"I can’t explain the reason why AI poses an existential risk in anything like detail required"
And I don`t believe anybody else can either..
I think that`s all that needs to be said on the subject
Top AI inventor Geoffrey Hinton reluctantly concluded that AI will probably humanity fairly soon
by slimboyfat 78 Replies latest social current
-
-
joey jojo
I didn't watch the video but AI is being introduced into a greater range of industries as it gets better.
Manipulation of AI for criminal activity is a risk with the technology as it is right now.
-
cofty
Can't we just pull out the plug?
-
TonusOH
nicolaou: Imagine a future where anyone with a bad motive can easily utilise AI to produce the narrative and results they want.
I think we're already there. I understand that we have ways to determine if a picture or video is fake, but there is an old saying that "a lie can circle the world while the truth is getting its shoes on." Even amateur deep-fakes are pretty convincing, especially if you do not know that they're fake and have your guard down. A talented enough individual (to say nothing of a well-funded group) should already be able to create content that would easily fool most of us.
We are entering a period where trust will diminish and misinformation will really begin to take off. All you need is one incident where people were fooled and the consequences were substantial and bad, and trust in all content will disappear. What will we do when our ability to confirm anything via audio, video, and images is compromised? When governments aren't trying to solve the problem, but actively taking advantage of it?
-
LV101
Disillusioned JW - you're certainly entitled -- the jab deaths are for real.
I don't believe anything until I see it and hope it's all conspiracy nonsense. I prefer remaining vertical and enjoying my life.
-
slimboyfat
Yes Disillusioned JW I accidentally pressed the button while I was trying to change the post title and a word is missing, but the ambiguity is growing on me I kind of like it now.
Yes SydBarrett the social consequences could be severe too, if the existential threat doesn’t materialise there might still be significant challenges.
smiddy the guy in the video is a widely recognised expert. He does explain it well.
Cofty yes that is a key point. These guys think we can’t pull the plug on AI once the AI is more intelligent than us because whatever we think to do to shut it down it will already have anticipated it. It’s more complicated than that but that’s my short version. (Edit: that’s one of the best jokes in the video - a member of the audience asks if we can ask AI itself to solve the problem of aligning our goals. Hinton says that’s possible but he isn’t sure if he would trust AI to give us he right answer. That got a good laugh.)
further edit: that’s another point some of these people make. We talk about “AI” as if it’s one thing, when really there are going to be many different AIs, and if they are correct about the risk, it only takes one of them to be misaligned with human goals, combined with super intelligence, is a deadly recipe. Even if we can align and/or shut down some AIs, we would need to succeed every single time in order to evade existential threat. It seems a pretty bleak prospect, if what there guys are saying is true.
-
careful
SBF, what verb did you intend to put in the last clause?
-
slimboyfat
My memory is getting worse these days, I think I had “kill” and wanted to change it to “eliminate”, but I’m not sure. In the end I somehow pressed the button and missed out the word altogether.
edit: but as it stands it could accommodate hopeful words such as “enhance”, “delight”, “settle down and marry and live happily with”, who knows? 🙂
-
LV101
The title of the Hinton youtube, "Humanity is just a passing phase to evolutionary intelligence" is up and coming.
-
notsurewheretogo
Other side of the coin from Neil DeGrasse Tyson:
https://www.youtube.com/watch?v=ukhmq5on-IA&
https://www.youtube.com/watch?v=1WAtYr0D57s
TL:DW Nothing to worry about.