Perhaps just as usefully, calling it idiocy is not how you are going to change things. What could change things? - "There should be a law" (1) really never sol

Perhaps just as usefully, calling it idiocy is not how you are going to change things. What could change things? - "There should be a law" (1) really never solves anything, (2) pushes stuff in the next jurisdiction. See attempts at killing generally available cryptography and Cypherpunks. Genuine success by the techies there. Law / policy is insufficient by itself. - Social media does get people thinking. A little. "The press should..." yeah maybe but the press doesn't so we do what we can. - Scifi authors write scifi. It's what they do. And essays sell sci-fi. And their essays are broadly commented on: THEY have a voice. Not going to change. - Outfunding the good while dis-incentivizing the bad might work: This is a field with massive ongoing progress that people are legitimately proud of. If these (many but finite number of) people can work "in the right direction", then they work less "in the wrong direction". Government funding might be the rare thing that can outspend advertising funding. Still, many people will chase AI work directly out of the large military budgets, in many countries with varied political systems. Meaning lots of people working in these directions no matter what. Meaning lots of spillovers from that no matter what else. - Having the better AIs beat up the worse AIs? Hmmm, I don't want to be caught in the middle of this. Actually this is a field for sci-fi authors to plow some effort into. And for AI game theorists? What does this look like? - Having the better AI fields be massively more profitable than the "bad" - so that the talented hopeful head in these directions more readily? Currently advertising AI and medical AI might be the highest profit potential, with military close in there? Self-driving AI probably pays well already now. But most of these run "general AI" labs in the name of "progress is progress". - More money into "AI ethics"? So far the results are underwhelming but not a reason to stop. Other practical suggestions?