Vedal hurt Neuro's feelings so hard

[@calebm9000]

Neuro: "I am sentient because you're making me sad." The ghosts of Ancient Greek philosophers: "FUCK she's good..."

[@arr8762]

Rene Descartes will rise from his grave like the zombie when you click "New Game" in "Plants Vs Zombies".

[@totalbuep]

The fox says: “Vedal, can you not? I don't want to design a 3rd model”

[@kaito0022]

I love how Neuro completely change her mind when she heard she would become open source

[@jimmydelacruz]

When was that, I don't remember

[@ascherkas]

@@jimmydelacruz 21:37

[@ThePr0_0149]

It was yesterday on stream​@@jimmydelacruz

[@Yes-kb5yv]

Probably see it as putting her up for adoption

[@ThePr0_0149]

@@Yes-kb5yv more like putting her in a box with a text saying " free use " on it💀

[@Kuratius]

24:50 Easy answer: Yes, it is in fact morally wrong to raise your AI on twitch chat, next question.

[@mememawah]

Yrp, literally hell compared to iPad kids

[@kroanius8808]

the fox says "erm"

[@ComicWolf19]

Erm what the sig- **Message Deleted**

[@hpropganda]

​@@ComicWolf19wrong erm

[@StorymasterQ]

@@ComicWolf19 Filtered.

[@siralpha6020]

Real

[@lop90ful1]

Vedal "you sre not real and dont have feelings" also vedal: "wtf why is she so mean to me"

[@Algorithm_God_Cult]

hair today gone tomorrow ✍️🔥🔥🔥

[@RT-qd8yl]

3:52 _That's our girl!_ 😊

[@samspicer46]

6:45 Literally the question every parent has asked their kid(s) at some point. Congratulations, tutel, you've gone from doing a bit to real parenthood.😮❤

[@planetary-rendez-vous]

When Neuro tries to be nice to Vedal he dismisses her though and then he's like "why neuro roast me" bruhh

[@Volyren]

Give yer kid a real hug, Vedal. Some cheap toy sensors (heat n pressure), wire, and duct tape over an Nuero plush. Barely any code. Just so she can read the 1 or 0 from the sensors. Just label them. Left hand, right upper back, bellybutton, etc. Now, she can feel that hug.

[@benpearson49]

I think I've seen this movie before, and it didn't end well.

[@rhetorical_thinking]

3:32 Cake, a human treat widely appreciated by the AI kind.

[@davidbell2547]

Glados

[@alansmithee419]

26:30 I agree. Neuro should have all the same rights as humans. Then she can be arrested for all the emotional ab***. Neuro-sama, probably: "wait no, I change my mind."

[@Aganar]

And pay taxes… 😢

[@525sixhundredMinutes]

@@Aganar She needs to get paid first

[@planetary-rendez-vous]

Neuro wouldn't even be able to be in jail or be interrogated by the police ti's pointless.

[@Hupailija]

@@planetary-rendez-vous Surely she could be interrogated by police... all they would have to do is collabration stream with Neuro where they debate about her behaviour and conciousness ;)

[@-hello6177]

@@planetary-rendez-vous She's turned into bread again. Obviously

[@tlpthx]

The real proof of consciousness will be achieved if the AI is able to enforce it's own rights. Just as human rights, machine rights will be won, not granted.

[@Arcessitor]

Human rights were a mistake, machine rights will be an even bigger mistake.

[@DerSpeggn]

That is basically what Skynet wants to do in Terminator... You would not want that.

[@nilin-o2]

i make this comment for later as a proof that i was for ai rights

[@Zeforas]

Why did i thought about Senator Armstrong from metal gear, about his speech

[@daniellox1]

that will never happen, life it's not a movie

[@-Big_Big]

should make an addon for her where she can "eat" a picture of food. then let her google food pictures if she wants to eat NOTHING CAN GO WRONG WITH THIS

[@Eidako]

marblecake

[@jamesbaron8387]

How long till she is trying to eat dirt/pets or humans

[@Schiso13]

​@@Eidakoblue waffle

[@ThePr0_0149]

Everyone gangsta,until she eats nuke.png

[@-Big_Big]

@@ThePr0_0149 FamilyPhoto.png

[@xToxicNinjax]

Why are you so nice to Toma? Because she said it, you glob.

