Is latest tech ‘game-changer’ just more of the same?

Maybe they’ve already thought of this. Maybe they just don’t care.

But building an artificial intelligence system that could leave one in five people without a job might not be the best idea in the world, or for the world.

Overseas manufacturing has already proven that cheap and sometimes barely functional is the enemy of the good: high-quality, locally manufactured products have their niche, but for the majority of sales, cost seems to regularly trump quality.

And if AI can make cheaper products — even if it fails to make better ones — well, the market will quickly pick the winners and losers.

In the short term, companies that can afford to invest will win with lower costs. But almost all of us — including those companies — could be on the brink of losing a lot more.

There’s plenty of dispute about what the effect of AI will be on employment — some, like AI company Anthropic’s CEO Dario Amodei, have suggested AI could take over virtually all entry-level white-collar jobs in just five years, driving unemployment to 20 per cent. Last June, Geoffrey Hinton, the former University of Toronto professor called the “Godfather of AI,” said the technology he helped launch could dramatically affect jobs.

“I think for mundane intellectual labour, AI will replace everyone,” Hinton said on the Diary of a CEO podcast.

Others, citing AI’s mixed successes and failures so far, suggest the impacts would be much smaller. But that may be more about tamping down fears than anything else.

You can understand why companies would opt for AI when they can: salaries and benefits are virtually every company’s largest cost, and the ability to automate large sections of a business show a clear cost benefit — which translates into a more financially attractive product.

AI, even tragically bad AI, may be preferable to actually paying humans a living wage when it comes to producing the cheapest product. Social media sites are already flooded with hopelessly bad AI-written stories and faked videos and photographs that people happily share as if they were gospel. I hang my head that a large number of people who liked and shared one particular video didn’t seem to notice that the sign on the front said the vehicle was a “Scoilbus.” Or notice that all of the “crashed” vehicles remained completely intact, failing even to lose a side mirror. Or that another showed three cars coming together in a freezing-rain accident — and then breaking apart miraculously into … four cars, found a great deal of online traction.

Some economists argue that AI will give companies the financial room to hire new employees — but to do what, and where? Even companies that pretend to be relocating manufacturing operations to the U.S. because of U.S. tariffs aren’t talking about manufacturing jobs when they discuss their presidential-mollification models.

With a clean slate to rebuild from the ground up, they’re talking about manufacturing in highly automated operations.

Remember, North America’s lean into the knowledge economy came after the manufacturing economy departed for places with far cheaper labour — cheap enough to offset even shipping expenses from the other side of the world — and AI has the potential to be the cheapest labour of all. And it is likely to take the biggest bite out of the ancillary surroundings of the knowledge economy — the code writers, middle managers and entry-level white-collar workers who do the grunt work.

But job losses and unemployment numbers are only a part of the issue — and certainly not the only one companies should be looking at.

Lost in the equation of cutting back workforces is what happens to an average worker’s weekly wage. They don’t bank the money — they spend it. Buying things.

These are not just your employees. They’re also your customers. And the knock-on effects of their inability to buy will radiate. With the comfort of years of low interest rates already luring people into borrowing far more than they should, there isn’t a cushion left.

AI may help make it cheaper to build a Tesla, but it won’t buy a single one. And neither will anyone replaced by AI.

I remember talking about the effects of outsourcing journalism to an unfettered world of social media, and the impact that was having on skilled, trained journalists.

Learn code, we were told haughtily.

But look where we are. We’re in the midst of planning to put all kinds of eggs in the AI basket, without realizing that, if it works as planned, we’re short-term thinking our way into a recession, or far worse. The very rich will get richer, or at the very least, be able to buy up most anything at bargain prices, and the rest of us will get poorer.

Potentially, much poorer, as we compete for fewer and fewer jobs with the only thing we have left to trade — accepting lower wages to be competitive in an environment with too many workers and too few jobs.

Either millions upon millions of people in professional positions lose work and a similar number of customers — and taxpayers — essentially vanish from the economy, or AI turns out to be a massive and unworkable bubble, and billions of dollars of investment by some of the largest companies in the world — not to mention large blocks of pension savings — goes up in smoke.

Neither is particularly palatable. Perhaps the best we can wish for is that this latest automation “game-changer” ends not with a bang, but with a whimper.

Russell Wangersky is the Comment Editor at the Free Press. He’s not an economist, but he’s seen this movie before. He can be reached at russell.wangersky@freepress.mb.ca