One thing that interests me is the "inductive leap" implicit in all modeling situations. You have empirical risk minimization, Bayesian, MML, MDL and others all taking a different approach to taking the leap. You also have the "mind projection" fallacy Jaynes liked to talk about.
Understanding what modelers do is a good way to understand what AI will be doing.
Comments
One thing that interests me is the "inductive leap" implicit in all modeling situations. You have empirical risk minimization, Bayesian, MML, MDL and others all taking a different approach to taking the leap. You also have the "mind projection" fallacy Jaynes liked to talk about.
Understanding what modelers do is a good way to understand what AI will be doing.