Advocating for allowing people to die (or otherwise be harmed) because they don't understand any given info space is immoral, no matter how "obvious" you feel that info space is. This is a very common emotional reaction to a wide variety of topics that represents an unrealistic ideology that simply isn't compatible with developed civilization. Other examples include: Internet/data security, industrial/mechanical/electrical safety regulation, self-driving cars, nutritional health, and many more.

This doesn't mean we can't have AI, but it does mean we need to mitigate the problems it presents. And that does not necessarily mean we must attempt to edit/moderate the AI itself (unless we develop effective ways to do so that don't hamstring the AI, but that seems intractable, and may never happen) - we can also address problems like this at other levels (UI level, social level, etc).

As technologists, it's our responsibility to not fall into the common trap of obsessively defending a technology against pragmatic non-technological problems related to it. If you find yourself saying something like "it's your fault for being stupid", that's a very reliable red flag that you might be falling into that trap.