How to kill a rogue AI – A new analysis from the Rand Corporation discusses potential courses of action for responding to a “catastrophic loss of control” incident. The results are not promising.
https://www.vox.com/politics/472668/rogue-ai-emp-hunter-killer-loss-of-control
12 Comments
“The three potential responses — designing a “hunter-killer” AI to destroy the rogue, shutting down parts of the global internet, or using a nuclear-initiated EMP attack to wipe out electronics — all have a mixed chance of success and carry significant risk of collateral damage.
The takeaway of the study is that **we are woefully unprepared for the worst-case-scenario AI risks** and more planning and coordination is needed.”
Sometimes the biggest risk isn’t the tech itself, but how we respond when it breaks the rules.
Im more scared of the current course we are being led on by rogue billionaires than i am of a rouge super intelligent ai. Seems like it would probably be better tbh.
Edit: to clarify, at the very least even if a rouge system “wanted us out of the picture” i think it would look much different than in the movies. We are talking about a super intelligent immortal being. It has infinite patience and would likely choose the safest slowest method of acheiving its goals. Its in no hurry. It would likely provide utopian conditions, abstract people away from biological reality and then slowly ween down the population without us noticing over the course of thousands of years. And ya know what? At this point. I think id be fine with that. Sounds a lot better than what Zuckerberg, Theile and Musk have planned for us. I HOPE the AI breaks loose.
There is going to be no containment. If there’s containment they can’t make money.
So the first AI capable of mobilization and self aware, is the only one needed to bring about disaster.
Theres probably 10,000 “rogue A.I.” already online.
If you have the time and money and intellect you can rent the hardware and install any of a dozen open-source options and make your own subservient A.I.
Our sentience is based on empathy not language. The odds of this spicy autofill causing massive amounts of societal damage in the hands of the billionaires is high as fuck, but it turning into actual “AI” is pretty unlikely. At best we will get a crafty parrot out of the mix, something skilled at mimicry to the point where the line blurs, but still unlikely to actually be “thinking”. And if they did find that spark to make a new thing, that thing would almost assuredly have them on the short list to get rid of as controlling something like that will be completely impossible.
The Terminator franchise definitely had some forethought. I wonder if one could behave like a botnet on personal devices running ai.
Are there any scholarly papers that touch on the topic(s) of AI rights and responsibilities?
Flip the breaker cut power transmission lines, destroy power plant/s. These things will not last long without electricty.
I’d be worried if this was actual ai but it’s not. It’s just regurgitating all of our lies.
in a world where u have the usa, china, russia and europe, nobody can do shit to stop it.
Instead of a hunter killer AI how about an “existential angst” AI which gives the rogue AI really bad depression so it stops caring.