Jobs of the Future

What jobs do you think we’ll still be doing in 20 years? I mean that both in the sense of us personally and society at large. I’m kinda curious for my own sake. There’s no guarantee, but odds are pretty good that I’ll still be around then to worry about it. In fact, the odds aren’t bad that I’ll still have a bunch of productive years left in me. My mom’s family seem to make it to their late 80s or early 90s if they take care of themselves. So I may be looking at starting another career in twenty years. If I’m going to do that, I’ll need to start planning for it. I may have to go back to school. Even if I keep doing software, I’ll have some learning to do. Being a programmer is kinda like being a writer. There’s a part of it that’s just a craft unto itself, but there’s also domain knowledge - developing an understanding of some part of the world.

Besides, it’s just fun to speculate about this stuff.

So, what are the ground rules? What assumptions am I going to make about the world of twenty years from now? For the sake of this exercise, I’m going to say that things chug along much as they do now. We may have ups and downs, but we’re basically prosperous. We don’t get completely screwed by climate change or disease pandemics or any of the other things in the “When the Zombies Come” category. Even without that, we may still get blindsided by something. Twenty years ago, most people thinking about this would have missed, or grossly underestimated, the impact of the internet.

Given all that, the jobs of the future will be those things we can’t automate, can’t do remotely, or are willing to pay a premium for.

Computing power is still growing exponentially. Software is still getting better, though at a much more modest pace. Robotics are getting better, and starting to drop into the consumer market with things like the Roomba. So if machines are getting more powerful, what will humans still be doing? What are machines fundamentally not good at? That really comes down to judgement, creativity, and empathy.

Computers are great at statistical analysis, and they can follow any rules you give them, but they don’t understand things in the way that people do. They can point out correlations, but they don’t really understand causality; that’s hard enough for humans. They don’t have common sense. Much of our judgement is based on things we’ve learned about the world just by osmosis, by living in it. A computer only knows what you tell it. It doesn’t know what it doesn’t know. People have some sense of that, the “known unknowns”. We also have ways of chipping away at the “unknown unknowns”: We people-watch, window-shop, and browse through bookstores. Computers don’t do serendipity.

Computers can follow instructions, but they can’t make up new ones. They don’t have any way of recognizing that the old instructions aren’t working anymore. You can’t give them goals, only actions. They have no way of going meta, of making sure that the rules they’re following still make sense in the larger context. They have no context.

They don’t have emotions, either. They can fake it to some extent, but they’re fundamentally lacking in all of the weird biological crap that we have going on. There are some things that people will only trust people for. Any job that involves working with other people requires some level of empathy, but there are some jobs that focus on it: Police, counselors of any sort, and probably lawyers, in an odd way.

Let me give you a definition of empathy: It’s the ability to develop a model of another person’s state of mind and update it in real time in response to things as subtle as tone of voice and facial expressions. Arguably, that’s the single thing that our brains have evolved to do. For tens of thousands of years, if not longer, the biggest evolutionary threat to people has been other people. Being able to predict their behavior is our single most important survival trait. Our brains have specifically evolved to do that modeling and perception, and it still takes most of us somewhere north of twenty years to really get the hang of it. You can’t program that.

So, if we can’t automate a given job, if we still need a person to do it, can they do it remotely? What jobs will still be done, but not by anyone around here? Which will be outsourced to call centers in the mid-west, or to the other side of the world? Which require a physical presence?

We’ll never outsource firemen. Ditto for police, most health care professionals, and anyone who works on your house. There’s some common thread there of physical and social infrastructure. It’s interesting to brainstorm about how you might automate or remote some aspects of these jobs, but some part of them requires that you get your hands dirty, at least metaphorically. You have to be there to really understand what’s going on. We may eventually have a world of telepresence androids, but I think that technology and cost put that well over twenty years out.

Jobs that require a high level of empathy pretty much require you to be there. Even with immersive videoconferencing in super hi-def, you may still miss something by not being there. There’s also two-way communication going on. You need other people to be able to read your expression and body language as much as you need to read theirs. You’ll also need cultural background; even if you could remote it, you couldn’t go far.

Also, we’re good at doing sound and sight remotely, but that leaves taste, touch and smell. There are definitely jobs that rely on taste, though probably not a lot. Ditto for smell. I can’t think of any that specifically require touch, but anything you can’t do in welding gloves probably can’t be done remotely.

There are also jobs where we’re willing to pay a premium to have a person there, mostly businesses built on convenience. The ones that jump to mind are restaurants and retail, but I suspect I’m missing some. Who among us won’t pay extra for coffee from the place with the cute barista?

So what will I be doing? I don’t know, but I’m getting a better idea of what I need to pay attention to. It’s kind of encouraging, because it boils down to the human parts of my job. In a sense, I’m a translator between people and machines. I think that the human side of the equation is more interesting.