Natural Language Processing with Neurons
by Chris Huyck
c.huyck@mdx.ac.uk
- People, macaques, rats and most things that think do their thinking
with their neurons.
- There are lots of things that we don't know about how that's done,
but we're finding out more all the time.
- One way to find out more is to build computational models. That's
what I do.
- On the other hand, there are some things that we AI people have
real problems with.
- Language is one of them, and in language semantics is one of the
trickiest.
- Neural simulations provide, I think the best, potential solution
to the problem.
- Build an agent in neurons. Let it wander around, use and learn an
environment. It will then be able to learn the semantics.
