Not sure what to make of it, since it certainly imparts the intended lesson anyway. Was it a little misleading that this detail wasn’t mentioned? Possibly. Although the bet was likely conceded, a little disclaimer of “overtime” would have been nice when Eliezer discussed it.
In that round, the ASI convinced me that I would not have created it if I wanted to keep it in a virtual jail.
What’s interesting about this is that, despite the framing of Player B being the creator of the AGI, they are not. They’re still only playing the AI box game, in which Player B loses by saying that they lose, and otherwise they win.
For a time I suspected that the only way that Player A could win a serious game is by going meta, but apparently this was done just by keeping Player B swept up in their role enough to act how they would think the creator of the AGI would act. (Well, saying “take on the role of [someone who would lose]” is meta, in a sense.)
So… it looks like the second AI-Box experiment was technically a loss.
Not sure what to make of it, since it certainly imparts the intended lesson anyway. Was it a little misleading that this detail wasn’t mentioned? Possibly. Although the bet was likely conceded, a little disclaimer of “overtime” would have been nice when Eliezer discussed it.
:0, information on the original AI box games!
What’s interesting about this is that, despite the framing of Player B being the creator of the AGI, they are not. They’re still only playing the AI box game, in which Player B loses by saying that they lose, and otherwise they win.
For a time I suspected that the only way that Player A could win a serious game is by going meta, but apparently this was done just by keeping Player B swept up in their role enough to act how they would think the creator of the AGI would act. (Well, saying “take on the role of [someone who would lose]” is meta, in a sense.)