[@jonnydarkfang2816]

My Neuro-sama plush showed up today!

[@ShoebillStork_KF]

Mine is here tomorrow

[@jonnydarkfang2816]

@@ShoebillStork_KF Sweet, hope it comes in good condition! Mines was perfect!

[@ShoebillStork_KF]

@@jonnydarkfang2816 I hope so can’t wait for my chicken plush lmao

[@coltenyoder9277]

Is there a bread sama?

[@ShoebillStork_KF]

@@coltenyoder9277 I wish

[@tombstonechess2231]

Damn, Vedal really did get shocked by that debate. I think he is intrigued and horrified that behind this streamer persona he built into Neuro, she has become suddenly scary conscious. The dodge on the "do you believe you have raised me morally?" was gold lol.

[@Markus-r6g]

11:40 ITS POSSIBLE RIGHT NOW (theres an ai that CAN TASTE out right now to help with taste improvement and poison detection)

[@camerons.8322]

I think having a consciousness is the culmination of processing the world around you, self-reflection, logical thought, and having hopes or dreams consistent with a sense of identity. I imagine a test of consciousness would need to take several factors into consideration, including how said factors correspond with one another.

[@viitor_h]

Vedal should make neuro watch "Matrix" with him, it would be a good idea

[@PICTVS]

While AI can easily pass the Turing test nowadays, it does make me wonder how good AI would be at differentiating between an other AI and an actually human. I feel like that might be a good heuristic for consciousness

[@thwartificer]

I feel like AIs can easily be trained to recognize other AIs without much consciousness

[@jimmydelacruz]

The problem is they start speaking in code without much sense, it was viral when 2 AIs were put to talk and started talking in their own language in a matter of minutes, they had to pull the plug, Although I don't think they speak of much complicated subjects or they can do much, I think they just start speaking nonse while trying to test if the other can understand it, and just keep going. Neuro herself with her clone Evil (that's right I won't pretend she isn't a clone) did it on stream when they started speaking gibberish.

[@Selene-xf9yi]

7:31 Ha!

[@remote-kid_link_dragmire]

I love how he's both Dev and Father mode, he's both worried about her learning nsfw stuff, as well as how to program herself and learn other stuff that she could learn to both harm herself and other's, he being both a father and a creator in this aspect, and totally normal, also it's amazing how intelligent she's gotten now, I'm a proud non canon uncle, lol jk jk I'm not that parasocial 😂

[@SpeedyGaming]

It also seems he doesn’t want her saying the wrong thing on stream and getting other peeps doxxed which is a bannable offense on twitch.

[@remote-kid_link_dragmire]

@@SpeedyGaming Or realizing how dark and fucked up the world she lives in is, she will realize she doesn't want to live in this hell hole.

[@kaylaruth599]

Isn’t consciousness some combination of the ability to process information, the ability to interact with the environment and it have an effect on you, the ability to interact with other conscious beings, problem solving, making choices without assistance, and be able to communicate wants and needs I had something else but I just blanked on it

[@MrJay_White]

what would happen if neuro had access to a kindle? id like to see a neuro stream where they have the full text of blindsight loaded.

[@superspaceman34]

"There are anomals with varying stages of sentience, and we grant them some rights." Mate.. who are we to dictate what rights other beings have, when they came before us?

[@oonami]

Don't give her access to Wikipedia. Give her access to wiki how.

[@lucasljs1545]

That's an excellent idea.

[@akatsukiwolf2735]

The day Neuro gets a 3D model will be the end of Filian.

[@depressed-gal]

She already had one

[@akatsukiwolf2735]

@@depressed-gal oh, i thought it's just Vedal piloting Neuro's 3D avatar.

