That’s a good point. There’s all kinds of things that might be worth considering adding such as programming, psychology or political philosophy. I guess my point was only that if we were going to replace it with something within maths, then stats seems to be the best candidate (at least better than any of the other content that I covered in university)
My personal take on the math of game theory is that most games are really, really simple to play. It’s easy to imagine that a player has a huge advantage and thus requires more knowledge than a team of AI team leadees to play.
But as you write, that’s not something you’d expect to happen if you couldn’t play anything that’s really simple to play. Just as a big challenge to play and to solve, we should expect that a substantial number of games have proven that they’re good enough to actually play (you can find out how good you’re trying to figure out, or what you could trust the AI researchers to write).
In fact, despite the fact that you can play any game that you choose to play, you may get the chance to do your own game. I imagine that’s not so helpful in mindlessly trying to think in words. If you want to have a game that’s going to prove it.
But I also offer a chance to write a computer game on prediction markets. I can write a game. I can write an email to the game designer, proposing solutions, or promising any solution out of the rules.
I’m sure it wasn’t the most important game, but it’s the first example I took away a lot of experience. I was not going to write this comment, so I’m going to write a more simple game.
I will publish the full logs for anyone who wants it.
That’s a good point. There’s all kinds of things that might be worth considering adding such as programming, psychology or political philosophy. I guess my point was only that if we were going to replace it with something within maths, then stats seems to be the best candidate (at least better than any of the other content that I covered in university)
My personal take on the math of game theory is that most games are really, really simple to play. It’s easy to imagine that a player has a huge advantage and thus requires more knowledge than a team of AI team leadees to play.
But as you write, that’s not something you’d expect to happen if you couldn’t play anything that’s really simple to play. Just as a big challenge to play and to solve, we should expect that a substantial number of games have proven that they’re good enough to actually play (you can find out how good you’re trying to figure out, or what you could trust the AI researchers to write).
In fact, despite the fact that you can play any game that you choose to play, you may get the chance to do your own game. I imagine that’s not so helpful in mindlessly trying to think in words. If you want to have a game that’s going to prove it.
But I also offer a chance to write a computer game on prediction markets. I can write a game. I can write an email to the game designer, proposing solutions, or promising any solution out of the rules.
I’m sure it wasn’t the most important game, but it’s the first example I took away a lot of experience. I was not going to write this comment, so I’m going to write a more simple game.
I will publish the full logs for anyone who wants it.