Next Generation AIs Get To Play Wing Commander (January 25, 2017)

ChrisReid

Super Soaker Collector / Administrator
The OpenAI project has announced a virtual environment - dubbed the Universe - where it will allow artificial intelligences to train in front of a variety of virtual computing interfaces. Ultimately the goal is to enhance AI so that it can more seamlessly perform tasks in the real world, and this simulation is an interim learning step. Some of the modules are simple computing tasks, but a number of video games will also be used to provide the AIs different environments to interact with. And of course, you would want your future robotic overlords to have good taste in video games, so Wing Commander 3 will be one of the games that they learn! There's no word yet on whether they will be reading End Run and Fleet Action to have a broader appreciation of the Heart of the Tiger's context in the larger Wing Commander Universe. You can learn more about the project here.

Universe allows an AI agent to use a computer like a human does: by looking at screen pixels and operating a virtual keyboard and mouse. We must train AI systems on the full range of tasks we expect them to solve, and Universe lets us train a single agent on any task a human can complete with a computer.

Our goal is to develop a single AI agent that can flexibly apply its past experience on Universe environments to quickly master unfamiliar, difficult environments, which would be a major step towards general intelligence. There are many ways to help: giving us permission on your games, training agents across Universe tasks, (soon) integrating new games, or (soon) playing the games.

With support from EA, Microsoft Studios, Valve, Wolfram, and many others, we've already secured permission for Universe AI agents to freely access games and applications such as Portal, Fable Anniversary, World of Goo, RimWorld, Slime Rancher, Shovel Knight, SpaceChem, Wing Commander III, Command & Conquer: Red Alert 2, Syndicate, Magic Carpet, Mirror's Edge, Sid Meier's Alpha Centauri, and Wolfram Mathematica. We look forward to integrating these and many more.






--
Original update published on January 25, 2017
 
Last edited by a moderator:
I read the article but not thoroughly... I wonder if they'll plop the AI into the game and expect it to work things out without an explanation of which keys perform which functions. Would be interesting to see how fast it learns how to do things. Wonder if it'll work out how to destroy capital ships... or unknowingly fire on friendly craft... I remember seeing demos of AIs playing old games like Super Mario Bros and while they'd get Mario killed by running into the first Goomba or letting him fall into the first pit, eventually they would perform quite remarkably, learning how to pick up the Mushrooms and Fire Flowers for power-ups, etc.
 
Back
Top