Free Will in Artificial Intelligence: What volition neuroscience says about free behaviour and implications for the design of autonomous artificial intelligence agents
Organizer and Chair: Catalin Mitelut, Center for Theoretical Neuroscience, Columbia University, New York, USA
The human capacity for free will or volitional (i.e. voluntary) behaviour has intrigued scientists and philosophers for thousands of years. Over the last few decades, neuroscientists have uncovered many neural correlates of voluntary behaviours and identified several decision stages involving specific neuroanatomy and dynamics (Haggard 2008). While a reward-optimized decision framework lies at the core of most fast decisions it is supported by slower time-course motivational systems that identify long-term needs (e.g. feeding, offspring care; Maslow 1943, Kenrick 2010). In the absence of naturally evolved motivational drives, autonomous general artificial agents will require the design of motivational systems that will pose unique challenges to our understanding of free will while offering creative opportunities.
This symposium seeks submissions focusing on extending evolutionary biology and the neuroscience of volition towards the design of internally motivated, freely behaving autonomous artificial agents.
Paper & Abstract submission deadline: May 31, 2020 (Anywhere on Earth)
Paper notification: July 15, 2020