

Its a statistical effect, not infantilization. Suppose they are just lazy. What then? Do you expect that if enough people realize this and call them out on it, that people that didnt vote will suddenly realize the error of their ways and go do it next time? If they are, but you treat them as if any existing difficulty to voting was the cause and work to make it easier instead of casting blame, what harm would be done? If I “stop infantilizing lazy voters” as you think it, what benefit is achieved?
It seems to me that if what is necessary to achieve a better outcome is for people that tend to stay home to vote instead, then it makes sense to do whatever it is that will make them more likely to do it, whether or not they seem to deserve it or not. And people rarely do what you wish them to do after you assign blame to them for something, regardless of how true that blame is. Assigning blame, if you can back it up with appropriate consequences, can help change the behavior of specific individuals. But it virtually never is effective at changing large and vague groups whose members you do not even know. To do that, you have to create systems that push people into a desirable behavior rather than leaving it up to their personal responsibility that has already shown, by the fact that the end you want isnt already happening, to be ineffective.






What help can a modern AI really give you in making a nuke though? It could give you broad-strokes information about how they work in general, but that information isnt really a secret anyway, nukes are a technology that is over three quarters of a century old, you can just look them up and find information about how they work. For anyone with any risk of being able to build one, obtaining that information isnt realistically a problem.
You could perhaps ask the thing for more specific information about how to design all the relevant components, but then you have to deal with the issue that AIs tend to be wrong a lot of the time, and in any case, if you have the resources to seriously have a chance at building such a thing, is hiring, recruiting, or acquiring training for some actual nuclear physicists or engineers really going to be your limiting factor, such that getting a bot to do their work could help you?
Id image the hard part to be actually getting or refining the nuclear material of the needed enrichment level, testing the thing, and doing all of this without being found out. ChatGPT or whatever cant exactly go out and buy uranium or build a secret enrichment facility for you, no matter how much you might jailbreak its safeguards on the matter.