How Forza's Racing AI Uses Neural Networks To Evolve | War Stories | Ars Technica
From Ars Technica · · 28 min

The player loads from YouTube only when you click. It uses YouTube's privacy-enhanced mode.
If the video does not play, the owner may have turned off embedding. Watch it on YouTube instead.
Description
Today on War Stories, Ars Technica is joined by Dan Greenawalt, Creative Director of the Forza franchise, who takes us through the colossal potential (and many challenges) of utilizing deep neural networks to build sophisticated racing AI. Beginning with the original Drivatar AI in Forza Motorsport for XBOX in 2005, Greenawalt and his colleagues experimented with computer learning, refining their methods from iteration to iteration. It was once the AI system was unshackled from local hard drives and placed online in Forza Motorsport 5, however, that the floodgates truly opened. As the AI rapidly evolved, producing unexpected new behaviors, the Forza team quickly learned how challenging it would be to wield this two-edged sword in creating realistic opponents.
00:00 Introduction 00:42 Chapter 1: The Setup 07:17 Chapter 2: The Problems With AI 11:00 Chapter 3: Drivatar 19:24 Chapter 4: Rubber-banding 23:32 Chapter 5: The Lesson
Thank you to the following YouTube users for gameplay footage: kevi86 TheYamiks pHiLaToR99 Tyerell testdrive426 Alex SandOr Paragleiber (Gleiber) GameplayOnly SpecAgentProduction POWERHOUSEHD John GodGames Freelancer Gamer Pirate Love SnoopZaKT1 ApexClone RapidKiller11b THE RED DRAGON
Connect with Ars Technica: Visit ArsTechnica.com: http://arstechnica.com Follow Ars Technica on Facebook: https://www.facebook.com/arstechnica Follow Ars Technica on Google+: https://plus.google.com/+ArsTechnica/videos Follow Ars Technica on Twitter: https://twitter.com/arstechnica
How Forza Uses Neural Networks To Evolve Its Racing AI | War Stories | Ars Technica