0096 - Hardcoded programming - 2019.08.05



Zoa's "ah, but if you are a bad person, you wouldn't want to do the right thing" logic relies on a certain Bizarro-like expectation of everything being topsy-turvy opposite. Of course, the very idea of an "opposite", when applied to a physical reality, often makes no sense. If something is red, then its opposite should be green... but it's also visible, which means its opposite should be invisible, and how can an invisible thing be green?

Of course, bad people don't work that way. Also, of course, evolutionary psychology doesn't work that way - people don't get depressed because it's a survival trait.

Anyway, as a piece of world-building, I will point out that truth-contexts are absolutely an element of the human consciousness and need to be an element of AI consciousness. That's the reason why you can say "this sentence is false" to Zoa and it won't cause it to malfunction and shut down: Zoa is more than capable of handling apparent paradoxes by segmenting out different contexts for every aspect of a statement.

It's actually quite similar to the way that Zoa was able to act on Lee's behalf earlier - it enrolled Lee in Rosenthal because of its own self-interest, but was able to establish a context in which its only priority was Lee's well-being and show that it could derive the same course of action.

It should be noted, as I said, that separating out reality-contexts is an important element of human consciousness. You should be very careful around someone who appears to be unable to do so. You should be more careful around someone who does it a little too well.


0096 - 2167/07/06/12:29 - Lee Caldavera's apartment, living room
LC: I guess it makes sense that most humans don't think of themselves as bad. I mean, if thinking of yourself as bad causes you to self-harm, and AIs to stop you haven't been invented yet... it takes you out of the gene pool. The trait of thinking that you're bad gets bred out of the population.
Zoa: You think people who consider themselves morally bad would necessarily want to kill themselves?
LC: If your existence is bad, wouldn't eliminating yourself be the right thing to do?
Zoa: Ah, but if you are a bad person, you wouldn't want to do the right thing, now would you?
LC: I suppose that makes sense.
LC: Wait a second, if you have it hardcoded into your programming that there's no such thing as a bad person, how can you form thoughts and statements about what bad people would or would not do?
Zoa: Oh, every truth statement has a reality-context in which it is true. It's the same reason I can know that unicorns aren't real, but if you asked me how many horns a unicorn has, I'd still know the answer is "one".
LC: Unicorns are real. I remember I saw one at the zoo when I was nine.
Zoa: Okay, unicorns that evolved and exist in nature aren't real.
Zoa: Also, I don't think that was a zoo.