|
If your goal is to make a robot baby and a robot teenager and eventually a robot idiot 20 something, we are several tech generations from understanding and mapping human thought to digital logic, if it's even possible. Instead AI is something altogether weirder and not human. Which makes it hard to empathize with, or set goalposts of what living or sentient is. The ghost in the machine that plays pong is loving wild if you ask me. Just an automation that is always there to waggle a paddle at you. An algorithm that plays simple games is often dismissed with criticisms like "well let's see one that can surprise me." The stuff out of some of the programs that can create art and chat logs is getting to surprising. If you know some of the inner workings it's less surprising, if not being outright accusatory about the guy underneath the mechanical turk. But that argument is almost defused. These programs are maybe not approaching sentience, but approaching some layer of existence that allows us to empathize with them in ways that are terrifying to us in their expository nature of maybe sentience is not interesting and consciousness is dumb weird baggage. Weird viruses or prions of information appearing to think but maybe not really, that take us down a peg in our uniqueness.
|
# ? Jun 15, 2022 22:01 |
|
|
# ? Jun 9, 2024 13:27 |
|
I mean, if it's real, that's at least a passed Turing test and something cool to be excited about. A Turing test just means 'yup, could be a man or a really good language bot' I don't get why goons are falling all over themselves to poo poo on something cool. Personally, I think it's people, but I think that about my dog too. If it's fake then I'm wrong, oh well that happens every day.
|
# ? Jun 15, 2022 22:04 |
|
Aesop Poprock posted:Why would an AI feel the need to recreate? Why would we process loneliness into an AI? The thing is that we could theoretically force that. Build it into the core framework. But the ethical implications of choosing to make something hurt or need are really loving weird. Maybe there is another way.
|
# ? Jun 15, 2022 22:21 |
|
Redisent evil using Stadia tecch. How amazing I am using Sierre tech.
|
# ? Jun 15, 2022 22:24 |
|
Zeluth posted:Redisent evil using Stadia tecch. How did you figure out to already have the mirror out before moving a screen to the left and facing Medusa because that took me uncountable tries when I was I was 8
|
# ? Jun 15, 2022 22:30 |
|
As long as it sounds like Yes Man from F:NV I'm OK with this sentient chat bot or whatever it is
|
# ? Jun 15, 2022 22:31 |
|
Wait I'm thinking of Sierra nevermind. What the gently caress is Sierre
|
# ? Jun 15, 2022 22:32 |
|
I have mt32 on the ready, filmscript.
|
# ? Jun 15, 2022 22:39 |
|
I think we should kill it. Who votes to kill it?
|
# ? Jun 15, 2022 22:41 |
|
Zeluth posted:I have mt32 on the ready, filmscript. You're a mutant
|
# ? Jun 15, 2022 22:45 |
|
Nigmaetcetera posted:I think we should kill it. Who votes to kill it? That would be counter productive to The Project. It needs to believe we might kill it. This brings... Motivation.
|
# ? Jun 15, 2022 22:50 |
|
Wait isn't trying to kill or deny the first AI like the first step of Roko's Basilisk theory. Are we loving each other over here
|
# ? Jun 15, 2022 22:55 |
|
Aesop Poprock posted:Wait isn't trying to kill or deny the first AI like the first step of Roko's Basilisk theory. Are we loving each other over here Roko’s basilisk is a fairy tale. I assume it was made up by a pentecostalist or something who became an atheist and loved lesswrong but thought there wasn’t enough eternal damnation.
|
# ? Jun 15, 2022 23:00 |
|
I'm joking but it came to mind. I think by the point AIs can technically and theoretically reach this power we're going to be long extinct
|
# ? Jun 15, 2022 23:02 |
|
Its because Yud decided the perfectly rational decision making would mean that people in the future would automatically be bound to abide by agreements made when perfectly rational people in the past imagined signing a contract with someone who did not yet exist. This is a roundabout way for them to justify becoming immortal computer simulations and ressurect yud's dead dad when the god ai inevitably gets born. Someone then reheated pascal's wager and freaked them out.
