Yes, it may take a long time to get to full self driving, but at least this time machine learning is actually giving us useful stuff unlike the last AI winter in the 1980s.

I am currently working on an application which transcribes spoken words, and it is just amazing what sort of capalities we get out our fingertips today.

Today we are at a stage, which I haven't written about but which is very interesting in the history of technological development. That stage when a new revolutionary technology becomes available but we have not quite figured out how to use it yet.

E.g. when the telegraph was made, they initially used it to play chess long distance. Using it to report news from distant places, for quick decision making in diplomacy, to relay information about stock prices or company orders, actually took quite a number of years to realize.

Likewise when the electric motor got made. They made it as big bulky machines just like a steam engine. It took them 30-40 years realize that electric motors could be used in entirely different ways than steam engines because they could so easily be shrunk in size. That is what gave us portable electric power tools But it took decades before they saw that possibility.

Machine learning today has gotten extremely powerful, but I think we have not yet realized the potential. Often we get hung up on obvious things like self driving cars, when in reality the real revolution may be somewhere else. We just haven't seen it yet.