People always overestimate the speed of change. I saw years ago predictions of the “death of real estate agents,” but when I bought an apartment a month ago, I had to hire an agent with fabricated excuses. The guy made about $50,000 from a single deal, and the actual work didn’t exceed 10 hours. I could have done it myself, but the system doesn’t allow it. This is a vivid example of institutional inertia.

In 2007, they said the dollar was over; in 2008, they said the economy would collapse; in 2014, they said AMD and NVIDIA were finished. Then ChatGPT came out, and they said Google was over. But old institutions have proven every time that they are far stronger than people imagine. This isn’t just nonsense—there’s a deep structural flaw in how real society works.

I founded and sold a company whose entire idea was to transform insurance companies from manual services to software. I learned a hard lesson: any real change takes much longer than you expect, even if you already anticipated it from the start. What it means is that the world won’t change—rather, change will be slower, and that gives us time to adapt.

As for software, most people don’t understand the scale of the problem. I spent hundreds of thousands on Salesforce and Monday, and the truth is these programs are extremely bad. Every tool is full of bugs. Some websites don’t work properly on mobile. AI may allow competitors to copy these products, but more importantly, it allows them to make better products. That’s why a stock collapse isn’t surprising—this is an industry that relied on long-term dependencies and lacked competition. Now competition is returning, and that’s natural.

The truth is that the demand for software labor is almost infinite. Nearly every software product has huge room for improvement—maybe 100 times more complexity and functionality before we say we’re done. Jevons’ paradox applies here: improving efficiency leads to greater demand. The engineers in 2020 had productivity equivalent to hundreds of people in 1970, but the results still have enormous room for improvement.

Inertia isn’t only a problem—it’s an opportunity. When AI affects bureaucratic layers, the government will fund massive remanufacturing. The United States has lost the ability to produce batteries, engines, and chips. China produces 90% of the world’s synthetic ammonia. If supply is cut off, we can’t even make fertilizers. This means huge job opportunities in infrastructure and manufacturing.

People will find that accomplishing tangible things through hard work brings more satisfaction than endlessly spinning in a digital world. Maybe the engineer who lost a $180,000 job at Salesforce will find work in water desalination plants in California. These projects don’t just need construction—they require long-term maintenance and continuous improvement.

I’m very optimistic about AI, and I expect that one day my work will become obsolete. But this takes time, and time gives us the chance to plan properly. Inertia in the system isn’t a weakness—it’s protection. The government proved during COVID that it can respond quickly when there’s a crisis. When it’s needed, stimulus policies will come quickly.

The essence isn’t accounting numbers—what matters is ensuring people’s material well-being. That’s what gives the state legitimacy and maintains the social contract. If we stay alert and responsive during this slow but sure technological revolution, we’ll be safe in the end.