A Route to AI

• I think this biological modelling provides a route to real AI (Turing test).
• We've been working on AI for a long time, and we have lots of really great systems from that.
• However, we seem to have problems with domain general systems.
• We have to craft the individual systems.
• By developing models based on neurons, that have accurate Psychological behaviour, we can gradually develop systems that increasingly approximate human behaviour.
• That's what my CABot agents have done.
• They're simple assistants in really simple virtual environments.
• They view, understand commands, plan and do some learning.
• There is clearly a long way to go to pass the Turing test.