Not hugely. I think we're headed for another "AI winter" (https://en.wikipedia.org/wiki/AI_winter https://en.wikipedia.org/wiki/AI_winter), which will be eithe

Not hugely. I think we're headed for another "AI winter" (https://en.wikipedia.org/wiki/AI_winter https://en.wikipedia.org/wiki/AI_winter), which will be either our third or fourth depending on whether or not you count the one after IBM Watson. But who knows how long it takes for that; you can know that the bubble would burst, but not when. Which is a shame, because neural networks are useful tools to have in your toolbox, they're just not the entire toolbox. I wish we could just use them like "finite state machine" or "model-view-controller" or any other useful-but-not-overhyped algorithm or code pattern.