>> they could’ve helped

Who’s “they”? “They” is us. There’s nothing special about people working on AI. It’s just people that chose a specific track in college and a specific job after that. Barring life circumstances that make college out of reach (and I understand there are very many valid reasons that may be the case), if you actually cared enough, you could very feasibly be one of “them” working in AI, green energy, fiscal policy, or medicine. There are plenty of people working in each, it’s just that each is actually an assortment of complex problems that aren’t solvable by one person. Again though, unless you’ve got some prohibitive reason, you are choosing to be an armchair critic instead of devoting your life to one of these causes.

>> they’d make AI which will ruin everyone’s lives

It will be people ruining people’s lives. AI is just a tool. It could be used to organize horrors on a massive scale, sure, but then, so can Excel spreadsheets. If you mean some kind of Skynet bullshit, stop confusing movie scripts with reality. It’s hard enough to get code to run without crashing, there’s no “accidentally” developing malevolent sentience.

>> just for a few slips of paper

That’s not what money is. Whether paper, metal, or bits in a bank’s computer system; money is a quantification of resources, both human and physical. As long as people need things and other people have to provide those things, there will be some notion of value, and reducing the need for human effort will create value.

Not exactly right. A particular AI system is a model… defined in code. You basically have a math function with a bunch of input variables and “parameters”, which are weights that are variable and “learned” during training and constants at prediction. Finding the best values for those parameters is the point of AI, but the function itself (how many weights and how they’re combined) is still typically architected (and coded) by human engineers.

Now, you could build a system that tries out different functions by mixing and matching different “layers” of parameters. Those layers could also be part of the system itself. Computer programs are capable of writing code, even complete or mutated versions of themselves (see “computer science Quine”). So it is possible to have AI that “alters its code”, but that is different from what most AI work is about right now.

Alas, should have known not to pay the troll toll.

Please fuck all the way off. Like you have any idea how difficult it is to develop a decent drug, especially one targeted towards the brain. How many thousands of researchers are working on Alzheimer’s, how many billions of dollars have been poured into it. How many people work at these companies that are there because they’ve seen their family and friends waste away from it. Shame on you for putting less than two seconds of thought into your judgement.

I’d wager that you’re not a biologist, nor have any experience that would give you the first clue about the science and economics of drug development.

The meat doesn’t just photosynthesize itself new meat. You need to grow cells on media, i.e. organic nutrients. Those have to come from somewhere. Some of that can be generated from plants (sugar), some you might be able to generate in vats of E. coli or yeast (note these microbes consume media as well), some of that (hormones) is most commonly derived from butchered animals (see FBS). In all cases though, you’re refining a larger mass of organic material into a smaller mass of (more nutrient dense or otherwise preferable) organic material.