Machine Learning

http://www.viralviralvideos.com/wp-content/uploads/2014/08/humans-need-not-apply-300x168.jpgAnother thought-provoking post at kottke.org about the effects of relentless automation on social equity.  We might blame China or capitalists, but the real job tumult in this country is caused by the fact that machines are faster/better/cheaper than humans are.  If there's a job that doesn't need to be done by a human, it is either going to be done by a machine or we are going to invent a machine to do it.

That counts as progress.  But it comes with a need to figure out how to harness that progress in ways that benefit all people.  That's a tough challenge, and there are no shortcuts to real solutions.  Just ask Europe, whose work policies are the envy of many of my liberal friends but which suffers from an awful entry-level labor market as a result.  (Consider this news story from three years ago, in which McDonald's hiring binge in the US was contrasted with its deployment of order-taking kiosks in Europe.)

There has been a lot of talk about becoming a barbell economy, in which there are a few high-end jobs and a bunch of low-skill jobs and not much in the middle.  It might be even worse, if innovation makes even the low-skill jobs more easily done by robots than by humans.   Again, I don't think the response is to kill the machines or jerry-rig the markets through over-regulation.  Whether it is through personal effort or public policy, I think the answers lie in figuring out what humans and only humans can do - no matter how advanced the machines get, there will always be something humans can do better - and make sure we are preparing ourselves and others to do those things. 

Post a Comment