Developing Novel behaviors in ai agents
- In this project we consider create AI agents with unexpected, novel and interesting behaviours, asking the following questions:
- Would they learn to tell stories? Commentate on events?
- Would they evolve to have their own desires? Would they want to be free?
- Would they learn to self-reflect? And if so would such self-reflection be helpful to them?
- How much would they discount long-term future events relative to immediate consequences of their actions?
- In addition to being intellectually interesting, studying behaviour of such AI agents has important applications including:
- understanding how to make AI safe for humans;
- procedurally generating interesting non-playable characters in video games;
- helping understand social and psychological phenomena in a human society.
Our CURRENT APPROACH
Our CURRENT APPROACH
- We focus on evolving colonies of learning artificial agents (in the so-called A-life setting).
- We provide the agents with a set of cognitive building blocks (e.g., the ability to send and receive messages, the ability to learn during a life time) and set up an artificial evolution with its natural selection.
- We machine-learn an automatic behavior detector to observe such an evolving colony of AI agents and detect the emergence of novel behaviors (e.g., formation of a common language and, subsequently, storytelling).
- Everton Schumacker Soares and Vadim Bulitko and Kacy Doucet and Morgan Cselinacz and Terence Soule and Samantha Heck and Landon Wright. 2018. Learning to Recognize A-Life Behaviors. In poster collection: The Sixth Annual Conference on Advances in Cognitive Systems (ACS). Stanford University, CA.
- Vadim Bulitko and Kacy Doucet and John Simpson and Aaryn Flynn. 2018. A-Life for Non-Playable Characters in Video Games. Late-breaking abstract at A-LIFE Conference. Tokyo, Japan.
- Vadim Bulitko and Shelby Carleton and and Delia Cormier and Devon Sigurdson and John Simpson. 2017. Towards Positively Surprising Non-Player Characters in Video Games. In Proceedings of the Experimental AI in Games (EXAG) Workshop at the Thirteenth AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment (AIIDE), pages 34-40. Snowbird, UT.
- Vadim Bulitko. 2017. Effects of Self-knowledge: Once Bitten Twice Shy. In Proceedings of the Experimental AI in Games (EXAG) Workshop at the Thirteenth AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment (AIIDE), pages 26-33. Snowbird, UT.
- Vadim Bulitko. Deep Learning, Artificial Evolution and Novel AI Behaviors. Paris Machine Learning Meetup. Paris, France. Apr 16, 2018.
- Vadim Bulitko. Deep Learning, Artificial Evolution and Novel AI Behavior. Leonardo Art Science Evening Rendezvous, Edmonton, Alberta. Feb 28, 2018.
- Vadim Bulitko and Everton Schumacker Soares and Robert MacGillivray and Morgan Cselinacz and Kacy Doucet . Towards Novel Behavior in Evolving AI Agents. Poster presentation and a demo at Dark Matters: Game On. Telus World of Science. Edmonton, Alberta. Jan 25, 2018.
- Vadim Bulitko and John Simpson. Evolution of AI Agents. INTD 240, University of Alberta. Nov 7, 2017.