The bad news is you’re in free fall, the good news is there’s no ground to hit. We’re in a weird, disorienting freefall where AI will keep improving, and that’s rightfully creating a lot of anxiety about jobs.
Imagine that automated jobs are a growing sphere that’s been with us for thousands of years. But it entered a phase of rapid expansion in the late 1700s and early 1800s, driven by the Industrial Revolution. For the last 200+ years, the sphere has grown, but humans have always been able to find jobs “outside the sphere”. So far, this has always worked.
The specter with AI in the 2020s is that “this time it’s different”. In an interview, Tristan Harris said, “The difference is that AI is automating unrelated fields at the same time, which has never happened before.” But he mic drops it there, as if the outcome is obvious.
Instead, I see two powerful forces at work. The invisible hand that has always found people jobs outside the sphere of automation, and this admittedly novel phase, where AI is fueling new, rapid, even scary, growth of the sphere. But unlike Tristan, I think we don’t know which force is more powerful, since we’ve never been here before.
I think what it comes down to is this: where is this sphere of automation? If the sphere is expanding in a fixed-sized room, then yes, soon the room will be filled, and there will be no more jobs.
But I don’t think there is a room. I think the sphere of automation is expanding into empty space, so there will always be jobs outside it. We’ll discover these jobs, as a society, through acts of human ingenuity that we cannot spell out in advance. And in fact, as the sphere’s surface area increases, there will be more places where humans can add value on top of all the automation, no matter how staggeringly large the volume of the sphere gets.
We truly can’t imagine what many of these jobs will be, no matter how long we think about it. The jobs will be products of an economy and culture that simply doesn’t exist yet. Much like every job today that involves computers would be hard to explain to someone from the 1800s.
This doesn’t say there won’t be massive disruptions to individuals and even whole occupations. But it suggests humans will continue to construct functioning societies around whatever automation exists at the time, which we’ve always done. This isn’t something to blindly believe; it’s something to watch unfold, year by year, forever.