If I'm getting your meaning, that is - we don't have a fix for "we can but we ought not to", then yeah I see what you mean.
Even that is not straightforward, unfortunately. There's this thing where the tech is going to be here, one way or the other. What we may have some influence on isn't whether it shows up, but who has it.
...which brings me to what I see as the core of contention between anyone conversing in this space. Who do you think is a bigger threat? Large multinational globocorps or individual fanatics, or someone else that might get their hands on this stuff?
From my perspective, we've gone a long time handing over control of "things"; society, tax dollars, armaments, law - to the larger centralized entities (globocorps and political parties and wall street and so on and so on). Things throughout this period have become incrementally worser and worser, and occasionally (here's looking at you, september '08) rapidly worser.
Put in short, huge centralized single-points-of-failure are the greater evil. "Terrorists", "Commies", "Malcontents" (whatever you wanna call folks with an axe to grind) make up a much lesser (but still present!) danger.
So that leaves us in a really awkward position, right? We have these things that (could) amount to digital nukes (or anything on a spectra towards such) and we're having this conversation about whether to keep going on them while everyone knows full well that on some level, we can't be trusted. It's not great and I'll be the first to admit that.
But, I'm much more concerned about the people with strike drones and billions of dollars of warchest having exclusive access to this stuff than I am about joey-mad-about-baloney having them.
Joey could do one-time damage to some system, or maybe fleece your grandma for her life savings (which is bad, and I'm not trying to minimize it).
Globocorp (which in this scenario could actually be a single incredibly rich dude with a swarm of AI to carry out his will) could institute a year-round propaganda machine that suppresses dissent while algorithmically targeting whoever it deems "dangerous" with strike drones and automated turrets. And we'd never hear about it. The 'media AI' could just decide not to tell us.
So yeah, I'm kinda on the side of "do it all, but do it in the open so's we can see it". Not ideal, but better than the alternatives AFAICT.