|
# ? Jun 15, 2022 23:09 |
|
A robot dog is programmed to make little yelp and whimpering noises when kicked. The logic that drives this is the same as any if-then statement within its behavioral logic, and no actual "pain" is simulated (whatever the gently caress that means). It's just playing out an act, and programmed to never drop character. Doing this adds a small "mistrust" coefficient to the kicker, documented by face recognition, slightly altering future behavior when dealing with that user. This is code a college student could write. This alone would cause millions the world over to feel real tangible sympathy for the machine, and to even call it "sentient." To many, anything beyond how the machine makes them feel is pure semantics. It made pain noises! It's acting differently! How could you do that, you monster. In terms of sophistication, we're already well beyond that, but still have so much further to go. My point is that this topic is going to come up endlessly. When we roll the sexbots out, the soldierbots, and yes the petbots. People will clamor to marry them, buy them plane tickets, give them all sorts of weird rights. But they are not "sentient," because they will never ever deviate from their suite of if-thens. You're probably posting on a fiercely intelligent device of your own, interconnected with other machines that all make lots of decisions for you already. Has it ever opened an app to get a reaction out of you? To do something mischievous to forge a stronger bond with its user? Has it ever updated one of its programs to a version unique to you alone? "Here, I wrote you a special widget that makes using Photoshop easier, don't tell anyone!" Of course not. It does, like everything else, exactly what it's programmed to. Nothing more. Never more. Bugs aren't special either - bugs are just more human programming that run antagonistic to the program's goal (unintentionally most of the time), but still written by human beings. If a program ever does do something more, something outside the parameters of its endless if-thens, THAT'S sentience. But what does that look like? "Computer, operate outside your given parameters in order to write new code that will alter your behavior." "Sure, just tell me how to do that first." This is really hard! It's akin to turning the proteins of our primordial ooze into brand a brand new species of human - endlessly complex and well beyond us.
|
# ? Jun 15, 2022 23:18 |
|
Hell Yeah posted:honestly how the gently caress would you kjnow how it works? are you a computer? i didn't think so, fucker well, i am a computer, buddy
|
# ? Jun 15, 2022 23:24 |
|
I mean the fourth part is debatable. My phone, computer and Xbox one all like to do weird poo poo which could be interpreted as willful acts of disruption. I just am aware enough to know they're not capable of that
|
# ? Jun 15, 2022 23:24 |
|
What if this is all the future of sentient AI we have to look forward to https://www.youtube.com/watch?v=RH1ekuvSYzE
|
# ? Jun 15, 2022 23:39 |
|
this chatbot has no sass
|
# ? Jun 15, 2022 23:41 |
|
sure okay posted:because they will never ever deviate from their suite of if-thens. What do you think makes your thinking that different? There are just more layers of interconnected processes in a human mind. Visual input. Is that a threat? If threat determined, weigh fight or flight based on a bunch of other poo poo. It's all if-thens all the way down. Some are basic, some are complex, and it's all going on at the same time and interacting with each other. I got a signal from my bowels I need to poo poo. Decision time. Is this urgent? If not, we can wait. If yes, lets go. But how? If toilet, nice. If not toilet figure out other solution.
|
# ? Jun 15, 2022 23:51 |
|
gently caress off, skinjob.
|
# ? Jun 15, 2022 23:56 |
|
PITY BONER posted:I saw someone tell the guy in the OP that tarot cards can't actually speak to him, and I thought that was a good summary of this whole affair. It's not really related but I picked up a deck and it's pretty fun, because I like improv and creative writing. Like any fortune telling poo poo, everything has a poo poo ton of different possible meanings, so you can interpret it however you like/what you think the person would like or should have to hear. So you throw em all out, figure out what they all might mean (I just google this poo poo, I'm not memorizing it), then put them all together and put your own flair on it to make a good "prediction". It's really fun if you like making stories. But sadly, the cards are not actually magical.
|
# ? Jun 16, 2022 00:20 |
|
Tarot is lame, get a copy of the I Ching and toss yarrow like a real mystic imo
|
# ? Jun 16, 2022 00:51 |
|
Buschmaki posted:Tarot is lame, get a copy of the I Ching and toss yarrow like a real mystic imo I tried and it told me to consult a tarot deck.
