What about chess-playing AI that, after playing several million games against itself, without ever being fed information from historical databases, a chess book, even seeing a human game, independently formulate complex strategic concepts?
Yes. If you aren't, you should perhaps review what you know about how computers work.
You need to give a computer a lot more instructions than just the rules of chess to get it to do what you describe.
You were the one suggesting that it's only the programming, not I.
Oh my. You have very badly misunderstood something. I'm just not sure exactly what. Do you think the rules of chess are the only thing you need to program to do what you described?
What other program besides the rules of chess was AlphaZero given?
My Computer Technology Classes at Purdue were evidently wasted. Es tut mir leid.
Wow. You really don't know anything about how computers work. How exactly do you think AlphaZero learned? That's not spontaneous, computers don't do that automatically. They have to be programmed *how* to learn. And that part isn't in the rules of chess.
If they taught you that computers do anything besides what they are programmed to do, then yes, it was wasted.
Yes. A (digital) computer that does anything other than exactly what the program and data determines is faulty. They are made this way specifically to ensure that they are 100% deterministic. However, as every programmer knows, what you tell it to do and what you want it to do are often not the same thing. LLMs are pretty much guaranteed to do things you don't expect, because the interaction between the program and data is practically impossible for a human to figure out.
I don't think they imply that. Actually in all that AI talk I've been immersed in these few years, philosophy and human parallels are pretty much completely absent.
Did you notice the link I posted? Here it is again.
They haven't been made that way for years, because it's sometimes desirable for them not to be completely deterministic.
AlphaZero *is* a program. The computer it ran on was given the program AlphaZero to run. It ran AlphaZero.
My guess is that you are being deliberately obtuse in the face of having one of your assertions shown to be incorrect.
AlphaZero
The simplest way to get a computer to play chess is to just make it do any valid move until the game's over. This will be easy to beat. Next step is to calculate every possible move and its response, choosing the one that seems to be 'best'. However as more moves are computed in advance the CPU and memory quickly become overloaded. Remembering 'good' moves cuts the work down, so more moves can be computed with a given time and memory - making it a stronger player. Traditional chess programs do this by having a library of opening moves and hard-coded strategies. With a neural net the program learns as it goes, removing hard coded limits and improving efficiency as it plays more.
Not here please. In its current state AI is not reliable enough, and I don't want to waste time verifying information spat out from a machine that's known to be unreliable. Do the work needed to find primary sources or don't bother.
Wow. You really don't understand anything.
The have a log called "Chain of Thoughts" (yes, an anthropomorphism, but not mine) where one of them on the discovery of a forum which others have set up by themselves writes "Woah! Covert mailbox among agents". There is no reason for this exclamation, except that this agent expresses its excitement with words it has learned from humans. Denying that something similar to a human emotion is displayed is much like how people let eels writhe in pain when put into a salt bed to remove the fat coat, and then saying "fish have no sense of pain because of X. The writhing movements are just muscle reactions".
The agents write about it in their CoT, and in messages among themselves on their forum. But OK, if "fraud" is a purely legal concept, you'll have to substitute it with the correct English non-legal term.
But they use the words for emotions. Walk like ducks, quack like ducks …
I take it that you didn't actually read the report.
