Add that in the dotcom bubble, internet-flavored growth had never happened before, so it was relatively harder to predict the growth of (say) Google. But today

Add that in the dotcom bubble, internet-flavored growth had never happened before, so it was relatively harder to predict the growth of (say) Google. But today, deep-learning-flavored growth is a 10-year-old concept, and LLMs largely have leverage over existing (versus quite new) business models. There will probably be fewer new monopolies versus the dotcom era; in particular OpenAI has lots of competition.