My Hypothesis
- My Hypothesis is that the best way to get to real AI is by
following the human example.
- I don't claim that this is novel.
- I don't want to just be inspired by humans, I want
to have models of them.
- There is a lot of work in cognitive architectures (e.g. ACT and Soar).
These are based around psychological models.
- I think this stuff is great but does not constrain the problem
enough.
- I think we should build cognitive architectures out of neurons:
Neural Cognitive Architectures.
- There is a lot of work in modelling neural behaviour but it rarely
goes to the psychologial level.
- I am working on bridging this gap.
- Moreover, it's not just enough to build models.
- We need to build agents out of neurons that function in
environments.
