We have figured out how to run everything, absolutely everything, in the 1950s.

The original computer "AI" craze was started by "cybernetic systems" and for good reason. You probably only know of the bastardizations of "cyber-" that don't have anything in common with the original concept.

The original concept goes like this:

1. set a goal
2. perform an action
3. measure how much impact that had, did it get you closer to your goal or not?
4. If you are at your goal, you're done,
5. otherwise adjust your actions, got to 2. (This is "feedback" and the reason that word is now so common. People at the time knew)

The faster you go through the loop, the faster you will figure out what works.

You can measure anything you want, as vague is you want. Happiness, money, productivity. It's the way democracy is designed to work, in which case the feedback is vague and the cycle time is measured in years. It runs your thermostats, in your home, big national power grid power plants. It's how autopilots autopilot.

The idea that "nobody could have predicted..." or "nobody responsible" is a myth. We have the science. We know how it works.

Every failure we still experience is a failure we allow to happen. Because of profit, politics, or whatever.

Didn't catch something "going on for years", maybe someone should check more often. "Crazy single individual causing a tragedy"? No, that's a person at risk, probably with social or mental problems you didn't take care of before, didn't flag, and didn't stop in time.

"Nobody wants to work on our open source project" Really, how is your onboarding? Do people take a look at the docs/culture and run away screaming? Yeah?