This commentary is by Tom Evslin of Stowe, an entrepreneur, author and former Douglas administration official. It is republished from the Fractals of Change blog.
I haven’t blogged for the last three weeks because I’ve been busy programming. My last post was about how much easier ChatGPT makes programming. Shouldn’t I have been able to accomplish my coding chores in less time than they used to take? The answer is an illustration of why technology almost always creates more jobs than it destroys.
Suppose you had a business skidding stone from the local quarry to wherever in town it was used. Now the wheel is invented. You and your competitors can move five times more stone per day. There’s only so much demand for rock in town and the price you can charge per ton-mile has gone way down. Do you all go out of business? No way unless you stick to using your skid for hauling.
- Your costs have gone down. It only takes one donkey to pull a wheeled cart and that one donkey is doing five times the work two donkeys used to do. The new technology lets you improve your business.
- Since stone delivered is now much cheaper than it used to be, stone substitutes more often for other building materials and projects are built which would not have been economically feasible before. Lower prices increase demand.
- Your local stone is now affordable in the surrounding communities where it was never practical to drag it. Lower prices increase the potential market.
- Since you were an early adopter and had to repair your own cart, you might open a new business repairing or even making and selling carts.
- You may invent a better axle and start an axle-making business. There was no market for axels before there were wheels.
- If you’re a builder, you gain because lower material costs mean more building.
- If you’re a consumer, you gain because you can afford a better house.
Summary: new technology improves the profitability of old businesses, expands existing markets, opens new markets, and creates opportunities to service and provide the technology and to make the technology even better (a better axle). Lower prices increase economic opportunity and spread the benefits of affordability. Useful new technologies always have all these benefits. The greater the disruption, the greater the opportunities.
In my case I’m programming more because I can now do things I couldn’t do before. I and many other people are also programming to make AI better, which only a few people were doing before.
Suppose you didn’t have enough capital to buy a cart when it was first invented. You work harder and harder with your skid but can’t afford to feed your donkeys at the lower prices your wheeled competitors charge. One of your donkeys gets sick and you can’t afford treatment or a replacement. At best you will sell your business or get a huge loan to modernize. The cost of servicing the loan may still leave you uncompetitive.
If you were in the skid-making business, you lose that business and only survive if you have the capital and foresight to get into carts. If you were selling wood for construction, you have new competition from stone.
If you’re a donkey dealer, you get a scare because there is less immediate demand for the beasts. If you can tough it out, you might have a bonanza because there is now much more hauling and more total work for donkeys and donkeys are more profitable for their owners who have carts.
If you’re a banker who lent money to skidders, you might never get paid back — unless you’re brave enough and well-enough financed to double down and lend them money to get carts.
We can’t just look at net effect. Yes, technology creates more opportunity and eventually wealth than it destroys. That net gain is no comfort to those who chose not to or simply can’t move into the new opportunities. AI is the same kind of threat to us (over-)educated knowledge workers that globalization was to those without a degree.
Stopping AI development is impossible as well as undesirable. It’s laughable to see those who didn’t get there fast enough like Google and Elon Musk calling for a “pause.” Should we halt e-vehicle development so Ford can catch up with Tesla?
Not only individuals but institutions must adapt. Our education system had become over-priced and nearly useless in preparing people for careers before the pandemic and AI. Now education is in crisis when it is needed the most. It is not a coincidence that the most recent banks to go under were those which catered to over-compensated knowledge workers.
We learned from globalization, which did create more wealth worldwide than it destroyed and lifted billions from extreme poverty, that we can’t ignore the plight of those who lose through change. Their distress is real. We are at risk from the disaffected who are easy prey for demagogues. Paying people for not working is a boon only to drug-dealers. The losers from new techologies are not “deplorables”; we can’t just consider them collateral damage. “They” might even be “us” this time around.
I don’t know how to solve the problems that rapid technological change causes, but we ignore them at great peril. On the optimistic hand, we have greater resources available for problem solving. We’ll need lots of intelligence — both human and the other kind — to use these resources wisely.