It reminds me of the old Roald Dahl short story, "The Great Automatic Grammatizator," in which an aspiring writer invents a machine that can write short stories.
I am reminded of the story "Melancholy Elephants" by Spider Robinson. It's actually a criticism of the extension of copyrights, but it is relevant here.
Robinson notes (through his protagonist) that the ways one can combine musical notes are FINITE. Tremendously large, but finite. And most combinations are crap; little better than noise. Composers don't really "create" a new piece; they just "discover" one of the pleasing combinations.
By turning computers loose on "creative" fields, we are reducing the amount of "art" that can be created (or discovered) by humans.
That's depressing. It's bad enough that automation is taking over Work - what happens when it takes over everything else?
Joe probably knows this, but Agadmator played Mittens against Stockfish 14 and Stockfish destroyed the poor little kitty. Video here: https://youtu.be/VWAG5OhhvUA
The article's main point--that machines are better than humans at chess, and that changes humans' relationship to chess--is much broader and more chilling. Pretty soon, machines will be much better than us at essentially everything. And then what's the point of doing or learning anything? The British SciFi show "Humans" did a good job showing human problems living in a world where, no matter what you learned or did, a machine already knew it or could do it better. The show was good. So is the article. Thanks, Joe.
I'm 82 so I'm not an expert in AI, but I do have a strong feeling that someone should get chess credit for programing such a vehicle that can beat everyone in chess.
Someone, or more likely a group of someones, wrote Mittens' code. I think that is a creative endeavor that is worthy of celebration and an artform itself.
Actually, not sure about that. And anyway, even if Mittens *was* written by a human, we are coming upon a time when the best chess programs will be written by computers, and the ones wtitten by humans just won't be as good. . . .
I'm seeing a bunch of stats in baseball comparisons to chess computers.
I think a key difference here is that professional baseball exists — and the reason it's a multi-billion dollar industry — is to be entertaining for people to watch.
I'm not sure how I'd define the purpose of chess, and to what extent it's just pure competition, but it's definitely not entertainment for the masses.
But pro baseball is supposed to be entertaining, and when it gets pushed away from that goal by statistical analysis or anything else, there's a responsibility (one they've abdicated until recently) to make changes to bring the sport back into alignment with its goal of being entertaining.
There's no doubt that the romantic perception is lost through statistical analysis and computers. It's not always a negative, though - the "way the game is played" and "everybody knows" is not a good thing (leads to crappy #2 hitters and lots of punting as a couple of examples). The key for me is that it doesn't devolve into right and wrong ways to play such that everyone plays the same. It's been happening in baseball with more and more players from the same cloth (power pull hitters who don't swing much, pitchers throwing with very high velocity, conservative baserunners, etc.) and similar coaching styles (avoid 3x through the order, defensive shifts, play station-to-station). The good news is the rules are changeable (yes, there will be outcry) to prevent or at least reduce the homogenization.
Chess can't change its rules. Hopefully, computer play and, for that matter, training and coaching via computer, doesn't result in most people playing the same way. I have a feeling that it won't. I think people will play a style that fits their personality and most will likely not resort to memorization to rack up quick wins and boost their rating. As Nate Solon pointed out, there may be more opening moves used which should open up more variety of positions. As for cheating, my answer has always been to not play them.
i, too, frequently 'rage' against what is happening in our world, our all-subsuming subservience to the machine(s) that we've made- but verily, i can't stop it, you can't stop it, none of us can stop it!!!
and frankly, that both saddens and terrifies me- and it's only going to get worse w/no chance of getting better...
"the machine that we've made can never save us, that's what they say" -Jimi Hendrix (1984, A Merman I Shall Be)
Reminds me of Sabermetrics. I started reading James’s newsletters around 1980 as an 11-year-old, and realized he was on to something. I was happy when the establishment finally conceded two decades later that so was he, and the game of baseball changed for good. Or so I thought. I didn’t foresee what was to happen, which was that once enough teams were employing optimal strategy, the fun from the game would be drained. For the first time since maybe 1973, I didn’t follow baseball last season. And I didn’t miss it. The battle against inefficiency has been won, and the sport is the worse for it.
