kane_magus (![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
kane_magus) wrote2023-02-25 11:42 am
kane_magus) wrote2023-02-25 11:42 am
Entry tags:
"Don’t worry about AI breaking out of its box—worry about us breaking in"
"Opinion: The worst human impulses will find plenty of uses for generative AI."
All the inane stuff about the possibility of terrorists manipulating ChatGPT or whatever into telling them how to build bombs (because if a terrorist really wants to know how to build a bomb, there are far easier and better ways to find that out than by the computer equivalent of cajoling a toddler into telling them how) and the like aside, I think the article touches on what is probably the bigger "threat" of these AIs.
Probably the biggest "threat" of current and future AI is less along the lines of the "Skynet takes over and nukes humanity into extinction" thing or the "HAL 9000 locks you outside of the spaceship" thing, and more along the lines of the "humanity drives itself to extinction by spending every waking moment fucking computer-made holograms in the holodeck" thing.
This kind of thing was somewhat explored in Her, a movie from almost a decade ago that I haven't actually seen myself, though I read the plot synopsis on Wikipedia.
But really, yeah, the biggest actual "threat" of AI is, of course, how humanity uses/misuses/abuses it, same as with any and every other piece of technology ever made in the history of humanity. You could use that large stick you picked up to fend off the saber-toothed tiger threatening your tribe, or you could use it in a fire to cook food... or you could use it to brain your caveman buddy and steal his food/mate/cave/whatever.
(And if you replace human customer service bots with AI customer service bots, it would be a lateral move at best, for the most part. If anything, we already have computers trying to "help" with customer service, given how so much of the "wait times" for customer service is, in large part, due to trying to wrangle the computer voice/chatbot talking to you on the phone/website into connecting you with an actual human, instead of continuing to waste your time. If humans were replaced with AI in customer service, that would absolutely not make me more confident or willing to utilize customer service, especially given how current chat AIs are so often confidently/belligerently wrong about things they claim. But then, it's certainly not like humans themselves lack the capacity for being confidently/belligerently wrong about things, either, including those who work in customer service.)
All the inane stuff about the possibility of terrorists manipulating ChatGPT or whatever into telling them how to build bombs (because if a terrorist really wants to know how to build a bomb, there are far easier and better ways to find that out than by the computer equivalent of cajoling a toddler into telling them how) and the like aside, I think the article touches on what is probably the bigger "threat" of these AIs.
Probably the biggest "threat" of current and future AI is less along the lines of the "Skynet takes over and nukes humanity into extinction" thing or the "HAL 9000 locks you outside of the spaceship" thing, and more along the lines of the "humanity drives itself to extinction by spending every waking moment fucking computer-made holograms in the holodeck" thing.
This kind of thing was somewhat explored in Her, a movie from almost a decade ago that I haven't actually seen myself, though I read the plot synopsis on Wikipedia.
But really, yeah, the biggest actual "threat" of AI is, of course, how humanity uses/misuses/abuses it, same as with any and every other piece of technology ever made in the history of humanity. You could use that large stick you picked up to fend off the saber-toothed tiger threatening your tribe, or you could use it in a fire to cook food... or you could use it to brain your caveman buddy and steal his food/mate/cave/whatever.
(And if you replace human customer service bots with AI customer service bots, it would be a lateral move at best, for the most part. If anything, we already have computers trying to "help" with customer service, given how so much of the "wait times" for customer service is, in large part, due to trying to wrangle the computer voice/chatbot talking to you on the phone/website into connecting you with an actual human, instead of continuing to waste your time. If humans were replaced with AI in customer service, that would absolutely not make me more confident or willing to utilize customer service, especially given how current chat AIs are so often confidently/belligerently wrong about things they claim. But then, it's certainly not like humans themselves lack the capacity for being confidently/belligerently wrong about things, either, including those who work in customer service.)