Ok, I'm a little scared to put my reputation at risk by saying this, but I think it's important and I think there's a chance I have enough of a platform that speaking up can matter. So here goes. I'm concerned that the meme of "AI risk" is being perverted to destructive ends.
-
-
I find this worrisome on free-speech grounds, on privacy grounds, on all the traditional civil libertarian grounds that seem so "uncool" in 2019.
Show this thread -
I also find it worrisome that there are essentially calls for slowing down and adding more governance to a relatively freewheeling and dynamic industry (software), and making it work more like "mature" industries. Mature industries innovate less.
Show this thread -
And it's *especially* disturbing that I don't see any pushback, not even verbal pushback *by* technologists who are generally chatty on the internet. Not even from my friends. Not even from "heterodox" right-wingers.
Show this thread -
I know: you look kind of weak and not with-it if you say "actually, I think we should be free to do our thing; our thing is mostly good", when you can sense that the currents of power are trending towards illiberalism. But maybe *don't be a weenie*?
Show this thread -
Do you *really* want to go from being a creative intellectual to making sure you get a lock on those sweet DOD contracts? Do you really think it's *good* for positive-sum creation to be turned into zero-sum rent-seeking? If not, then maybe say so?
Show this thread -
People talk about all sorts of dumb and misguided stuff. There are *so* many open racists on Twitter, for instance. Surely it shouldn't be scarier to say "rent-seeking is bad" than to say the far more heinous stuff that "politically incorrect" culture warriors do?
Show this thread
End of conversation
New conversation -
-
-
I would argue that any sufficiently large assemblage of computing power should have no backdoor, on the grounds of AI safety
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I think there will always be people yelling "Give me a grant to study risk X because it's really bad" and these people are unfortunately obscuring the fact that there really is a new, novel type of risk with AI that's far more concerning (it's the longer-term "strong AI" risk)
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Do you seriously think that there are no backwoods already... Let's say USA sells latest tech bombers to Saudi, it will be dumb if there is no backdoor to deactivate the fleet remotely
-
*backdoors. Backwoods becoming scarce
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.