How games teach AI to learn for itself

This podcast episode features Julian Togelius, associate professor at NYU Tandon, who speaks about the intricacies of artificial intelligence in game structure. According to Togelius, games like Chess “became core focuses of artificial intelligence research very early on,” but “there are so many more possibilities, so many more interesting challenges in the other games.” 

While it is true that the scope of what AI has come incredibly far, Togelius argues that AI is still doing relatively simple procedures. Compared to how humans function, AI falls short, argues Togelius, who notes that AI can't multitask or apply the same concepts across a variety of obstacles. "AI is good at what it does, but only what it does. 

“If we had a system that could reliably play, like with some proficiency, the top hundred games on a computer game top list, like Steam or the AppStore or something, then we would have something akin to general intelligence,” says Togelius.