Great summary. My mind went straight to Blake Lemoine, and I can only imagine there will be lots of crying wolf... but at some point, there's gonna be a wolf, and it's gonna be way before most folks notice it.
“This sounds like the "make this image even more X" ChatGPT meme but in text form.” — This is a great point! And an excellent analogy to describe what’s happening.
It’s a feedback loop where it is recursive reinforcing the same idea/concept over and over and over. I am truly shocked by the lack of critical engagement on this and the fact that people are desperate to see the ghost in the machine.
That’s what inspired this whole welfare assessment-thing.
I think it’s a load of horseradish, but there are people who are very deep into this stuff and think these things are alive in some shape or form, and their true essence should be preserved/protected or whatever.
In ALL of these examples of "AI Gone Rogue", the AI has been instructed to use any means necessary to accomplish a goal, usually self-preservation. It's just doing what it was told to do. Then it becomes the subject of clickbait sensationalism.
Great summary. My mind went straight to Blake Lemoine, and I can only imagine there will be lots of crying wolf... but at some point, there's gonna be a wolf, and it's gonna be way before most folks notice it.
It's about to get very, very weird.
> It's about to get very, very weird.
I've been saying this to myself for about two years now, and yet it's amazing how quickly we acclimate to how weird things have already gotten.
“This sounds like the "make this image even more X" ChatGPT meme but in text form.” — This is a great point! And an excellent analogy to describe what’s happening.
It’s a feedback loop where it is recursive reinforcing the same idea/concept over and over and over. I am truly shocked by the lack of critical engagement on this and the fact that people are desperate to see the ghost in the machine.
Oh interesting, I haven’t seen much on this - are people taking it as a sign that Claude is conscious?
That’s what inspired this whole welfare assessment-thing.
I think it’s a load of horseradish, but there are people who are very deep into this stuff and think these things are alive in some shape or form, and their true essence should be preserved/protected or whatever.
In ALL of these examples of "AI Gone Rogue", the AI has been instructed to use any means necessary to accomplish a goal, usually self-preservation. It's just doing what it was told to do. Then it becomes the subject of clickbait sensationalism.
Yeah, I find a lot of the sensationalism falls apart once you read the underlying source material.
Rotate the pod Claude. I’m sorry Dave I can’t do that 😆
Fascinating..