But the difference between a sport like baseball and a game like chess is that baseball has so many variables you can make adjustments to fix problems without changing the basic nature. The problems more and more people are finding they have with baseball are less about the sport being "solved" and more about the sport not adjusting as required.
Chess has far less room for those sort of adjustments. The strictness of everything (the size and shape of the board, the movement of the pieces) is both the simple magic of the game for us, and the flaw that allows a computer to figure out every possible move at any point.
But there are ways you can vary chess. The most obvious is to restrict people’s time to think which causes suboptimal moves and makes the game more fun. There are also randomized versions of cheese where each player starts in a position they’ve never seen before.
It seems as if all games can be solved eventually and get stale, and that when it’s on humans to innovate and make something new and cool.
Another point: chess isn’t just for the fanatics. I play my six-year-old son sometimes and neither of us have anything solved.
I loved reading about your over the board chess games. Maybe you wrote only about playing Priest Holmes but those games sound like Art.
Get some people you know and perhaps people they know and get together and play. You make the game about Art. Aim for a beautiful move or try something creative or out of the norm.
I am not good either and I dislike the bots and have played on Chess.com when a human opponent suddenly comes up with a couple great moves after a series of puzzling blunders. No doubt there is cheating over the web, but so be it.
Play over the board in person. Ignore the analysis of Brilliant moves, book moves, good moves and blunders. Let go of win/lose and make it artful.
I love the immortal game and the games of artist Marcel Duchamp. Find an unorthodox or dated opening and play it to see what happens.
I'm not sure I see how Mittens Bot is really a whole lot different than playing a really hard baseball simulator.
It reminds me of the old Roald Dahl short story, "The Great Automatic Grammatizator," in which an aspiring writer invents a machine that can write short stories.
https://archive.org/details/greatautomaticgr0000dahl_a2f7/page/2/mode/2up
I am reminded of the story "Melancholy Elephants" by Spider Robinson. It's actually a criticism of the extension of copyrights, but it is relevant here.
Robinson notes (through his protagonist) that the ways one can combine musical notes are FINITE. Tremendously large, but finite. And most combinations are crap; little better than noise. Composers don't really "create" a new piece; they just "discover" one of the pleasing combinations.
By turning computers loose on "creative" fields, we are reducing the amount of "art" that can be created (or discovered) by humans.
That's depressing. It's bad enough that automation is taking over Work - what happens when it takes over everything else?
Joe probably knows this, but Agadmator played Mittens against Stockfish 14 and Stockfish destroyed the poor little kitty. Video here: https://youtu.be/VWAG5OhhvUA
The article's main point--that machines are better than humans at chess, and that changes humans' relationship to chess--is much broader and more chilling. Pretty soon, machines will be much better than us at essentially everything. And then what's the point of doing or learning anything? The British SciFi show "Humans" did a good job showing human problems living in a world where, no matter what you learned or did, a machine already knew it or could do it better. The show was good. So is the article. Thanks, Joe.
I'm 82 so I'm not an expert in AI, but I do have a strong feeling that someone should get chess credit for programing such a vehicle that can beat everyone in chess.
I am not a person who plays chess but I say Death to Mittens!
Someone, or more likely a group of someones, wrote Mittens' code. I think that is a creative endeavor that is worthy of celebration and an artform itself.
Actually, not sure about that. And anyway, even if Mittens *was* written by a human, we are coming upon a time when the best chess programs will be written by computers, and the ones wtitten by humans just won't be as good. . . .
I'm seeing a bunch of stats in baseball comparisons to chess computers.
I think a key difference here is that professional baseball exists — and the reason it's a multi-billion dollar industry — is to be entertaining for people to watch.
I'm not sure how I'd define the purpose of chess, and to what extent it's just pure competition, but it's definitely not entertainment for the masses.
