Nope. The best thing it could do would be retrieve my dead friends and family. But if we’re talking about whether I should sign up for cryonics, I’m assuming that’s the only way somebody gets to be not dead after having died a while ago. If we have an AI that’s so brilliant that it can reconstruct people accurately just by looking at the causal history of the universe and extrapolating backwards, I’m safe whether I sign up or not! And if we have one that can’t, I think I’m only safe if I am signed up with at least one loved one.
The best thing it could do would be retrieve my dead friends and family.
Out of curiosity—how accurate would the retrieval need to be? For instance, suppose the FAI accessed your memories and reconstructed your friends based on the information found there, extrapolating the bits you didn’t know. Obviously they wouldn’t be the same people, since the FAI had to make up a lot of stuff neither you nor it knew. But since the main model was a fit to your memories, they’d still seem just like your friends to you. Would you find that acceptable?
My initial reaction is that I would really hate this. It’s one of the things that makes me really uneasy about extreme “neural archaeology”-style cryonics: I want an actual reconstruction, not just a plausible one.
You can think of no scenarios between those two that would entice you to sign up?
Nope. You’re welcome to try, though, if you value my life and don’t want to try the “befriend me while signed up or on track to become so” route via which several wonderful people are helping.
Nope. The best thing it could do would be retrieve my dead friends and family. But if we’re talking about whether I should sign up for cryonics, I’m assuming that’s the only way somebody gets to be not dead after having died a while ago. If we have an AI that’s so brilliant that it can reconstruct people accurately just by looking at the causal history of the universe and extrapolating backwards, I’m safe whether I sign up or not! And if we have one that can’t, I think I’m only safe if I am signed up with at least one loved one.
Out of curiosity—how accurate would the retrieval need to be? For instance, suppose the FAI accessed your memories and reconstructed your friends based on the information found there, extrapolating the bits you didn’t know. Obviously they wouldn’t be the same people, since the FAI had to make up a lot of stuff neither you nor it knew. But since the main model was a fit to your memories, they’d still seem just like your friends to you. Would you find that acceptable?
No. That would not be okay with me, assuming I knew this about the process.
My initial reaction is that I would really hate this. It’s one of the things that makes me really uneasy about extreme “neural archaeology”-style cryonics: I want an actual reconstruction, not just a plausible one.
You can think of no scenarios between those two that would entice you to sign up? Your arguments seem really specious to me.
Nope. You’re welcome to try, though, if you value my life and don’t want to try the “befriend me while signed up or on track to become so” route via which several wonderful people are helping.