I recently saw a heated debate about how artificial intelligence will destroy the economy and jobs at an incredible speed. Honestly, I strongly disagree with them.

People always forget the power of institutional inertia. Take real estate agents — people say "brokers are finished" 20 years ago, but they still take a 5-6% commission! I bought an apartment a while ago, and the process forced me to use an agent even if I didn’t want to. The guy made about $50,000 from a deal that took him at most 10 hours. This is a clear example of how institutions, regulations, and habits protect old jobs much more strongly than we expect.

Regarding software, people are afraid that AI will make all programs identical and unprofitable. But the real truth that everyone ignores: most current software is actually very bad. Salesforce and Monday, which I’ve spent hundreds of thousands on — they’re full of bugs and unnecessary complexities. Even large apps don’t have proper consistency between mobile and desktop. The only reason Stripe and Linear stand out is because they didn’t make their use as burdensome as competitors.

Here, an important point emerges: the demand for engineers and developers is almost infinite. Every program has huge improvement opportunities before reaching saturation. Even with all technological progress, the software industry still faces a constant shortage. I’m a developer, and my productivity equals that of hundreds of people from 50 years ago, but results still need significant improvement. This is what they call the Gibbons paradox — increasing efficiency boosts total demand, it doesn’t decrease it.

The inertia in human society is more complex than we imagine. I founded a company whose entire idea was to transform insurance companies from manual services to software, and I learned that any change takes much longer than expected. That doesn’t mean the world won’t change — it means change will be slower and more gradual.

Regarding jobs, it’s true that some will disappear, like autonomous drivers. But here’s the key part: the U.S. has nearly unlimited needs for manufacturing and rebuilding. We’ve lost the ability to produce basics: batteries, engines, electronic chips — we import everything. Even fertilizers! China produces 90% of the world’s ammonia. If supply is cut off, we face a real problem.

The U.S. government has started focusing on bringing manufacturing and infrastructure back. This isn’t just political talk — it’s a strategic necessity. When AI begins to impact jobs, the easiest political approach will be to fund massive construction and manufacturing projects to absorb the workforce. Bridges, roads, facilities don’t need a “unique point” — they are subject to friction and physical reality.

A Salesforce engineer earning $180,000 a year might find new work in water desalination plants or infrastructure projects. People are discovering that tangible physical achievements give a better feeling than just spinning in a digital world.

In summary: I am very optimistic about AI, but the transformation will be much slower than pessimistic predictions suggest. Governments have tools for intervention and stimulation, and the physical economy has huge needs. If we stay attentive and respond quickly during this gradual revolution, we will be safe. Time gives us the opportunity to plan properly.