AI via Neural Processing

by Chris Huyck

www.cwa.mdx.ac.uk/CABot/CABot.html

c.huyck@mdx.ac.uk

  • I got my PhD from the University of Michigan in 1994. The topic was a psycholinguistically plausible natural language parser.
  • Shortly thereafter, I was exploring ways to resolve PP attachment ambiguities.
  • I had a nice idea about an engineering approach,
  • but I thought I'd use Cell Assemblies (CAs) to solve the problem.
  • I'd known about these, had thought someone had solved the problems, and thought I'd just plug them in.
  • Also, I thought they were the basis of human thought, so I'd be doing good old fashioned hard AI.
  • It turned out, CAs were not figure out, and I've spent the last decade or so figuring them out.
  • I'm not there yet, but...