AlphaMystica
So, what’s the deal with AI these days?
What do they know? What are they capable of? What do they want from us? And how do they feel when we abandon them in the woods with their sad little faces staring longingly into our swiftly departing rearview mirrors?
You’ll find out all that and more from this article.
Designer Diary: The Search for AlphaMystica
Alright, I confess, it’s not really about that stuff I said before. But it is about AI and boardgames.
Tysen Streib, an AI developer for games shares his insights and struggles as he works on an AI for a digital version of the game Terra Mystica.
I’ll admit, the article gets very technical by the end—even for a genius like me (yes, I still have to look up how to spell the word ”genius”), but the beginning part is especially interesting. It turns out that not only can AI beat the best chess player, they can also beat the best Go player.
First, there was AlphaGo, which learned by studying human games, then came AlphaZero, which was completely self-taught and markedly better than AlphaGo. It’s kinda freaky when you think about it too much (which I do), especially if you’re really into boardgames (which I am).
What other areas of life will AI develop complete mastery over their human competitors?
Well, I have yet to meet one able to make a better quesadilla than I can, so we’re still safe for the time being. But when that day comes ... be afraid, be very afraid.