Welcome Back to XcessAI

A few weeks ago, we explored how Artificial Intelligence is rapidly becoming mandatory inside organizations.

The argument was straightforward: once every company adopts AI, the competitive advantage disappears. AI stops being a differentiator and becomes infrastructure.

But that raises a deeper question.

If every company uses AI to become more productive, what happens to the people whose work is displaced?

And perhaps more importantly:

Who buys the output?

A recently published paper titled The AI Layoff Trap tackles this question through an economic lens. Its conclusion is both simple and provocative: competitive markets may encourage companies to automate more work than is ultimately healthy for the economy as a whole.

Whether the authors are right remains to be seen. History has repeatedly shown that technological progress creates new industries, new jobs, and new forms of demand. Yet the paper raises an important question that deserves serious attention because it sits at the intersection of artificial intelligence, economics, and business strategy.

The question is not whether AI will increase productivity.

The question is what happens if productivity grows faster than purchasing power.

Because if AI eventually performs a significant portion of the work, someone still needs to buy the products and services being produced.

For decades, business leaders have pursued the same objective.

Higher productivity. More output. Lower costs. Better margins.

Artificial intelligence appears to be the most powerful productivity tool ever created.

Unlike previous technologies, AI can automate not only physical work but also many cognitive tasks. It can write reports, generate code, analyze contracts, summarize research, create marketing campaigns, answer customer inquiries, and increasingly perform activities that were once considered uniquely human.

From an individual company's perspective, the incentives are obvious.

If AI allows a firm to operate with fewer employees while maintaining or increasing output, shareholders benefit.

Profits rise. Competitiveness improves. The decision appears rational.

The problem emerges when every company reaches the same conclusion simultaneously.

The paper describes what economists call an externality.

Imagine a company replaces 500 employees with AI systems. The company saves money. Productivity improves. Shareholders benefit.

On its own, this appears entirely rational.

Now imagine thousands of companies doing the same thing.

The savings remain private. But the consequences become collective. The workers who lose jobs are not only employees. They are also consumers. They buy houses, cars, clothes, food, streaming subscriptions, vacations, etc.

If enough people lose income faster than the economy can create new opportunities, consumer demand begins to weaken. This creates a feedback loop.

Companies automate to reduce costs. Workers lose income. Demand falls. Companies face slower growth. They respond by reducing costs further. Which often means more automation. The cycle reinforces itself.

No individual company is behaving irrationally. In fact, every company is doing exactly what markets encourage it to do. That is precisely what makes the problem interesting.

At this point, many readers may be thinking:

"We've heard this before."

And that is a fair response.

The Industrial Revolution displaced countless agricultural workers. Mechanization transformed manufacturing. Computers automated clerical work. The internet disrupted entire industries. Yet unemployment did not permanently explode. New industries emerged. New jobs appeared. Living standards improved.

History suggests that human societies are remarkably effective at adapting to technological change. That remains the strongest counterargument to the paper's conclusion.

Perhaps AI will create entirely new professions. Perhaps productivity gains will generate new forms of demand. Perhaps entirely new industries will emerge that we cannot yet imagine.

History certainly provides reasons for optimism.

The authors argue that AI deserves special attention because it is not merely automating muscles. It is increasingly automating cognitive tasks.

Previous technological revolutions primarily enhanced human capabilities. AI increasingly replicates them. That distinction may prove important.

A tractor could replace a farmer's physical labor. It could not become an accountant. A factory robot could assemble products. It could not draft a legal contract.

AI is beginning to cross boundaries that previous technologies largely respected.

If that trend continues, the adjustment process could become more challenging than past technological transitions. Not necessarily because jobs disappear. But because new jobs may emerge more slowly than old ones vanish.

Regardless of whether the paper's conclusions prove correct, executives should monitor a few key indicators.

Labor's share of economic output. Consumer spending trends. Wage growth relative to productivity growth. Corporate profit concentration. And perhaps most importantly, the relationship between productivity gains and demand growth.

If productivity accelerates while purchasing power stagnates, the dynamics described in the paper become more relevant. If demand continues expanding alongside productivity, history may once again prove the pessimists wrong.

The answer will not be found in academic debates. It will be found in the data.

Artificial intelligence promises extraordinary productivity gains. Businesses should embrace those opportunities.

But productivity is only one side of an economy. Demand is the other.

For more than two centuries, technological progress increased both. The question raised by The AI Layoff Trap is whether AI might eventually weaken that relationship.

Nobody knows the answer. History suggests adaptation. The model suggests caution.

Reality will likely fall somewhere between the two. But the question itself is worth asking.

Because the greatest challenge of the AI age may not be producing goods and services more efficiently. It may be ensuring there are still enough customers to buy them.

Until next time,
Stay adaptive. Stay strategic.
And keep exploring the frontier of AI.

Fabio Lopes
XcessAI

💡Next week: I’m breaking down one of the most misunderstood AI shifts happening right now. Stay tuned. Subscribe above.

Read our previous episodes online!