My Patron supporters got this podcast early. Support me on Patron to do the same and get audio versions of other articles.
Playing with and against other humans is great in many ways (and not so great in others) but the fact is that gamers spend a lot of time interacting with computer-controlled agents. Enemies, shopkeeers, quest givers, teammates, other NPCs –they can all be controlled by a game’s artificial intelligence. AI has come a long, long way thanks to advances in the field and increased processing power on our gaming hardware, but some games are still better than others at making us feel that an NPC or enemy bot is acting like a human.
But to make something inhuman act human, you have to know something about how our fleshy meat brains work. You have to know a thing or two about human psychology. Humans don’t always act rationally. They take social information like reputation into account when dealing with people. They use mental shortcuts in their decision making that produce weird results. Their perception of a scene can be affected by their attentional resources and the contextual baggage their puny minds bring with them. Can you teach a computer to emulate all that?
My guest on this episode of the podcast thinks so. His name is David Mark and he’s an expert on developing AI for video games. Mark has also made it a point of studying psychology and applying its lessons to creating AI that seems human if you’re willing to suspend a bit of disbelief.
To get the podcasts delivered straight to your device of choice, search for “Psychology of Games” or use one of these links:
To listen to the podcast right now in a browser window, click the play button below.
Info on this week’s guest:
- “Robot Motivation” by The Polish Ambassador, licensed under Creative Commons: CC BY-NC-SA 3.0
- “Winner Winner!” by Kevin MacLeod Licensed under Creative Commons: By Attribution 3.0
- “Hidden Agenda” by Kevin MacLeod Licensed under Creative Commons: By Attribution 3.0
- “Unwritten Return” by Kevin MacLeod Licensed under Creative Commons: By Attribution 3.0