Open-endedness, curiosity-driven AI and automated search in artificial life

Presentation made by Pierre-Yves Oudeyer at The 6th International Workshop on Intrinsically Motivated Open-ended Learning (IMOL 2023)
Slides: tinyurl.com/op...
Topics in this talk:
What is open-endedness? Is this a subjective notion? How can we approach it at the evolutionary, cultural and developmental levels?
The role of curiosity-driven learning in open-ended cognitive and social development
How to build open-ended curiosity-driven autotelic AI systems?
Explanation of the concept of autotelic artificial intelligence
Role of language and LLMs in creative autotelic exploration, examples of the IMAGINE and VOYAGER AI architectures
Short history of research on curiosity-driven learning and intrinsic motivation in AI and machine learning
How autotelic exploration can help solve hard engineering problems
How autotelic exploration algorithms can help phycisists and chemists explore and make discoveries in complex systems, e.g. building maps of possible self-organized structures with a limited budget of experiments
How autotelic exploration can help artificial life researchers automate search of open-endedness and self-organization of interesting structures in continous cellular automata such as Lenia or Flow Lenia