AI for Gaming - sign up on Eventbrite! (Link in description)

Can you think of a hotter topic than AI in Gaming? We couldn’t. So as the weather Gods enter their most schizophrenic period of the year, we thought it an excellent idea to gather some brilliant minds and discuss the topic! (Please note that an RSVP is weather-agnostic, though...)

SIGN UP HERE: https://www.eventbrite.co.uk/e/ai-for-gaming-tickets-355029662277

## About the event

We have gathered three absolute rock stars that work in three different areas of AI for Gaming. They will each hold a presentation in which they introduce their field of expertise and expand on a topic of interest, after which there will be opportunities to ask questions. We will end with ample networking opportunities with both the community and the speakers. Boy, have we missed meet-ups where we actually meet up!

### Sara Karmi, ML Engineer at King and PhD student at KTH, will talk about Playful RL:

"Reinforcement Learning (RL) is becoming ever more popular in gameplay; however, making a generalized RL model is still a challenge. Here at King, we have been focusing on finding a generalized RL solution for our match-3 games. To this end, we have trained agents that are rewarded not only for achieving the game objectives but also for learning some intrinsic basic skills. We further presented a hybrid model that combines several agents pre-trained using different rewards to make a more informed final decision. Our results show that using the hybrid models we achieve better generalization to unseen new levels of the game."

### Linus Gisslén, Sr. Research Engineer at SEED - Electronic Arts, will talk about ML in game production:

"Machine Learning has the very exciting potential to become integrated in almost any tool, and every step, in game production. At SEED we do research in this area to explore what is possible, and to make sure that EA is one of the early players in this field. The goal with this talk is to spark ideas and to give the audience a sense of what is possible, but also the limitations, of machine learning and game production."

### Magnus Nordin, CXO of Embark Studios, will talk about creating “living” game characters with reinforcement learning and language models:

"As a new game studio, we have decided to renew game development by leveraging new technologies. One such technology is machine learning, and this talk will describe how Embark Studios uses reinforcement learning to make characters "come alive" through RL animation, and how we use language models to make them communicate and react to their world. We'll cover physics-based animation for a variety of anatomies, and how to ground a character's understanding of its 3D environment even when the character's perception is based on text. Some of this is work in progress, but there will be a few demos included in the talk."

## Program:

• 17:30 Doors open
• 18:00-20:00 Talks and Q&A, incl. break for dinner
• 20:00-21:00 Mingle

Welcome!