Techstars Startup Week Boise 2019

Machine learning as a concept has existed as long as practical artificial intelligence, since the invention of the Perceptron in 1957.  But you would be correct to say that machine learning's popularity within the AI community has waxed and waned over the years, with a lot of interest in recent years due to the successes of deep neural networks. Today, a lot of these and other techniques are still used as part of more complex AI systems, though machine learning can often be used in some part of the system to help it improve.

So what was the AI community investing in when machine learning seemed less promising? Today we'll discuss logic and automated reasoning in the early days of AI research, how we managed to encode a lot of simple facts into a machine, and how those machines use inference and deduction to reach an answer.