I think there’s a lot of armchair simplification going on here. Easy to call investors dumb but it’s probably a bit more complex.
AI might not get better than where it is now but if it does, it has the power to be a societally transformative tech which means there is a boatload of money to be made. (Consider early investors in Amazon, Microsoft, Apple and even the much derided Bitcoin.)
Then consider that until incredibly recently, the Turing test was the yardstick for intelligence. We now have to move that goalpost after what was preciously unthinkable happened.
And in the limited time with AI, we’ve seen scientific discoveries, terrifying advancements in war and more.
Heck, even if AI gets better at code (not unreasonable, sets of problems with defined goals/outputs etc, even if it gets parts wrong shrinking a dev team of obscenely well paid engineers to maybe a handful of supervisory roles… Well, like Wu Tang said, Cash Rules Everything Around Me.
Tl;dr: huge possibilities, even if there’s a small chance of an almost infinite payout, that’s a risk well worth taking.