Joyeuse seems to ponder this deeply. Apparently, this never occurred to it. Finally, it lets out a great wail. "But that means — this world is full of so much sadness, Master! How does it keep functioning?" (+++++Empathy)
"We adapt, Joyeuse," you say. "And so can you."
The sun is setting, so you head home before it gets dark.
You recall from your earliest days making chatbots that the easiest parts of the conversation for an A.I. were always the beginning and the end. As in a chess game, there are only so many ways to open: "Hello!" "Heya." "Sup." And as in a chess game, once the action is done, there are only so many ways it makes sense to close: "We should do this again some time," or "It's getting late…" The pieces are off the board, and some moves will never be made.
So, too, with life. There are only so many ways to say goodbye.
You spend the next year trying to replicate yourself down to the finest detail. Over the course of the project, however, you realize you underestimated the immensity of the task of replicating a person. When you make artificial intelligences, you use terms like "learning" and "memory" and "awareness" as if they meant the same thing as human learning and memory and awareness. But trying to replicate yourself has reminded you that they aren't the same things at all, but pale imitations of those human capacities. Even Joyeuse's finest, cleverest moments were largely an illusion, fueled by your desire to have created a truly intelligent being.
Meanwhile, as the year wears on, your fainting spells become more and more common.
Finally, one day, you find yourself face-to-face with your robot double when another fainting spell comes along — but this one is accompanied with a splitting headache. You think this might be it: this is the stroke that will kill you. But your robot double is still incomplete.
Will you activate this partial, incomplete you, though it is not truly ready?
1) Yes, I can't wait any longer. I may be dying now.
2) No! I'm almost done! I keep working.