Traditional economics thinking has two strong principles, each based on abundant historical data:

Principle (A): No “lump of labor”: If human population goes up, there might be some wage drop in the very short term, because the demand curve for labor slopes down. But in the longer term, people will find new productive things to do, such that human labor will retain high value—in other words, the demand curve will move right. Indeed, if anything, the value of labor will ultimately go up, not down—for example, dense cities are engines of economic growth!
Principle (B): “Experience curves”: If the demand for some product goes up, there might be some price increase in the very short term, because the supply curve slopes up. But in the longer term, people will ramp up manufacturing of that product to catch up with the demand—in other words, the supply curve will move right. Indeed, if anything, the price per unit will ultimately go down, not up, because of economies of scale, R&D, etc.

Now consider Artificial General Intelligence (AGI), i.e. a combination of chips, algorithms, electricity, and teleoperated robots that can autonomously do the kinds of stuff that ambitious human adults can do—stuff like founding and running new companies, research and development, learning and applying new skills, working in collaborative teams, skillfully using teleoperated robots after only a few hours of practice, and so on.

So here’s a question: When we have AGI, what happens to the price of chips, electricity, and teleoperated robots?

(…Assuming free markets, and rule of law, and AGI not taking over and wiping out humanity, and so on. I think those are highly dubious assumptions, but let’s not get into that here!)

Principle (A) has an answer to this question. It says prices will be high. After all, if AGI can really do all the things that ambitious entrepreneurial skilled labor can do, then there will be no “lump of labor” for AGI, any more than there has been for humans. However much AGI there is, it will keep finding new productive things to do. And the prices will reflect that high value. (Incidentally, if that’s true, then it would imply that human labor will retain a well-paying niche—just as less-skilled labor today can still get jobs despite more-skilled labor also existing.)

Principle (B) has a different, contradictory answer to this question. It says prices will be low. After all, if AGI is basically a manufactured good, then manufacturing will ramp up, creating ever more AGI at a cost that decreases with scale (and with R&D). And the prices will reflect that low cost. (Incidentally, if that’s true, then it would imply that human labor, now forced to compete with a far-lower-price substitute, will become so devalued that we won’t be able to earn enough money to afford to eat.^[1])

Anyway, I sometimes see unproductive debates that look like this:

One side treats Principle (A) as an unstoppable force. The other side treats Principle (B) as an immovable wall. Instead of grappling with the contradiction, they just talk past each other. As a very recent example of such arguments, check out the blog post AGI Will Not Make Labor Worthless by @Maxwell Tabarrok, and its comments section.

Who is right? Well, at any given time,

Either the price is high, and the supply curve is racing rightwards—since there’s a massive profit to be made by ramping up the manufacture of AGI “labor”.
…Or the price is low, and the demand curve is racing rightwards—since there’s a massive profit to be made by skilled entrepreneurial AGI “labor” finding new productive things to do.
…Or the price is in between, and both the supply curve and the demand curve are racing rightwards!

The price at any given time depends on which curve is racing rightwards faster. I have opinions, but that’s out-of-scope for this little post. If people are even trying to figure this out, that would already be a step up from much current discourse.

But more importantly— What happens when an unstoppable force is slamming into an immovable wall? Common sense says: a big friggin’ explosion.

…So that naturally brings us to the school of thought where we expect AGI to bring >>100%/year sustained growth of the global economy—see for example a discussion by Carl Shulman on the 80,000 hours podcast.

I think this is the correct conclusion, given the premises. Indeed, I think that, if you really try hard to hold Principle (A) and Principle (B) in your mind at the same time, and think through the consequences, then truly explosive economic growth is where you will inevitably wind up.

Of course, that collides against yet a third principle of traditional economics, also based on abundant historical data:

Principle (C): Wait, you said >>100%/year of sustained growth of the global economy? What are you, nuts??

But, that’s where we’re at. It’s a trilemma. All three of (A, B, C) are traditional, time-tested economic principles. But it’s basically impossible to simultaneously believe all of them at once. People still try to do so, including professional economists, but I think they wind up tying themselves into knots of self-contradiction.

(Of course, those economists are still a step up from the economists who dismiss AGI as sci-fi nonsense!)

(Again, my actual main expectation is AGI takeover, which renders this whole discussion kinda moot. But if we’re gonna talk about it, we should get it right!)

^{^}
At least, probably not. We don’t know for sure how much compute and electricity it will take to run superhuman AGI, since it doesn’t exist yet. But my own guess, based on how much calculation a human brain does, is that it would probably be well under $0.10/hour at today’s prices, and lower in the future as we go down the experience curve.
Edited to add: Also, needless to say, another way that humans can turn out OK (in principle), even if there are AGIs that can do everything that humans can do but better and for $0.10/hour, is for there to be human ownership of capital or political power in the long term, and/or AGIs that choose to donate their resources to humans, etc. Whether we should expect any of those things is an important question, but out-of-scope.