Goat Simulator 3: Exploring the Power of Artificial Intelligence to Learn More Using Lessons Learned from a Video Game
Frederic Besse, a research engineer at DeepMind, said in an email that “SIMA is greater than its parts.” “It is able to take advantage of the shared concepts in the game, to learn better skills and to learn to be better at carrying out instructions.”
Harley said the team chose games that were more focused on open play than narrative to help SIMA learn general gaming skills. If you have played the game, you know that the point is for it to be random, but Harley said they hoped it would learn tricks from it.
Goat Simulator 3 is a surreal video game in which players take domesticated ungulates on a series of implausible adventures, sometimes involving jetpacks.
DeepMind has created a video game project that shows how artificial intelligence could do more than simply chat and generate images, by controlling computers and performing complex commands. That’s a dream being chased by both independent AI enthusiasts and big companies including Google DeepMind, whose CEO, Demis Hassabis, recently told WIRED is “investing heavily in that direction.”
As Google, OpenAI, and others jostle to gain an edge in building on the recent generative AI boom, broadening out the kind of data that algorithms can learn from offers a route to more powerful capabilities.