Commentary | AI Isn’t Taking Our Jobs. We’re Deciding What It Is For

People are losing their jobs to AI. That is already happening. But the debate is often presented as a false choice: either we stop AI from replacing people, or we accept mass unemployment as the inevitable price of technological progress.

There is another possibility.

The real question isn't whether AI can replace people. Increasingly, it can. The question is whether replacing people is what we actually want to use it for.

AI is different from many technologies that came before it. Machines traditionally replaced physical labor or automated repetitive tasks. AI is moving into work we considered inherently human: writing, coding, research, analysis, design, translation, customer service and parts of professional decision-making.

That creates an uncomfortable economic reality. If a company can produce the same amount of work with 100 employees that previously required 150, there will be enormous pressure to employ 100.

And telling those 150 people to "learn AI" doesn't solve the problem. If AI makes everyone 30% more productive, the company may simply discover that it needs fewer people.

But productivity is only valuable to society if we decide what to do with the gains.

What if AI's purpose was to make jobs better rather than make workers unnecessary?

If ten people can now do the work that previously required 15, a company could fire five people. Or it could give the remaining employees shorter working weeks, more vacation, higher pay, or more time for work that requires judgment, creativity and human interaction.

AI could turn productivity into time.

That isn't such a radical idea. Technological progress has repeatedly made societies more productive, and societies have sometimes chosen to turn those gains into leisure. The five-day workweek is one example. We could have used industrial productivity simply to make people work faster. Instead, some of the gains eventually became time away from work.

AI could give us another opportunity to make that choice.

The problem is that individual companies may have little incentive to do so. If Company A uses AI to cut 30% of its payroll while Company B uses it to give employees a four-day week for the same pay, Company A operates at lower cost and undercuts Company B on price. Unchecked market competition naturally rewards the payroll cutters.

So this cannot be solved entirely by asking companies or workers to make better choices. It becomes a question of how societies distribute the productivity dividend.

That could mean shorter working weeks, profit-sharing tied to AI-driven productivity, incentives for companies that augment workers rather than replace them, retraining for genuinely new roles, and stronger safety nets for people whose professions actually disappear.

And then comes the harder question: what happens if AI becomes capable of doing so much economically useful work that there simply aren't enough jobs for everyone?

We have tied employment to far more than a paycheck. Jobs provide status, structure, social interaction and a sense of purpose. We have built an economic system around the assumption that a decent life requires selling your labor.

AI could challenge that assumption.

That doesn't necessarily mean a future in which nobody works. Humans are unlikely to stop wanting to create, teach, investigate, build and care for others simply because machines can perform some of those tasks. But it could mean that having a job is no longer the prerequisite for having a decent life.

There is also a middle ground that gets surprisingly little attention: humans and AI doing more together.

A doctor using AI could see more patients while spending more time talking to them. A journalist could use AI to search vast amounts of information and spend more time investigating. An engineer could test ten ideas instead of two. A teacher could spend less time grading and more time teaching.

The difference comes down to what happens when AI makes a worker dramatically more productive.

Does management say, "Great. Now we can fire 20% of you"? Or does it say, "Great. Now let's see what we can accomplish that we couldn't accomplish before"?

That is not a technological decision. It is a human one.

AI may eventually give humanity something previous technologies only gave us in small doses: the ability to produce more while working less. Whether that becomes a world of mass unemployment and concentrated wealth, or one of better work, more leisure and a higher standard of living, will depend less on what AI can do than on what we choose to do with it.

Full disclosure: I consulted AI while writing this piece, using it to challenge my thinking, explore arguments, help with the writing process — and, admittedly, curb some of my rambling. I didn't ask it to tell me what to think. And perhaps that distinction is exactly the point.

follow us -Google News