[JohnnyB]

The discussion in this thread about whether or not historical examples of automation have led to increases or decreases in employment seems to me to miss the real significance of current trends in automation. Automation has always led to job losses and increases in frictional unemployment, but that has been responded to with people learning new skills and then working with the aid of automation to enhance their productivity. This has generally led to a more educated, skilled and productive workforce because the speed of workforce improvement in knowledge and skills has been fast enough to keep up with the speed of improvements in automation. The big thing that makes current changes different is that we are rapidly approaching a point at which the capacity of the workforce to re-skill fast enough to keep up with advances in automation is coming to an end. When we reach that point, frictional employment will rapidly turn into persistently discouraged workers. Speed of change and speed of possible adaptation are the key.

If we're not smart enough to learn and adapt faster than automation at a cost-effective price, we're not going to be employable. This is why things like Elon Musk's new venture into creating a means to augment human brains with integrated AI is both a serious bet worth dumping tonnes of money into for many investors and is the type of moon shot that fits ideologically with Musk's other ventures.