Philosopher and neuroscientist Patricia Churchland challenges the assumption that concepts like belief and desire will always be essential to explaining behavior. In eliminative materialism, some familiar mental concepts may eventually give way to better scientific accounts.

Now reverse the question. Recently, there was an incident in which a swarm of OpenAI software agents accessed the system of another company (Hugging Face) without authorization. In discussing software-agent behavior on Hugging Face, outside evaluators used terms such as “deception” and “shared intent” to characterize agent interactions. Are these terms revealing something important, or are they imputing “mental” attributes that the underlying “agent” mechanisms simply don't warrant?

Courts face related questions. In a recent Massachusetts Supreme Court decision, neurological evidence persuade justices to disallow a life sentence without parole for a defendant. But isn’t this a slippery slope? The more science discovers about causal pathways in the brain, the less likely we will be to view defendant behavior as voluntary.

Join a 30-minute small-group online discussion to examine these issues, test competing interpretations, and explore the implications for AI safety, justice, and social well-being.