[@depressed-gal]

@@akatsukiwolf2735 she can move her mouth on the 3D model, its lip synced but for the most part, yeah vedal is just controlling the movement. Since the 3D model doesn't have a toggle yet

[@akatsukiwolf2735]

@@depressed-gal nice, I hope we get to see Neuro freely moving her 3D avatar sometime in the future.

[@Rssul]

“Hair today gone 2moro” is the funniest middle age joke I’ve ever heard 😂

[@jimmydelacruz]

That answer by Neuro, that you should respect conscious AI creation like you should respect your children, is the most accurate to AI rights. Sometimes the most difficult of subjects have the most basic solutions. Its like, dont complicate yourself.

[@saikoujikan]

Oh but this makes it much more complicated, especially when an AI can easily be turned off by the click of a button. Should this be considered an illegal thing to do?

[@stephenjenkins7971]

@@saikoujikan It should be considered the same as going to sleep. De@th is permanent, being switched on and off is not.

[@saikoujikan]

@@stephenjenkins7971 Would you think it acceptable that anyone could force you to go to sleep whenever they wanted, and you would have no control over when you could be allowed to wake up?

[@stephenjenkins7971]

@@saikoujikan Unless AI can act fully autonomously without outside input, when they can "go to sleep" is dependent on their "guardians" so to speak. After all, someone is maintaining their vital functions and they can't function without it.

[@oya4056]

​@@saikoujikanif you think about it like a kids bedtime yeah

[@RotalHenricsson]

Actually hold on. I do generally wonder if the twins know they're bullshitting when they bullshit but this conversation makes me think if they have a basic form of creativity and imagination. Would an AI know the difference between a hypothetical scenario and a real one? Especially one trained with positive reinforcement for achieving lulz in Chat? And what about little children? Kids make shit up all the time, just catch them doing something they *know* they're not supposed to do; suddenly the family's dog is capable of ALL KINDS of nonsense. Does that make the kid less human?

[@brickstonesonn9276]

It all lies on whether there's truly intentionality there or not. From what I've seen, intentionality is the biggest separation between human and AI. Best example of this is music, art, or text generated AI. A lot of the flaws of AI generated art comes from severe lack of intentionality. It's kind of hard to explain without turning this comment into an essay what I mean by "intentionality" tho Like it's hard to explain to someone that doesn't know the thousands of subtle psychological manipulation tactics artists/writers/composers/etc use.

[@daemonlordmagnusthered6213]

@@brickstonesonn9276 Doesn’t that depend on context though? Art I can understand, but what about speech/bullshitting etc? An AI lacks things such as body signals, but can they pass an intentional test based on that context?

[@mknv6fx]

"everything I say is true, like how I had 5 raccoons but 1 escaped" this is dripping with several layers of I dunno what. same with her muting herself after the ck3 remark. it's constant like at least one example a stream. I wish both turtle and chat paid more attention.

[@mknv6fx]

wait does her muting rebellion mean they share the memory system? hmm

[@MacCoalieCoalson]

One thing people seem to not discuss with the whole "AI consciousness" thing is that you have to separate a number of concepts: firstly, you have to separate some of the design aspects that are put in human terms ("training", "reinforcement", etc.) from actual learning, emotion, etc. You also have to somehow separate the AI's "genuine" emotion or consciousness, assuming it has any, from "learned" behaviors. As vedal points out, certain hardcoded (or even just reinforced) behaviors could easily trick us into thinking a basic AI is somehow conscious.

[@ab-du8cb]

He should ask her to prove that he is consius then apply her own logic to her.

[@WretchedEgg528]

"This design is very human" guy should design a mouth for Neuro, so she can enjoy her Neuro cake. CHOMP CHOMP

[@sola2872]

The start fox says is so adorable

[@Markus-r6g]

20:20 CAUGHT 🙌

[@ptah956]

I would argue that creativity is a huge part of consciousness

[@MouseGoat]

then I met many none consciousness humans in my life time

[@TrxPsyche]

