My work on Epistemology and Artificial Intelligence converged when I faced a specific problem in my Epistemology research. Specifically, I needed a method or process by which the human brain would prioritize certain data. Let me put it in context:
One of the leading theories in cognition is that our brain is designed to classify experiences into categories. This theory is attractive because a lot of what we do involves classification. Unfortunately, it has some severe problems outlined on the Eviduction site in The Classification Challenge (new window).
I was forced to find an alternative solution and ultimately did. Indeed, the solution was so elegant that it was able to explain a significant number of logical fallacies (see the logical fallacies section of Eviduction for more info). When combined with my epistemological solution, I felt that I had a very comprehensive explanation for human behavior (especially the irrational parts).
I figure that I had two interesting popular non-fiction books ready. All I needed to move forward was to find a Neuroscience basis for my theory on cognition. I dived into the field with great excitement only to find that the field was still working at a much lower level. Nothing in the field directly contradicted by theory, but they also weren’t talking in the same terms. By coming into the problem backwards, I was confident that I had the answer to the BIG question. Unfortunately, I had take on the problem directly, leading to my work in Neuroscience.