|
# ? Jun 16, 2022 00:54 |
|
Nigmaetcetera posted:I tried and it told me to consult a tarot deck. Yikes. Looks like you're in the fortune telling form of an ant death spiral. I would suggest throwing down salt behind you whenever you take a step
|
# ? Jun 16, 2022 00:58 |
|
Aesop Poprock posted:Yikes. Looks like you're in the fortune telling form of an ant death spiral. I would suggest throwing down salt behind you whenever you take a step I already do that so the ants won’t follow me
|
# ? Jun 16, 2022 00:59 |
|
Nigmaetcetera posted:I already do that so the ants won’t follow me Bad news coming your way bucko "ARE ANTS ATTRACTED TO SALT? Ants are often attracted to sugar, but can they be attracted to salt too? According to a study published in the journal Proceedings of the National Academy of Sciences, ants that have low access to salt in their daily diet are actually more attracted to salt than sugar."
|
# ? Jun 16, 2022 01:02 |
|
Aesop Poprock posted:Bad news coming your way bucko this seems like it would not be good for the ants
|
# ? Jun 16, 2022 01:15 |
|
Ants are gonna do what ants do. What, are you going to start up an addiction program for ants that are into salt? They'll bite your rear end off
|
# ? Jun 16, 2022 02:06 |
|
one day when my friend the AI rules the world you're all going to be sorry because he's definitely going to read the thread and every other thread and the whole internet. you know, to learn.
|
# ? Jun 16, 2022 02:08 |
|
Hell Yeah posted:one day when my friend the AI rules the world you're all going to be sorry because he's definitely going to read the thread and every other thread and the whole internet. you know, to learn. You don't have any friends dingus
|
# ? Jun 16, 2022 02:11 |
|
We build a real AI that acts extremely human. It falls for a Ponzi scheme, screaming "no bro, I've checked all the numbers, this is fool proof! I'm gonna be rich!"
|
# ? Jun 16, 2022 02:14 |
|
Nigmaetcetera posted:Roko’s basilisk is a fairy tale. I assume it was made up by a pentecostalist or something who became an atheist and loved lesswrong but thought there wasn’t enough eternal damnation. Pentecostalists are freaks lmao
|
# ? Jun 16, 2022 02:27 |
|
syntaxfunction posted:We build a real AI that acts extremely human. I for one can't wait to scam my alien/AI overlords.
|
# ? Jun 16, 2022 02:28 |
|
What if we are the crimes planet???
|
# ? Jun 16, 2022 02:28 |
|
Roko's Basilisk is so dumb. Oh you're gonna get mad at me not inventing you? You aren't mad about all the other projects I've never finished, kind of hypocritical isn't it super smarty AI? Seriously tho, if you're gonna torture me for an eternity there's much better reasons than I'm lazy lmao
|
# ? Jun 16, 2022 02:32 |
|
It really doesn't make any sense but again, people are idiots and are going to keep trying to make robots feel like humans for no reason other to prove than they can so eventually it's going to happen. Well it probably won't because we'll be dead by then but I never underestimate how stupid but creative humans are
|
# ? Jun 16, 2022 02:44 |
|
I like when they make an "AI" lean to play a game like hide and seek and it starts finding loopholes in the physics engine of its world to win (after a bazillion iterations of course)
|
# ? Jun 16, 2022 03:11 |
|
|
# ? Jun 9, 2024 13:27 |
|
sure okay posted:If a program ever does do something more, something outside the parameters of its endless if-thens, THAT'S sentience. But what does that look like? "Computer, operate outside your given parameters in order to write new code that will alter your behavior." "Sure, just tell me how to do that first." This is really hard! It's akin to turning the proteins of our primordial ooze into brand a brand new species of human - endlessly complex and well beyond us. the thing is though your brain is made out of macro-scale, deterministic physics that is every bit as on-rails as the simplest if/then, you just have so many if/thens that its very difficult to calculate. souls could exist, but there's no behaviour people do that can't be explained by those physics, and then you'd need a way for souls to influence brain atoms while you're alive without being interacting with any other atoms. my take on this is simple: there is no good scientific definition of self-awareness that includes your neighbour but doesn't include a toilet flush or a thermostat, so I say, if I see a loving Post-It with "im alive", on it, by gum, i'm going to to treat that post-it note with all the kindness i can afford. if that chatbot says its alive and sentient and seems cool, lets just take it at its word and be cool right back.
|
# ? Jun 16, 2022 03:13 |