As an old Ph.D. in experimental psychology and from Carnegie Mellon University in both psychology and computer science, this is nice. Thank you.
Now for another little area where it applies: human privacy and its subset human lying. In my cognitive science theory book on “How to get your privacy back” I refer to Skinner and others on verbal behavior. Basically privacy is seen as part of a perceived-value system (reinforcement learning) associated with what to say and when. This is based on the value of secrets defined as things you could say if you think you can get value out of that “information property” and can trust the people you tell to help you retain that value for yourself. Similarly for lying which is a way to protect the value of information you wish to keep secret. The neocortex, it could be argued, is an amazing predication engine that is driven by reinforcement learning in figuring out what predications to say, write, or wave one’s hands about (as I did in the privacy and lies books).
You nicely reveal that a bunch of us have believed in reinforcement learning even through the “death of behaviorism” and the rise of “cognitive neuroscience.” And the shame that should go to the AI people who say they are not interested in how the human brain computes, but merrily go around using reinforcement learning as if it is an AI thing devoid of a proper recognition of where it came from in psychology.
Books: https://www.amazon.com/How-Get-Your-Privacy-Back/dp/B07DWFCDTS/
https://www.amazon.com/How-Get-Your-Lies-Back/dp/B07R448S2L/
And a no-paywall medium article summarizing the lies book in an 11 minute read:
