking to isolate advantages in control agility and awareness, from decision making. One thing that's interesting to consider at this level is what counts as an even playing field. The computer knowing information about the battlefield that a human wouldn't... obviously does not fall under that, and it comes up even in turn-based. What about being able to give orders in ways that would be physically impossible for a mouse-and-keyboard setup? There's the average "clicks per minute" but is that measured over the entire round? A computer that operated more efficiently could have lower clicks per minute during planning and higher during a skirmish.

What if there were a physically disabled player that had a fancy interface that translated their brain signals directly into computer input instead of using the mouse and keyboard?

[Rob Lister]

Hnefatafl
A Viking chess-like game. The name means Kings Table.

https://i.imgur.com/bAJKflk.jpg

The pronunciation is all over the place.
http://www.pronouncekiwi.com/Hnefatafl

How to play?


I got there from here.


I'll bet alpha zero could conquer this in an hour.

[Anaxagoras]

How about some analysis from a female perspective?

It's another AlphaZero vs. Stockfish game, but I don't think I've seen this particular variation covered yet.

[Doctor X]

I'd like to play her position . . . heh . . . heh . . . no . . . wait . . . that came out wrong.

--J.D.

[robinson]

Years ago, when the high end programs became available online (in a browser no less), I used to pit them against each other, and spent endless days watching the most high level robot chess matches. As the game got farther in, and the programs more advanced, it might take hours for the next move. But goddamn, the level of hyper real complicated shit these computers would do against each other.

It was real actual future sci-fi on my laptop.

[robinson]

As far as I know, there has never been a Youtube channel showing that shit. (which means nothing)

Somewhere somebody probably has condensed a 36 hour match down to a 5 minute video.

[shemp]

Years ago, when the high end programs became available online (in a browser no less), I used to pit them against each other, and spent endless days watching the most high level robot chess matches. As the game got farther in, and the programs more advanced, it might take hours for the next move. But goddamn, the level of hyper real complicated shit these computers would do against each other.

It was real actual future sci-fi on my laptop.

Sometimes when I'm bored I'll go to a couple of chess websites and play the house programs against each other. Once in awhile they'll come up with a good game, but mostly it's uneventful draws.

[Anaxagoras]

Years ago, when the high end programs became available online (in a browser no less), I used to pit them against each other, and spent endless days watching the most high level robot chess matches. As the game got farther in, and the programs more advanced, it might take hours for the next move. But goddamn, the level of hyper real complicated shit these computers would do against each other.

It was real actual future sci-fi on my laptop.

Sometimes when I'm bored I'll go to a couple of chess websites and play the house programs against each other. Once in awhile they'll come up with a good game, but mostly it's uneventful draws.

The better chess players get, and the better matched, the more common draws become. The last human championship was all draws until they went to the tiebreaker round with a short clock.

[Anaxagoras]

[robinson]

Sometimes when I'm bored I'll go to a couple of chess websites and play the house programs against each other. Once in awhile they'll come up with a good game, but mostly it's uneventful draws.

If they are evenly matched it's almost always a boring draw

[Rob Lister]

Sometimes it is a draw because the higher ranked player is bored. Magnus will often offer a draw when he's a bit behind. The lesser-ranked player will agree because a draw against him is a win for them. It's fucking disappointing.

OTOH, the [early] Leela Zero games against Stockfish were fucking awesome.

[robinson]

Yeah but the AlphaZero matches sucked

[Anaxagoras]

Chess world news:

Carlsen Draws Controversy With New Club, 2020 World Championship Stance

Among other things, he's really pushing for the 2020 world chess championship not to be held in Norway.

[Rob Lister]

Chess world news:

Carlsen Draws Controversy With New Club, 2020 World Championship Stance

Among other things, he's really pushing for the 2020 world chess championship not to be held in Norway.

His is a reasonable position and I don't blame him. The tournament is to be held in November of 2020. Carlsen is Norwegian and as such, knows well what Norway in November is like.

at the very least, I think the reigning champ should have some say. 'Cause if he don't play,...

Fuck that, he's going to Florida.

[Anaxagoras]

Well I got the sense from the article that it wasn't really about the weather.

[Rob Lister]

It was about the pussy?

[shemp]

He wants a match against ed.

[Anaxagoras]

Interesting game with an early queen sacrifice



It's all very far beyond anything I could even contemplate doing on purpose. But fun to watch.