History is full of companies that rush to profit from bubbles. If/when the AI bubble bursts in the next few years, Nvidia and the data centre providers will have made out like bandits from all the people rushing to throw money at them to not be left behind in the new gold rush. It is the AI companies (and those that have invested in them at such lofty valuations) that will be left holding the bag.
This comment of yours has no relation with what I was replying to, which is about calling the "cloud" a b_ll sh_t.
The cloud is not a "bubble", it's not a speculation based on projected gains. Cloud computing services are well-established, constantly growing and replacing on-premise deployments at pace. They are used by large and small companies alike, which depend on them. Amazon, Microsoft and Google are making profits from the "cloud", so it's not a b_llsh_t as stated in the previous comment.
If/when the AI bubble bursts in the next few years
So it's an "if" or a "when"? In the long run, every invention will eventually be replaced by a new one. If everything is a "bubble" because eventually will end, then the word "bubble" means nothing.
Note that the most recent and infamous "bubble" (the dot-com bubble) actually realized and exceeded its expectations. Sure, there were a lot of losers when it bursted, but other survived and recovered, new players emerged and thrived in the "dot-com economy".
In other words, betting that Internet was becoming one of the most important technological revolutions
was the right thing to do. Forecasting who would have gained from it, deciding where to invest money and getting the timeline right is entirely another thing, much more difficult.