But pro baseball is supposed to be entertaining, and when it gets pushed away from that goal by statistical analysis or anything else, there's a responsibility (one they've abdicated until recently) to make changes to bring the sport back into alignment with its goal of being entertaining.
Has there been a skynet reference yet, because if not here it is. They are coming.
I would give that 88 a B+ and maybe even an A-! Not just a B!!!
Joe sounds remarkably like someone complaining about sabermetrics ruining baseball or three-pointers ruining basketball. Which doesn’t make him wrong.
There's no doubt that the romantic perception is lost through statistical analysis and computers. It's not always a negative, though - the "way the game is played" and "everybody knows" is not a good thing (leads to crappy #2 hitters and lots of punting as a couple of examples). The key for me is that it doesn't devolve into right and wrong ways to play such that everyone plays the same. It's been happening in baseball with more and more players from the same cloth (power pull hitters who don't swing much, pitchers throwing with very high velocity, conservative baserunners, etc.) and similar coaching styles (avoid 3x through the order, defensive shifts, play station-to-station). The good news is the rules are changeable (yes, there will be outcry) to prevent or at least reduce the homogenization.
Chess can't change its rules. Hopefully, computer play and, for that matter, training and coaching via computer, doesn't result in most people playing the same way. I have a feeling that it won't. I think people will play a style that fits their personality and most will likely not resort to memorization to rack up quick wins and boost their rating. As Nate Solon pointed out, there may be more opening moves used which should open up more variety of positions. As for cheating, my answer has always been to not play them.
There have ALWAYS been common strategies in sports. That's not the problem - the problem is the type of game those strategies create.
i, too, frequently 'rage' against what is happening in our world, our all-subsuming subservience to the machine(s) that we've made- but verily, i can't stop it, you can't stop it, none of us can stop it!!!
and frankly, that both saddens and terrifies me- and it's only going to get worse w/no chance of getting better...
"the machine that we've made can never save us, that's what they say" -Jimi Hendrix (1984, A Merman I Shall Be)
And anyway… it would be against the will of God, and the grace of the King!
This one is definitely gonna be discussed with my psychiatrist this week.
Reminds me of Sabermetrics. I started reading James’s newsletters around 1980 as an 11-year-old, and realized he was on to something. I was happy when the establishment finally conceded two decades later that so was he, and the game of baseball changed for good. Or so I thought. I didn’t foresee what was to happen, which was that once enough teams were employing optimal strategy, the fun from the game would be drained. For the first time since maybe 1973, I didn’t follow baseball last season. And I didn’t miss it. The battle against inefficiency has been won, and the sport is the worse for it.
But the difference between a sport like baseball and a game like chess is that baseball has so many variables you can make adjustments to fix problems without changing the basic nature. The problems more and more people are finding they have with baseball are less about the sport being "solved" and more about the sport not adjusting as required.
Chess has far less room for those sort of adjustments. The strictness of everything (the size and shape of the board, the movement of the pieces) is both the simple magic of the game for us, and the flaw that allows a computer to figure out every possible move at any point.
But there are ways you can vary chess. The most obvious is to restrict people’s time to think which causes suboptimal moves and makes the game more fun. There are also randomized versions of cheese where each player starts in a position they’ve never seen before.
It seems as if all games can be solved eventually and get stale, and that when it’s on humans to innovate and make something new and cool.
Another point: chess isn’t just for the fanatics. I play my six-year-old son sometimes and neither of us have anything solved.
I loved reading about your over the board chess games. Maybe you wrote only about playing Priest Holmes but those games sound like Art.
Get some people you know and perhaps people they know and get together and play. You make the game about Art. Aim for a beautiful move or try something creative or out of the norm.
I am not good either and I dislike the bots and have played on Chess.com when a human opponent suddenly comes up with a couple great moves after a series of puzzling blunders. No doubt there is cheating over the web, but so be it.
Play over the board in person. Ignore the analysis of Brilliant moves, book moves, good moves and blunders. Let go of win/lose and make it artful.
I love the immortal game and the games of artist Marcel Duchamp. Find an unorthodox or dated opening and play it to see what happens.