Events Calendar
Sign Up

32 VASSAR ST, Cambridge, MA 02139

#csailalliances csail
View map

Games and Machine Learning – an MIT Independent Activities Period offering

This course will combine a series of lectures and hackathon-style projects, the latter chosen from a list of examples or otherwise devised, to introduce students to game creation, current game-related research and an exploration of the technology, the art and the fun of video games.

Sony Interactive Entertainment : The Nexus of Games and AI

Description: Computers and gaming have grown up together. Since Bertie the Brain learned to play Tic Tac Toe in 1950, computers have hosted, played, and designed increasingly sophisticated games as they have grown in power. For every computer science paper on the arXiv – from computer vision to LLMs to personal immersion to cognitive science and general AI – there are a half dozen use cases you can name in the creation of video games. Furthermore, video games provide new worlds and synthetic data that test and stretch the capabilities of machine learning models – so the relationship is synergistic.

This IAP course will give a brief, project-oriented survey of some of the main research topics that are integral to the development of the future of gaming. Lecturers from Sony Interactive Entertainment, the company behind PlayStation, as well as contributors from MIT and other universities will discuss topics such as player immersion, persuadable chatbots and realistic NPCs, sentiment analysis, rendering, asset and game creation with Unreal Engine and the use of LLMs for agent planning and the development of game tutor agents. Students will choose from a list of potential course projects (or invent their own) and the final week will include student project presentations.

Audience: The course is targeted at those who have an interest in video games and machine learning. There are no specific prerequisites. We encourage students with an interest in art, psychology, writing, social impact and design of games to participate.

Structure: Project-oriented course combined with lectures.

Readings: Readings will be dependent upon project choice and be chosen from the literature.

Lectures

  • Number of lectures: 9
  • Lectures:
    • 1/6/25     Mike Stopa – The lay of the land: AI in video games
    • 1/8/25     Keri Carpenter – Creating a game concierge for our games
    • 1/10/25   D. Fox Harrell, MIT: Considering the Avatar Dream: Designing and Evaluating Characters and Avatars
    • 1/13/25   Gale Lucas, USC – Cognitive science and realistic NPCs
    • 1/15/25   Ram Barankin – Using GenAI to listen social media
    • 1/17/25   Erick Flores – Neural methods to improve real time game streaming
    • 1/22/25   Logan Olson, Haven Studios – “0-to-1” game development with Generative ML
    • 1/24/25   TBD
    • 1/27/25   Final projects + TBD

Project examplesstudents will be asked to complete a hackathon-style project for the class. The topic is completely open although several categories of projects will be suggested by our lecturers – who can then provide guidance in executing the projects. These projects do not need to be computational. Pure artistic projects designing characters, for example, would be welcome.

  • Unreal Engine (or Unity) simple game creation (Mike)
  • Cognitive Science and Realistic NPCs  (Gale)
  • Creating Novel Game Mechanics with LLMs  (Logan)
  • Video Enhancement, Video Quality Measurement or Codec tuning using Machine Learning  (Erick)
  • AI Generated Gamer Concierge   (Keri)
  • A Generative AI-Powered Sentiment and Insights Tool for Social Listening (Ram)

Event Details

See Who Is Interested