it is very hard to think of solutions to problems of such magnitude. My goal has been to highlight the problem, pushing towards societal agreement that we have a major problems on our hand. we need to bring many people to this discussion: computing researchers, economists, philosophers, and the like. Right now, the job problem is just starting to be recognized as a major problem.

With respect to climate change, we are at least holding a vigorous societal discussion. We are not holding a similar discussion with respect to machine intelligence.

I agree that society tends to ignore such long-term problems.

Economists promise that there will be a stream of new jobs to replace the lost jobs, but I do not find their promises to be supported by a serious analysis of machine intelligence and robotics.

Our society makes a huge investment developing machine-intelligence technology, but practically very little thought and investment in thinking about the consequences.

Steven Clausnitzer asked me to join this discussion. I see that the discussion here is focusing on the inevitability of machine intelligence. But this issue, as well as the issue of trying to stop the march to machine intelligence, was not the focus ofthe article. The focus of the article was the hypothesis that machine intelligence will cause, and is perhaps now already causing massive job losses, which will not be compensated by the development of new jobs. If you accepts this hypothesis, then we could start the discussion of what should be done about it.

Moshe Vardi