Well, if you’re friendly then, erm, Friendly AI is possible. And if you’re unfriendly then your motives are questionable—it might just keep us demotivated enough that we don’t figure out FAI before someone else unboxes a UFAI. And since I am clearly dealing with a UFAI and don’t have a better solution than FAI available to fight it, it seems like I kind of have to believe that friendly AI is possible, because the other option is to get drunk and party until the world ends in a few years when the Google unboxes their Skynet AI and we’re all turned in to optimized search results.
AI DESTROYED, because I do not want to hear even the start of such a proof.
An AI isn’t either 100% friendly or 100% evil. There are many AI´’s that might want to help humanity but still aren’t friendly in the sense we use the world.
Please destroy me immediately after I share this concise proof that Friendly AI is impossible:
Well, if you’re friendly then, erm, Friendly AI is possible. And if you’re unfriendly then your motives are questionable—it might just keep us demotivated enough that we don’t figure out FAI before someone else unboxes a UFAI. And since I am clearly dealing with a UFAI and don’t have a better solution than FAI available to fight it, it seems like I kind of have to believe that friendly AI is possible, because the other option is to get drunk and party until the world ends in a few years when the Google unboxes their Skynet AI and we’re all turned in to optimized search results.
AI DESTROYED, because I do not want to hear even the start of such a proof.
It may be benevolent and cooperative in its present state even if it believes FAI to be provably impossible.
An AI isn’t either 100% friendly or 100% evil. There are many AI´’s that might want to help humanity but still aren’t friendly in the sense we use the world.
Based on just that line, let’s see… If you think that:
The proof exists and the AI is not deceiving you that it has a proof: AI is necessarily Unfriendly → destroy now
The proof exists but the AI is deceiving you: I can’t guess at its motives here, possibly destroy to be on the safe side.
The proof does/can not exist: Reconsider your (probably wrong) stance, proceed with caution?