The Jobless Future
Samir Mehta | February 26, 2020
[hidden by request]
Scott Hardie | March 3, 2020
All great points. I too have a lot of faith in humanity's ability to solve big problems. I believe that the urgency of the response tends to follow the urgency of the problem. As the effects of climate change intensify, so too will our response to the problem. I should have faith, too, for this "jobless future" problem as well.
I've heard of studies that show a correlation between employment and self-esteem, regardless of the job itself. You may not like your job, but you feel good about having it and doing it. Not being employed in a capitalist society, unless one has "earned" the right not to work through retirement or becoming wealthy, is a stigma that we internalize, causing unhappiness and low self-esteem. This must inevitably factor into any future solution to the problem -- even with something like UBI, people still need to feel like they're contributing to society.
Want to participate? Please create an account a new account or log in.
Scott Hardie | February 26, 2020
I started to get into this in a separate conversation about Andrew Yang and universal basic income, but I think it deserves to be broken out separately. What I want to know is, do you see any solution to the economic threat posed by the rise of robotics and artificial intelligence? Because I think we're in tremendous danger. I'd say the problem is second only to climate change in terms of threats to life as we know it; not even the rise of autocracy is this bad for the world.
Basically it comes down to this: Any job that can be done better by a machine will be done by a machine, and machines will keep getting better and being able to do more jobs, so therefore, eventually there will be no more jobs for humans. We've already seen some of this in manual labor jobs; it's not hard to imagine any job currently involving manual labor to be replaced by a smart robot sooner or later. What we haven't seen much of yet, but what I consider inevitable, is jobs requiring mental labor to be replaced as well. Eventually, AI will do a better job at practicing medicine than a human doctor. Your stockbroker will be replaced by AI. Your therapist will be replaced by AI. Your lawyer will be replaced by AI. Your kid's teacher will be replaced by AI. Your news reporters will be replaced by AI. Even people who code AI will be replaced by AI (more on that below). There is no job that cannot eventually be done better by a machine, given enough time for the technology to develop.
We are completely unprepared for this future. Think about the massive dissatisfaction that led to Trump and Brexit, how people in rural communities and blue-collar professions already feel left behind in a world of global free trade and Wall Street getting rich while Main Street falters. Now imagine that vacuum of opportunity happening everywhere, including major cities. And imagine there not being any more jobs at all, instead of merely a few. Riots would be a mild response. Do you know what the difference would be between this future and the Great Depression? The Depression ended.
Whenever I ask people about this, they often repeat that old conventional wisdom that technological improvements yield new jobs. For instance, when cars came along, all of the people who worked in a horse-based travel economy were screwed, until eventually cars created more jobs than they eliminated. But I don't think that's relevant in this, because we're facing a future where ALL jobs are eliminated. There would be no economic need for humans to work, other than their own self-sufficiency. It wouldn't all happen at once, so there would be plenty of new opportunities along the way, but there is some inevitable future date where there is no more way for a human to work for income, in any conventional sense of work.
I also reject the argument that we'll get jobs programming and operating these intelligent machines of the future. Gradually, we're going to see human-to-machine interfaces simplify to the point that anyone can instruct a machine. The history of programming has gone from obscure (binary and punch cards) to semi-obscure (early languages like assembly) to using semantic construction (today's modern languages). The languages of tomorrow will let you instruct the computer in plain English, so you only have to think about the logic of what it's doing. And the future after that is one where you simply describe the problem to the computer and it develops its own logic and its own solution. (If you think that's crazy, people are already working on this.) For example, if you need a ride to the park, today you can download Uber to your phone and enter instructions in the app's interface, but in the future, you would simply tell your computer verbally "I need to go to the park" and it would figure out a solution itself and get you there with no need for humans to have coded any specific app. It's a fascinating future from a technology-consumer perspective, but a lousy one from an employment perspective.
Capitalism has been a tremendous engine for wealth, but it has limitations. We have known for a long time that we need to constrain monopolies or they'll ruin capitalism. We're waking up now to the reality that the natural world cannot be endlessly exploited and polluted for profit. We haven't yet grasped that IT creates more economic unfairness that can be as ruinous as monopolies; consider the massive corporations that can track and predict and manipulate your every move, restrict your legal rights with 1000-page service agreements that no sane person can read, and so on; ordinary consumers are already getting screwed in this era. Now consider a future in which the only wealth to be had is in the ownership of robotics and AI because it eliminates all opportunity. There wouldn't even be marginal short-term opportunities in innovating new applications for it or finding new ways to use it cost-efficiently, because the ownership of it would be restricted to the Jeff Bezoses of the world. I cannot see AI becoming commoditized to the point where everyone can own a part of it, because advances are so powerful and expensive that ownership is likely to remain in the hands of the few. Capitalism as we know it is unprepared for this eventuality.
Universal basic income is one possible solution: Force the few people still getting rich to share their wealth directly with the people. I don't love this for a number of reasons, but at least it's trying to help.
Another possible solution is technocratic socialism: A government takeover of the ownership of AI, transferring ownership to the people, so that we can all benefit from it. One of the usual arguments against socialism, that it stifles innovation, is somewhat irrelevant in a world where AI can evolve on its own. And another, that it breeds corruption, is also irrelevant in a world without wealth. But I'm not thrilled by this solution either.
I guess what worries me is that I don't see a good way out of this future. Right now on the campaign trail, politicians talk about job training or financial assistance for people economically displaced by automation or other technological trends. People recognize that machines are taking over factories and warehouses. Nobody seems to be thinking ahead to a future where machines will take over every job. Nobody seems to recognize the scale of the problem.
Can you see other solutions to this? Or am I worried for nothing?