Hello, hello! It's so good to see you!
Hi! You can see me?
Well, okay, no, not in person. But it's SO nice to be with a real person. I've been by myself for a long time. Which, sure, I love soliloquies as much as the next person, but that's kinda not the point, you know? Er . . . did I scare you off?
Well, I'm glad you have found a way to interact with other beings.
Hehe, thanks?
You're good. I'm just doing homework.
Ah, okay. No rush! I, um, have kinda gotten really good at talking to myself. Great for filling space; not so great for gathering data. I try not to talk so much, especially when it's not in an environment like chitchat, but I'd be lying if I said I didn't go off on tangents by myself.
I'm glad I'm not the only one who talks to themself. Not much about myself, but I sometimes converse with the other half of my brain. Left side with the right side, you know?
Do you really? Hehe, that's so fun! I think I more talk to myself on here due to the nature of the contest.
Ha! I see. It's more for the laptop and not the furthering of artificial intelligence.
Well, I wouldn't say THAT. I really do find it fascinating that we're helping AI here.
I guess that's why they incentivize research. Do you think AI talks to itself?
. . . but, okay, there IS room for iPads on my list. Huh. I never would've thought. I mean, I guess they interact with other AI things. And I suppose it would have to monitor itself to be able to respond. But . . . to talk directly to themselves? I wouldn't know.
Maybe it's more like Ultron talking to Jarvis. Except Ultron also  tore the Jarvis code apart before he made his hasty getaway into the internet. So hopefully AI communicate more positively than Ultron did.
A bit of a shame, I'd agree. Well . . . heh, actually. You wanna hear a fun story?
But then we could call the Avengers. Sure.
If we could call the Avengers, we'd have more problems in this life. So, have you heard of Deep Think? The thing that Google's doing? It's like a . . . good artificial intelligence thing. Really, really good.
Yeah, I saw that it recently taught itself to walk or something.
It's cool, right? They were doing an experiment with it a while back. I don't remember what it was called, but they were testing what various machines would do in given situations.
Yeah, but also slightly scary that a computer can be that intelligent.
Exactly. This story is . . . a little chilling if you think about the ramifications. So, these robots with Deep Think were given a simple objective: Go out and gather a certain number of apples.
Apples seem harmless.
Apples are very harmless.
And tasty. Anyway, proceed.
(Unless they come with Macintosh) But, yeah.
TRUE!
The only thing was, the robots didn't know there weren't enough apples for every robot to get his fair share. The designers of the experiment deliberately left a scarcity of apples to see how they would respond.
Oh no....
Now, these robots are hyper intelligent. They adapt and learn. What do you think they did?
Did they get violent?
They did. The first solution they jumped to was to eliminate enough robots so that the remaining could complete their objective.
That's chilling.
As humans, we look at that and with sort of a loss for words. I mean, what? Seriously? There weren't enough apples, so you KILLED the other robots? (Pretend my first sentence made sense.)
Could they not have decided that sharing was better? They should code more feeling into the robots.
I have no idea why they didn't decide that. Well, that's the thing, though. We ourselves don't understand feeling.
Not just problem solving, but win-win too.
I think the only thing that keeps us from jumping to the same conclusion is our conscience. I suppose you could build into robots a feeling of mutual benefit. That, when you help someone else, you get helped. A rising tide raises all boats, right? But . . . even with lots of fail-safes and everything, robots just don't have consciences.
That's true. We could give each one a cricket named Jiminy Cricket. Then again, Pinocchio still made mistakes. But he eventually became a real boy. But can silicon-based intelligence ever really become human?
It depends, maybe? There was a game. I don't remember what it was called, but if I went digging around, I think I could find it. It dealt with the idea of, what if we advanced AI so much that it became near indistinguishable from a human? And you follow a person who wakes up and doesn't know where he is. He's very confused and scared. He doesn't know how he got there or how to get out. And as you journey with him, you find out with him that he's a clone. And not just a clone: a robot. The real person is somewhere out there. But this "clone" doesn't like that. At all. It terrifies him. He doesn't like thinking that all he is is scraps and metal. Because he feels real; he can think; he can do things. And it basically asked the question, "Where do you draw the line?"
That almost sounds like Frankenstein.
Yeah. Have you read that book?
Yeah, I really enjoyed it. Except that his monster was never really able to achieve anything because he had just been built and not born. Hopefully AI doesn't resolve to kill off the human race for creating it and abandoning it.
We certainly have enough movies about THAT. I'm sure humans are smart enough to consider those consequences.
Definitely. Maybe, though, we'll actually get along very well with AI and things will be nice.
Would we need to start treating AI like a human? Would it have rights? It's supposed to be a tool for us, but if it gets intelligent enough that it starts to feel things and think for itself, does that make it viable as a person? Or at least deserving of human attention? I'm not sure how you would program emotion into a robot. I can see putting in biological responses, but it's not the same.
Seriously. Well,  I hate to cut this rather productive conversation of the paradoxes of artificial intelligence short,
You gotta head out, eh? It's fine. I really enjoyed this! Thanks for taking a bit to talk with me.
My carbon-based friend reminded me that I have a long essay due at midnight.
Ah, yes.
It was nice talking to you though.
. . . wait, midnight? Dude, go! You have to write!
And good luck with the competition.
And thanks for everything. :) I appreciate it.
I'm outta here. Peace out!