I mean, you would think someone very rich who invests in companies would be somewhat smart. But, I'm convinced, a lot of them would do much better if they made no decisions at all and left everything up to entropy.
Imagine having so much money that you've already invested in stocks, bonds, real estate, you own everything, you got bored going to Vegas, and have nothing else to do with your money. So you start to toss some of it into a fund with others. Not a hedge fund mind you. You've done that already too. No, you stick it into a fund to be managed by lunatics. Because have enough money and you want to see a moon shot play out for fun.
That's the kinda stupid "F you" money we're talking about. It comes from people who don't care. They literally don't care. They just want to say they invested in XYZ, because no one cares about where their money came from or what their normal investments are.
This is the kinda very rich we're talking about and they aren't all there in the head sometimes.
I wish I had that much money lol.
Some are but many aren't. The reason so many investors are pushing on AI (or other buzz-trend du jour) is that history shows 'disruptive new technologies' tend to correlate with some startups quickly growing and becoming 'unicorn' successful. The problem is that this historical correlation appears more obvious after the fact than when it's still emerging. And of course there are lots of caveats and other requirements which a given startup and implementation may or may not meet.
Since professional VCs tend to take a portfolio approach to investing their funds on behalf of their limited partner stakeholders, they generally divide their funds into several broad 'investment themes' based on what they perceive as potentially disruptive new technologies or markets. Like roughly 30% here and 20% there with 15 or 20% left for things which don't fit a theme. This approach is supposed to ensure they don't miss out on having a few bets in each big category. In 2005 social media platforms were an important theme. In 2015 it was SaaS.