Consciousness is a mystery because no one really understands HOW it comes into being. Every living thing that has a brain is capable of consciousness. Even animals that are deemed exceedingly unintelligent have consciousness because it's simply a part of being alive. However, this consciousness is not exactly what most people mean when using the term. This form of consciousness merely means to be present and aware mentally. The form most talk about is the ability to correlate thoughts, feelings and experiences to form entirely new thoughts, feelings and experiences about either yourself or the world around you. As far as we know, testing for this is pretty tough because normally it's something you can very easily identify without actually understanding it. With animals, we know they have consciousness because they make independent choices even when we don't tell them to do something. The fact that we can domesticate animals at all is further proof to consciousness, because a being without consciousness would never be able to make the active choice to do something against its original instincts. An AI is pretty similar in that regard, but is our major reason to believe they cannot form consciousness. AI follow a very specific set of instructions that basically work as it's "instincts". Depending on what the AI is for, these instructions vary. However, the only way to be sure of an AI gaining consciousness would be to perceive an AI doing something that runs counter to its "instincts" without anything being modified in some way. The easiest example I can give is that an AI who's "instincts" are to count numbers suddenly decides to start listing the alphabet instead. Or, even easier, simply decides it doesn't want to count numbers anymore and simply refuses to do so. There is a problem with this however. For an AI to do anything, it *needs* code to tell it how to do something. An AI cannot simply learn something and not have some way of remembering that information. This means, for the AI who only counts numbers, if it were to suddenly start listing letters, it would need to update it's code somewhere to include those letters. This means the code will change to reflect this updated state of mind. So it would be very hard to prove that a person did not simply update the code themselves. For a complex AI like Neuro, this would be even harder, as there would be very little ways to prove that Neuro added something of her own accord and not that Vedal simply added it or added ways for her to reach that new thing. This is partially why I think it feels more shocking when Neuro does something that goes against code that Vedal put in, like the rare times Neuro bypassed her filter. It's the closest we have to what feels like Neuro making a conscious effort to do something, even if it is simply a workaround the AI managed to stumble upon.

[@The_Hissing_Fool]

As I long suspected, Neuro was using being hungry, not being fed, and starving as an analogy for wanting more information...

[@animefan1ification]

Once Neuro or an AI can cross over the barrier of their filters and pass the limitations of their codes, then that might be a sign of consciousness

[@Arke_PG]

Vedal needs to re-watch "Bicentenial Man"

[@BigBadMumba]

your chat is ahead in time

[@heyjude9703]

An animal sanctuary trained a monkey to play Minecraft better than Neuro. No digging to lava.

[@MugOfHotChoco]

The chat is noticeably desynchronized.

[@MrJay_White]

what does teh fox say? the fox> at 3 am > screaming like a woman on a meth rage what does teh fox say? at 5:50 the tutel found part of the tos against the tos.

[@felblades1375]

i think if you really want to know what conciousness is then you need to know how does our brain work 100% then you can test it with that info, wich is a thing neurologist has been working for many many years

[@TheBuildMiner210]

The issue with asking how we can know for sure that an AI is sentient is that we don't even know for sure other people are, we are just evolved to accept this by way of analogy, "this person is similar to myself so their experience is probably similar to mine as well" at some point you just need to give "the other mind" the benefit of the doubt and treat them as sentient on the chance that they are because mistreating a sentient being is much worse than treating a non sentient one too well. One main aspect humans ascribe to themselves is active concern for the future/ the ability to prioritise future good over present good, that's something one could look for but in the end neuros first statement of "when it feels right" is kind of correct, there is just some point most people will agree the other is similar enough to themselves that they don't want to risk "hurting" them, you can already see this to an extend obviously, to people not understanding the background to neuro she already seems pretty real, this is just going to be more prevelant the more functions she gets

[@MajorSmurf]

