From Steam Engines to Superintelligence: The Next Growth Revolution

Aug 6

BOOM!

Imagine a world where the economy doesn’t grow slowly—it goes BOOM. Not a regular boom like a good quarter for Apple or the Industrial Revolution “boom,” but a boom so big your brain does a cartoonish backflip trying to comprehend it.

That’s the wild ride this article takes us on. You see, for most of human history, economic growth was a sloth. From year 1 to about 1700, global output grew at a glacial 0.1% annually. That’s one-thousand-year-double-your-moneyspeed. Then came steam engines, spinning jennies, and the idea that maybe kids shouldn’t work in coal mines, and growth started jogging. In the 20th century, it started sprinting—2.8% a year! We felt rich.

But what if AI—specifically AGI (Artificial General Intelligence), which can do pretty much anything a smart human can—turns that sprint into light-speed? Silicon Valley’s starry-eyed optimists think AGI could push growth to 20–30% a year. That’s the kind of growth where GDP doubles every three years, not every 25. It’s like the economy chugging Red Bull, forever.

How?

AGI wouldn’t just automate your desk job—it would innovate. It could solve problems, write code, discover cancer cures, build better AGIs, and figure out how to grow tomatoes on Mars. Unlike humans, it doesn’t sleep, procrastinate, or spend hours arguing in meetings. It just builds and learns and builds again. Once AGI starts improving itself, you get what some call “recursive self-improvement.” That’s when the “boom” gets scary.

But what does this mean for regular people, like, say, you?

Meaning…

First: jobs. AGI might make humans obsolete for a lot of tasks. Wages could tank because why pay Bob $60K to do accounting when RoboBob does it better for $6/month?
Second: capital. In this future, the winners are the people who own the machines. If you don’t own a piece of the AI pie, you might be stuck trying to make a living in jobs AI can’t (yet) do—like plumbing, babysitting, or teaching yoga to toddlers.
Third: stuff. The stuff AI makes—digital entertainment, food, manufactured goods—could become insanely cheap. Think near-free Netflix, apples, and iPhones. But the stuff that still needs humans—childcare, massages, date-night sushi—gets expensive. That’s Baumol’s cost disease: the more essential and un-automatable the service, the more it costs.
Fourth: markets. With the economy possibly rocketing into the stratosphere, investment becomes weird. Interest rates might soar. Stocks could either moon or crash depending on whether earnings beat the gravitational pull of higher discount rates. The models disagree, like a room full of economists at a wine tasting.

And yet—markets don’t act like this future is coming. Despite ChatGPT and DeepSeek blowing minds, bond yields haven’t budged. Investors seem to think this AI thing is either hype, far off, or will get stuck in a web of human bureaucracy.

What Now?

But here’s the kicker: if the future really is “information producing information producing more information,” we could be a few breakthroughs away from an economic singularity—a word that basically means all bets are off.

So… maybe invest in a robot. Or a yoga studio. Just in case.

Read the full article in the Economist What if AI made the world’s economic growth explode?