AI Transitions

A different way of thinking about the advances of AI is to discuss the transitional techniques or technologies versus describing them as waves.

The figure illustrates advances in AI which became transitions for powering new AI solutions. Expert systems which use the knowledge of human experts for programmers to specify explicit rules and logical relations for solving a problem represent an early advance in AI. However, the hype of AI backfired as the promises with the technology at hand failed to meet expectations.  Then in 1966 government funding of research on NLP halted as machine translations were more expensive than using people. Governments canceled AI research, and AI researchers found it hard to find work, there was a dark period for AI, a winter.

AI was a dream, and it became an accepted premise--AI doesn't work.  This thinking continued until analytics became the moniker.  Machine learning techniques and predictive analytics were born and the 2nd AI transition was underway.  Then a new transition, deep learning, new algorithms like backpropagation, GPUs, and massive amounts of data advanced AI. There has been an explosion of new thinking by industry and academia, and there is a renaissance going on today, largely because of deep learning.