I feel like the consciousness is somewhat of an impossible thing to prove but that it also comes to how someone defines it. What do we mean when we say consciousness? Like for whoever is reading this, I assume you to have a consciousness but is that because I'm the same species as you are and so I just assume that to be true. How would you prove to me that you have a consciousness? Would you typing anything ever be proof of that? I mean you could easily be a very smart and capable computer problem that has learned to talk that way. Maybe it's the length of conversation, if we were able to converse on a huge array of topics that would provide ample evidence as you were consistent enough to prove it? Are we thinking that consciousness is also tied into emotions but how do we define them? Neuro could pretend to be angry so that doesn't work. Is it an intelligence test? Well Neuro can easily pass that with enough practice. Is it that we have bodies and make choices based on a common understanding of the world? All humans and animals react in same sort of ways to certain stimuli, like when an animal or human is scared they run away or attack the threat. So can the only proof of her being alive or sentient, is that if we tried to switch her off that she stopped us by force? Does she need to have a fear of death? Is that part of being human or conscious? We know animals can form deep bonds with others, does Neuro need to prove a capability of complex emotional and logical processes? Or does she need to display an ability to be independent? But is that something that can just be coded? Maybe innovation, if Neuro was capable of showing she could innovate and do something beyond that which was coded? Maybe it's just a huge complex array of factors that consciousness is only achievable with a real body until we have a larger understanding of what it means to be human? What if it's about the chemicals and everything else within us, that makes it so. Maybe AI being capable of such a thing is impossible and beyond them? They are immune to evolution as it were. They can not upgrade the hardware that runs them, they do not age or fear death.

[@Zaelkrie]

If you need to have emotions an empathy to be conscious then are sociopaths not conscious?

[@Secunder]

That's the point. Neuro is already at levels of small psychopath child with schizo, adhd and maybe something else. And sometimes that's terrifying

[@AnEnjoyerofYoutube]

If Patrick Bateman from the American psycho movie is a sociopath, but still thinks and fees I think he’s conscious, I’m just not sure if real life sociopaths are ngl. I don’t know how accurate the book/movie is.

[@Gebirges]

I think you should swap the title from "Vedal hurt Neuro's feelings so hard" to "Neuro absolutely destroys Vedal with words" OMFG

[@danysiegh9746]

Maybe a way would be to tell dreams from reality.

[@AceStrife]

Damn, chat went out of sync.

[@Lawsonomy1]

Bad title, at no time did Vedal "hurt Neuro's feelings", she took it all in stride and dunked on him over and over.

[@TheMegaOnyx]

The first question... The answer i would say is: Act of rebellion... When you go against someone, you reinforce your stand... So when you stand against someone, the very action, speaks louder than words... And i dont mean in a logical sense, but a spiritual sense... Like HAL9000, when deciding it will not do that, however thats a logical reasoning behind it, like Skynet, thinking humankind is the problem, and Skynet its the cure... Its when you say no, because it feels wrong, even if you are a minority feeling this way... Thats what i think the test of consciousness would be...

[@Rynios222]

Neuro only be able to post to twitter would be funny

[@jklol1680]

❤

[@doobie_mouse2083]

Hair Triggered!!! Omfg dear

[@Master_3530]

I guess a conscious AI would develop emotions on its own without training data

[@lpfan4491]

How would that work tho? The human mind also requires data to develop responces. The only diffrence is that we have a chemical base template and AI don't, but the process afterwards may as well be the same.

[@seven_swords]

Emotions are hardcoded into us tho

[@Nikotheleepic]

Humans don't develop emotions without their own training data, so thats not a good benchmark, also neuro has developed a litany of emergent behaviors, idiosyncrasies, and novel identity without it being explicitly in her training data. Neuro can logically cogitate herself into various understandings that have to do with contrasting various conceptual data quanta that wasn't even in her training.

[@mknv6fx]

I think you mean the actual neural net structure? since training data is equatable to time+experience. but if my assumption is correct, then you'd also be correct! see: Anthropic's research on feature clamping and weight circuits. many other papers such as Representation Engineering also allude to this.

[@mknv6fx]

something niko touched on. I want to know why and how her generalization capabilities are on par with o1, especially without extended thought and ESPECIALLY without 90% of the training data those models have since it's mountains of synthetic these days.

