Good evening. You are listening to Three Moves Ahead, and I'm your host, Rob Zakeney,

continuing on his crummy hub headset mic as he awaits a real microphone replacement.

So I apologize for poor audio quality.

But sounding rich and resonant, as always, on his professional recording gear, is Julian Rabbit Murdoch.

Yes, the problem is the content, not the delivery.

We also welcome back, our friend Soren Johnson, Civilization 4 designer, and, you know,

he seems to like spending all his vacations with three moves ahead.

Hey, it's my favorite place to be in August.

And isn't that an endorsement?

We also welcome back, Vic Davis from Cryptocomet.

Vic, welcome back to the show.

Thank you.

It's a pleasure to be here.

So one of our most requested topics on three moves ahead, and certainly one of the areas

that I think strategy gamers love to complain the most is artificial

intelligence, you know, what it should do, what they tend to do badly, and what we really want

out of them, and then what are the obstacles to creating a good artificial intelligence?

And so it seemed like since we've got Soren for the duration of his vacation and since Vic was willing to come on the show, it seemed like a good idea to finally address the topic of artificial intelligence in strategy games. So I'll just throw this open to either one of you. When you're thinking about the role of AI, what do you think is, what is an AI obligated to do in a strategy game?

So usually when I talk about AI and strategy games, I talk about it in terms of continuum.

And what defines the continuum is how integrated into the gameplay the AI is.

And what that means is on one end of that continuum, you have sort of a pure game, you know, like chess, which is, you know, not going to change.

And it really has nothing to do with the AI itself. And the AI is simply there to, you know, not going to change. And it really has nothing to do with the AI itself.

...