When Uber launched, it didn’t just provide rides, it subsidized them. By pouring venture capital into cheap fares, it hooked both drivers and riders, building a massive user base before competitors could catch up. The economics were never sustainable at face value, but the strategy worked: the network scaled, user habits hardened, and switching costs rose.Today, AI is running a strikingly similar play. The true cost of training, hosting, and running large language models is enormous. Yet to e...