The Pandemic Has Us Hurtling Toward an Era of Automation. Tech Reporter Kevin Roose Says That Doesn’t Have to Be Bad for Humanity.

Kevin Roose spends a lot of time thinking about how to stay ahead of the robots.

The New York Times technology columnist describes his new book, Futureproof, as a "guide to surviving the technological future"—a future that, due to the events of the past year, is coming at us even faster than before.

Even before the pandemic, it was getting harder to make the case for hiring humans for many jobs. Then, COVID-19 hit and humans became huge liabilities. Any in-person interaction became a potential brush with death. Researchers at the Brookings Institution say the pandemic is going to accelerate the pace of automation, both because of infection dangers and because the resulting recession will drive employers to cut costs more aggressively.

It sounds scary, like something along the lines of Battlestar Galactica or The Matrix. But Roose is actually optimistic about how tech can improve our lives. It just depends on whether humans make the right moves.

WW: The three things that scare me most in the world are climate change, job automation, and partisan gerrymandering, in that order. After reading your book, I think I'd better make job automation number one. What do you think?

Kevin Roose: Climate change is probably more dire, but it's good to have some healthy concern about all three of those things. I think it's important to balance pragmatism and realism with the understanding that all is not lost and there are actually things that we can do.

Sometimes I feel a little paranoid worrying about this because for centuries people have been predicting that machines would replace us. But, before the pandemic at least, unemployment was at record lows, even as the machines rose to greater power.

The thing that's different this time is the pace of change, and the nature of it. Automation and AI are happening really, really fast. Compared with previous tech transformations, which took decades, this one is happening in years. The typical defense is that AI is going to do all the worst parts of your job, and the parts that are left over are going to be the fun parts. But in practice, that's not how it works. There's a whole genre of software that I call "bossware" that uses AI to track worker productivity. It's putting workers at the mercy of these algorithms, and a lot of jobs are getting worse rather than better. If you're a worker in an Amazon warehouse, you are taking instructions from machines. You wear a bracelet that tracks your productivity.

You write in the book that we all need to make sure we're not endpoints. What does that mean?

In software development, an endpoint is something that connects one machine to another. It's the connective tissue between one algorithm and another, and that's essentially the role that a lot of workers are playing in the economy today. Whether it's DoorDash delivery people or workers in Amazon warehouses, or even journalists, there are people whose job it is to take information from one machine and plug it into a different machine.

The journalism thing is hitting a little close to home, Kevin.

People are very overconfident about their own irreplaceability. Something like 75% of Americans think that automation will destroy jobs, but only 25% think that it will destroy their job.

I have a friend who's an actuary. He had to go to school for a really long time, and now he makes really good money trying to figure out how long people are going to live for big pension funds. After reading your book, I got to believe he is in the Borg's crosshairs.

Oh yeah. I'm actually shocked that his job hasn't been automated already. It feels like saying you have a friend who's in buggy-whip sales.

My kids are 17 and 15. What should they study in school?

There isn't any one degree that is more robot-proof than another, but there are more robot-proof skills than others. People working in technology often believe that the only employable workers will be engineers, computer scientists, and math majors. That's just not true. It lulls people into a false sense of security into those in those fields. The things that are going to be durable are what machines can't do: the softer, human skills like leadership, collaboration, effective communication and public speaking. That's where we should be focusing our educational system.

The American middle class was built in large part on well-paying factory jobs, which are being automated by the thousand every day. Are we going to be able to replace those jobs with everybody becoming a comedian or an influencer or something?

The thing that created the middle class in the 20th century was not technology, it was how workers responded to technology. Unions could say to factory owners, "OK, you're making many more cars, steel pipes and whatever because of these new factory robots. We, the workers, should see some of that." It wasn't just the technology that made those workers better off. It was their ability to advocate for themselves. That's what you're seeing at Amazon warehouses right now. Workers are saying, "You guys are making so much money with automation, and we need that money to be fairly distributed."

So, if we get the political part of this right, it's going to be a worker's paradise?

I think it could be amazing. We could be working three days a week. We could be more prosperous. We could be less stressed out at work. We could have many more fulfilling hours in our day. It matters how we do this. I don't think we should stop using technology and start farming by hand again. But I do think we need to understand that this technology is coming and set up systems that are going to allow workers to benefit from it. The technology is not the issue here. It's the humans who are using the technology to exploit other humans.

Sometimes I think, in this economy, you need other things working for you. You need rental houses kicking off rental income, stocks kicking off dividends, and an army of bots doing…what? Mining Bitcoin? Is there way for an individual to create a droid army and turn the tables on the machines and the huge companies that control them?

We all have them. I've got Siri and Alexa and the Netflix algorithm. We all have armies of robots that we encounter every day. It's interesting to ask who is actually in charge in those relationships. Am I Netflix's boss, or is Netflix my boss? Is Netflix telling me what to watch and what I'll enjoy? Are we letting the robots choose things that we actually want to buy and pay attention to, or are they sort of steering us in ways that maybe we don't want?

You talk about avoiding "machine drift," where we are shaped by recommendation engines, where, if I read it correctly, our personalities become determined by Instagram and TikTok. I can see how avoiding these things is important to having a fulfilling, self-actualized life, but does it also help us keep the machines from taking our jobs?

We are making ourselves easier for robots to replace, full-stop. We are not making it very hard on them because we have outsourced so many of our decisions, personality traits, and values to machines already. There's an entire generation of kids who have grown up internalizing the incentives of algorithms, behaving in certain ways because it will get them more views, likes and shares. Part of making yourself hard to replace in the labor market is making yourself more human and distinctive and not behaving in formatted and predictable ways. "Machine-directed" is the shorthand I use for turning your free will over to algorithms and AI. I think we've all done that already to a certain extent. Clawing that back is important. You have to know yourself better than the machines know you.