How it used to be: a startup made one function more convenient than large corporations. It took about six months. They launched the service, people started using it, it gained an audience and captured a small market from the corporation. It was easier to buy than to squeeze out, because the corporation would also have to spend 5-6 months on development, and by that time the startup would have grown or even sold itself and its audience to competitors. That's why investments were needed: while the corporation was "itching," the startup, using the investor's money, scaled up and captured the largest possible piece of the market. Then it sold to the corporation at a significantly higher price. If the startup didn't take investments, the corporation copied it in six months to a year and squeezed the entire market back.

Then AI came along, and all developers rejoiced: now they could quickly and easily, literally single-handedly, build apps in a week and launch them. But corporations also learned to do the same, maybe a bit slower, but not critically. And here the math for the startup doesn't add up at all. If a corporation sees a wave of demand, it's easier for them to copy the code and roll out an alternative than to go through an acquisition. All legal issues would take longer. There's no longer any reason to buy a startup, especially if the corporation adds some value—integrates the new function into its ecosystem, provides more data, or simply makes the service better.

Example? The app "Where's the gas." The creator made it literally in a few days and immediately got an audience of several million people and tons of media coverage. But while he was thinking about how to monetize and develop it further, everyone copied it. T-Bank, 2GIS, Yandex rolled out alternatives. But they added big data and became more accurate: in the original, users mark whether gas is available at stations, while Yandex simply pulled data from Yandex.Gas Stations and shows exactly where gas is available, and through navigation access, it also shows where the queues are.

This is a reason for startups to think: creating code has become cheap and fast not only for you but for the entire market. Innovations made and sold by someone else cannot be your competitive advantage. Now AI has changed the IT market, and in the coming years, robots will change the real sector. As before, you need to find a unique selling proposition that is difficult or time-consuming to copy, and that is not code.