Cell Assemblies for Artificial Intelligence

• Our long-term goal is to build real AI.
• We're interested in real neurons doing real things.
• Hopefully, eventually, we'll be able to do human-like things.
• There is a long way to go.
• We've made a start.
• We're looking for collaborators, ideas, and critics.

We do not understand the way neural circuitry works. One of the chief lacks is the understanding of reverberation. A second lack is understanding the dynamics of learning and how this is affected by reverberation. CAs use reverberation and learning, so CAs can be used to study them. By simulating computational and psychological tasks with networks of neurons that have CAs, we can better understand human intelligence. Moreover, by understanding systems of CAs we can build systems with novel capabilities.