The issue is: 1) Overhype can cause a “winter” where the word itself will become toxic once the bubble pops, leading to significant underinvestment for the per

The issue is: 1) Overhype can cause a “winter” where the word itself will become toxic once the bubble pops, leading to significant underinvestment for the period following. This actually has already happed twice with AI, and is part of why we use “Machine Learning” for concepts that used to be called AI… because AI was a toxic word that investors ran from. 2) A bubble sucks all the oxygen out of the room for all other technological endeavours. It’s strictly a bad thing as it can crush the entire technology sector (or potentially even the economy). 3) Bubbles might cause a return to form, but the internet was more like the railroads. Once built the infrastructure was largely existing and it being sold “for cheap” lead to a resurgence. AI has fewer of these core bits of infrastructure that will become cheap in a bust.