A collection of potential research realizations I had. Perhaps to be explored in the future.
The 'Knock-down' experiment of symbolicism and connectionism. Find a semi-Godellian process in the brain and see whether both of them can handle it. We know symbolism generally cannot, but can connectionism?
The philosophy of infinite spaces, and claiming much anything about them at all could be flawed. "Representative examples" of the infinite space as said by the error curve, etc.
Are ANN's Turing-Complete?
All ANN's rely on 'representative examples' for your inputs. For modelling real phenomena (say economics) such samples do not really exist for the process is not understood. There must be some math to determine what are truly the most 'representative samples' for a given set of data. Perhaps that could be useful.
Post the Pragmatic prevail of machine functionalism to CogNet
TODO: (in order to be famous and rich in academia
Write a book (and write it well) that anyone can read. Doesn't have to be good.
Have flashy presentations.
Model Ex: Steven Pinker.