Cognitive Models

Another area that I do know about is cognitive models.
These are, now, typically programs that tell you how people are going to think (on a restricted set of tasks).
I'm particularly fond of cognitive architectures (like Soar or ACT) that claim to be the basic mechanism of thought. They claim, all that is needed is a particular program in the language to perform the behaviour like people will.
Cognitive architecture models have been used to, for example, influence policy on driving while on a mobile phone. They show that it leads to cognitive overload.
In the 60s there were box models that showed parts of the psychological system, like long and short term memory and how they interact. Now we have programs that show the interactions.
These can also be used to perform usability tests and simulate human behaviour (in for example air combat simulations).
I work on neuro-cognitive models that show how the neurons behave to generate psychological behaviour.
My hope is that we can start building neural systems that are sophisticated enough to learn from environments to help people in those environments.
We might then move on towards full fledged AI.