← Back to home

Robots will Take Our Jobs, But Not Ours

2019-2-7

Just based on the amount of new automation in manufacturing and supply chains and various other simple-services, it’s easy to say “wow, these robots and low-form AIs are going to take some of our jobs.”

From here, there are three main ways to go:

  1. Yes, these robots will replace current low-skill jobs. But technology has put humans out of work in the past - see the industrial revolution (or really any piece of new technology for that matter). People always adapt and find new jobs. This is nothing different.

  2. These robots will probably replace all the low-skill jobs, and people won’t be able to find new ones. The problem with the above argument is that there isn’t going to be new low-skill labor for those jobs to move to - as the robots could just take these jobs as well. It’s likely that the only jobs left are going to be the real high-skill ones: programming, or CEOs, or something similar.

  3. Robots will replace all jobs.

Please note that I’m ignoring the case where the robots kill us all (or something similarly terrible occurs) - on purpose or just due to a bit of research getting out of hand.

In any case, I’m not sure which of the three are most likely, but let’s make some claims about them:

  1. The new jobs that will likely emerge when current low-skill labor will likely either be easily automated, or either will not be considered low-skill jobs. The reasoning seems pretty straight-forward: a low-skilled job is usually considered to be one that is “mechanical” - and what better task to automate.

  2. Assuming humans don’t enhance their own abilities, humans will remain, on average, about as low-skilled as past generations were. As such, it seems that those who fall in this category will likely be put out of work, if robots manage to replace low-skill jobs. By the above point, it’s unlikely they will be unable to move to new positions.

  3. The definition of “real high-skill” jobs is both a) fluid, and b) going to change dramatically in the next 50 years. Really, being a programmer is just satisfying a specification efficiently (and AIs will likely be able to optimize in a well-defined domain very well!). Fine - being a CEO is just optimizing for NPV cash-flows in a much-more nebulous environment - but who knows how nebulous it is, really.

The reason I make the above points is two fold. First, I think that even with the most pessimistic view of automation, we need to account for a case where some large portion of the population is permanently job-less. Second, I don’t think anyone can be sure that they don’t fall within that group. The real point being: we definitely have to reconsider the role of the human in society, and that human being might be you - no matter how safe you’re sure you are.