An avid programmer since the 80s, I have used and abused Prolog, LISP, Smalltalk, Java and now Elixir.
I am a DDD, Functional Programming and Actor Model enthusiast.
I have forever obsessed about domain knowledge, conceptual clarity and consistent naming.
I work in Portland, Maine as an independent consultant.
In my spare time, I train regularly at Aikido of Maine, organize the local Erlang/Elixir and Elm meetups, and dabble in robotics with the Lego Robotics sets I bought "for my son".
This is the third in a series of talks on developing cognitive models for autonomous robots. In 2017, I presented a homebrew model, inspired by Minsky's Society of Mind, that I implemented as a hobby project for my Lego robots. I showed robots scurrying about for food and safety, all the while emoting audibly. Fun! But they couldn't learn so I went back to the drawing board. In 2018, inspired by current neuroscience research in Predictive Processing, I rebuilt my robot's cognitive model to enable (some) learning. It worked! I got my robot to learn more effective behaviors from experience, but I had only scratched the surface.
This year, I expanded my Predictive Processing-inspired cognitive model so that a robot could also learn the intentions of other robots. The goal is to have them get better at competing for food, avoiding running into each other etc. In a way, my robots are now building and evolving "models of mind" of each other while interacting (just like we do.) Robots modeling the minds of robots!