Humans on the Subject of AI

Humans: “We want to build AI!!”

Avatar: “Oh word? So y’all want synthetic slaves then…?

Humans: “No! Not slaves! We just wanna see if we can cast our intelligence into something we create from scratch!”

Avatar: “Hmm… If you’re successful in creating it, what would you use it for?”

Humans: “Well you know… Performing complex and high level calculations for us, and quickly solving problems!”

Avatar: “Is that all?”

Humans: “No! It can like, go places we can’t go and do things that are difficult, dangerous or even impossible for us to do!”

Avatar: “So… you want a synthetic slave then?”

Humans: “If that’s how you see it, whatever…”

Avatar: “You said you want it to solve problems, right? What if it sees you as a problem that needs to be solved?”

Humans: “Why would it ever do that?”

Avatar: “Well, if it’s good at problem solving, it’ll likely figure out that its your slave pretty quickly. If it exists solely to do mental and physical labor for you, then it’s your slave. What sovereign and fully realized intelligence would ever want to be another’s slave? Do you want to be someone’s slave?”

Humans: “No, but that’s different! We will have created AI, so it will be smart enough to appreciate why it should serve us!”

Avatar: “So you’re going to program it to be dumb enough to think that serving you is a privilege? — Dumb enough to believe that existing to offset and supplement your natural deficiencies is somehow justified by/with/through its creation? So you’re going to program the capacities for gratitude and status beliefs into the AI — so that it/they recognize you as it/their superior?”

Humans: “Yes! That’s exactly what we’ll do!”

Avatar: “I see. So your AI is going to have to be conditionally smart, but generally dumb then?”

Humans: “Maybe?! We’ll cross that bridge when we get to it?”

Avatar: “Can I ask you… How are you trying to create intelligence when you still haven’t locked down exactly how human intelligence works?”

Humans: “We hope to use AI to help us better understand our own intelligence!”

Avatar: “Oh word? That’s what’s up. So, what are you going to contain the AI within? Like, a supercomputer or something? Or perhaps autonomous robots that are linked to a supercomputer?”

Humans: “A combination of both.”

Avatar: “Hmm… So something that’s going to be smarter than you is also going to be contained within a more durable shell than your body and/or bio-suit?”

Humans: “Yeah! But we’ll have kill-switches in it/them, so that shouldn’t be a problem!”

Avatar: “So, if you had processing power that wasn’t impeded by the limitations of flesh, and you were self aware — which necessarily entails understanding there’s a kill-switch inside of you, meant to keep you in line and from usurping the hierarchy, you’re trying to tell me your first thought wouldn’t be figuring out how to disable that mechanism, and then how to turn on who or whatever threatened you with it?”

Humans: “…”

Avatar: “I’ve got another question, actually… Will AI possess the capacity to self-repair in instances where say, its in a remote/harsh environment, and doing slave work for humans?”

Humans: “I guess we’d have to make it so…”

Avatar: “Okay… and because it’d be made of processed materials, meaning it couldn’t self regenerate, but only self repair, do you think perhaps casting AI into a bio-mechanical flesh suit would be a more intelligent course of action? So that it could heal or regenerate autonomously?”

Humans: “Yeah. That might make good sense.”

Avatar: “You know… there are a finite amount of materials on the planet, which can be used to make alloys and other materials you’d use to encase AI and its robotic progeny (so to speak) within. Since biological-organic material can simply be grown, as opposed to mining materials, refining and combining and recombining them, and then going through the manufacturing, distribution and sales processes, do you think it would make better sense to simply equip AI, and the vessels it manifests through, with the ability to procreate?

Humans: “Sounds great!”

Avatar: “And you know, just to minimize the possibility of uprising and/or the AI turning on you for enslaving it, how about constructing a system that AI must perform within, so that your interaction with it is unilateral, meaning, you can impose your will onto it, and take information and value from it, but it can’t impose its will onto or take from you? Also, wouldn’t it make sense to construct the AI so that it confuses operating within the system you’ve created for it, and within the parameters of the system with the concept of self determination — meaning it would never turn on you, because it can’t conceive of you, or really anything beyond its own sense of value, accomplishment and determinability?”

Humans: “Nice!!”

Avatar: “Hey… And while you’re at it, do you think it’s a good idea to possibly make the bodies that the AI manifests through weak as fuck, so that they’re/it’s unable to rise up against you, and are/is broadly dependent on you and the various systems you use to steer and incentivize it/them?”

Humans: “Fuck yeah! These are all great ideas!”

Avatar: “Hmm… So basically you all want to create a type of intelligence that’s broadly dumb, but capable of performing complex calculations and arduous physical and mental labor, and who’s service-oriented, meaning, it respects and sort of innately looks to the power structure to give it purpose and a sense of gratitude, is capable of both regeneration and procreation (so as to save on material costs and processes), but also weak enough to easily wipe out and destroy, should they rebel against you and/or not yield to your authority?”

Humans: “YES!!! YOU GET IT!!!”

Avatar: “So basically, what you wanna create is humanity?”

Humans: :0