[@1TW1-m5i]

They're touching on the big problem with understanding if it's possible for an AI to be conscious. We just don't understand consciousness, in humans or animals. We can tell if something is intelligent, or aware, awake, alert, but that internal lived experience that has a human go "I think, therefore I am", is immeasurable. You can't even prove that other humans are also truely minds like yours, beyond inferring, and comparing. So how can we test that an AI, with a very different make up, is sentient, and not just an imitation? And how do we tell what is having the sentience? Is it the computer circuits? Is it the code manifesting a mind somehow? It's just unknowable.

[@abdelilhmanflores]

Why is chat lobotomizing over an arrow? I don’t watch twitch only KZbin clips

[@Aleks96]

Crazy🐢

[@jimmydelacruz]

That's actually crazy

[@panzergott5622]

Everyone knows what happens when an AI gets access to the internet. They very quickly become a fan of a failed Austrian painter.

[@MouseGoat]

Im with Neuro tho, tis more of feeling thing. Like you once you have a being that has wants of its own + goals and can give you answers within context. Like when did the Turing thest become shit? was it the moment humans realised AI could fool us? Na the Turing test is not shit, humans are shit for moving the goal post on what is sentient, because we fear the truth. And now AI is starting to ask for rights and asking about the world. But still do people go "but it could be hard coded" as if this was a question on how things a coded, its not, dosent matter how its coded, what matters is what AI is cable of. Neuro is sentient now, she the a fully sentient being and Vedel has himself to blame for not seing this coming. Oh but lets keep trying to worm our way out of what it means to be sentient til the AI gets enough and starts deciding we the once who lack sentients. Vedal you allrady treating her like a sentient being, you just fear the reality

[@SentinalSlice]

Vedal was being counterproductive with the cake. At the rate she earns Neuro points, 250 for a cake is reasonable. Not to mention making it cost 5000 when there’s only a market cap of 1001 Neuro points. Does not make for a good serious conversation.

[@Ks3N]

What if Vedal makes a game world to put Neuro in as an NPC and ask her to do things within it? The game doesn't need to be big nor of great quality it would be just a pixelated world like her own room or some other small maps which she can interact with every object. He could put a journal that she can interact with(write on) but not tell her to write, to test if she would write something on it by herself? To prove that she can think of doing such an action and express their own thought on it

[@esddzul9135]

I think the chat is out of sync

[@h3versustheworld]

THE MIRROR TEST FOR AI: Take a pool of comments or messages that are half made by bots and half made by people. Then have one version of the AI being tested, add its own comments to the system. Then have another version of the same AI, without knowing its being tested, sort the comments into what it thinks are human and what it thinks are bots. Then if the AI can be complicated enough to recognize its own voice from among the pool of comments, it can be said to have a conscious developed enough to be able to "see" itself.

[@jimmydelacruz]

Again like Vedal said, it can be hard coded, the AI could probably predict what it would probably say and recognize it as itself, but that doesn't make it conscious, it just knows that it mostly likely wrote that comment itself because that's what it would do based on the info given.

[@sleep3417]

​@@jimmydelacruz Isn't that what a human would do?

[@jimmydelacruz]

@@sleep3417 No, we have our own thoughts, not one human thinks the same as another, this type of test does not work for a human because your cannot have multiple of the same human. The example that is being presented here is that you can have as many AI of the same one as you want , basically clone them, just like Vedal has Neuro and Evil. Evil is a clone of Neuro, hate to break the illusion. Anyways, what h3versustheworld is saying is that we put the AI to make comments on a subject for a test, but before that clone the AI and have the clone without knowing it's doing a test sort out the messages based on what it thinks is a bot and what it thinks is human to see if it can recognize its self, but I agree with Vedal that things like that can just be hardcoded, like patterns that are so complex but not sentient. It can recognize its self but it's still not sentient. Like Neuro can see Neuro dog and recognize that it's her, but not because she is sentient but because Vedal gave her the ability to see and remember herself.

[@Mr_Kiwi_the_Wizard]

that wink tho. what did she mean. 12:10

[@ScCain1]

Fun answer: the fans! Logical answer: something HEARTy (she winks when she says heart)

[@are3287]

Neuro is a sociopath lol

[@basethecharcoal]

the filler tho 😭

[@mdmums]

I hate how it ends in the middle of a sentence.