From my understanding 'Strange things most people don't care about' may fail (and indeed, probably will fail) but they have also been major lottery tickets in the past. When you have enough money to throw around 'buying every lotto ticket' has a record of paying off better as a strategy than taking a 'rational' approach of only investing in those which you have strong fundamentals in. It takes a stupid amount of money to diversify in such fields so it isn't for everyone of course. Tech just has the best record for it so far with highly scaleable businesses.
They just care about the returns. If instead of what we know as 'tech' the big field was, say, genetic engineering or for the sake of absurdity - literal magic. If it somehow scaled better they would throw their money at that.
These people think they have Big Ideas and to them it isn't really different whether they hire creatives to make art, manufacturing nerds to make semiconductors, software nerds to make SaaS, or biochemists to make pharmaceuticals. It's all inscrutable mundane details to them and their magical thinking. After a career observing such things, I've come to see abstraction as something that can cross from virtue into malignancy.
I think the AI dream is driving some of this type towards madness. As someone in another sub-thread observed, It resonates strongly with the way they engage with the world, with delegation of things you don't really understand. But it also promises a kind of pure capitalist scaling where their revenue dreams can be untethered from labor constraints. The current scramble is the attempt to pivot every business dream into one that uses this new set piece.