There are many examples of new, transformative technologies that increased the efficiency, effectiveness, and—most importantly—scope of the underlying processes. Agriculture used to employ at least 70% of the population; it’s now about 2%; that didn’t cause 68% unemployment. James Watt’s new steam engine cut coal usage per unit of work by 75%; coal consumption soared. This particular case gave the world “Jevron’s Paradox”. The Industrial Revolution increased worker productivity by between 50 and 100 times; we employed more people, building more products. The introduction of internal combustion engines dramatically decreased the cost and increased the speed of transportation; we shipped more stuff at greater distances than ever before. Consider the impact of computers on countless clerical activities; far more jobs were created than destroyed.. Computers once needed to be coded in assembler. The introduction of compilers was supposed to devastate programmer employment; of course it did the reverse. Finally, one field that was supposed to be devastated by AI was radiology. However, hospitals that employed AI to analyze scans ended up hiring more radiologists as the cost of scans declined and usage increased. The efficiency of the radiologists increased and their role evolved, but eventually there were more of them. It’s not that these revolutions were not without a certain amount of painful adjustment. However, neither utopia nor dystopia resulted, and we ended up better after than before. On the other hand, each revolution had its unique aspects, and it is possible that AI is one of the most dangerous developments we face as a race. Handling the problem properly will require us to understand the broad spectrum of risks and rewards of this challenge. If I had one suggestion relative to AI, it would be that we must build a moral compass into such systems, not as an add-on or set of external constrains but as a deeply integrated and dominant “super ego”.