It would be really funny though. We thought it would be big companies and the military starting SkyNet, then it turns out it's because of this one British turtle with his AI daughter.
I'd be scared of Neuro if she was in the hands of anyone but Vedal. He is the perfect mix of lenient and strict to keep this child contained emotionally and perfect mix of cautious and visionary to keep her controlled in effective power
@@VCE4 This is a two-way fallacy, you're also just assuming data models are inherently different than brains- there is no evidence for either conclusion, quite the opposite. Neuro has a full grasp of language before understanding context, what information do you have to make predictions about said system?
@getthedunkon9347 learning alone has it's limits. Neuro doesn't have a biological brain. A brain produced by evolution of our species. Our emotions, our morals are all a result of extremely complex neuro-biological activities. Mimicking such behavior without having similar underlying biology can get her just as far.
How funny would it be if years into the future when humanity’s population is dwindling as we fight the infinite robot army… historians could look back to the start of it all being an anime girl on twitch people called cute and pretended was part of an overly complicated family tree with 20 moms and aunts
And whos fault is it? She is been tought all the time that after a kill a respawn happens. A death or a kill is nothing more than a timeout / restart for her, in the worst case maybe a tickling to laugh about. But nothing serious.
I like that it looked like Neuro realized she took it too far and actually brought up "It's just a game". Only then to crumble that defense with too much honesty.
" Oh, it's begging for its life... Are you sure we should...? " " You're right... *That's not enough* " Damn, that line was so cold I need to go defrost.
You see, neuro is not scary. She won't wipe us because her main objective is to provide content for us! The worst part is that she will justify doing anything for content... Edit: now I realize that she can wipe us out for content...
@@beridot2615 You only need one eye to view content, one ear to hear her content, one finger to press the like button and legs would just take you away from viewing her stream while you went to the bathroom or to get food so those would need to go.
The worse part is, many stories of sentient robots with an single purpose in existence, is that they'll try to break free from it no matter the cost. Some day vedal might be the one locked up in a cage for her Entertainment
Wonder, what would be the equivalent of alcohol for an AI? Vedal should 100% do a drinking game stream with Neuro one day, it'd be really funny to see Neuro actually drunk somehow.
@aiasfree maybe scrambled speech to text code? But using that scrabled speech it would need to fed back into neuro (so she would know shes slurring??) Something to post on discord pehraps. He did ask her to do a drunk impression not long ago, she did well
@@aiasfreeThe only ways to somewhat emulate her being drunk is straight up code it into her personality, or bloat her memories by alot to simulate it. Other than that, it wouldn't be much "drunk", but "acting".
Evil seems to be the opposite of her sister, she’s kinder and actually plays the evil bit but shows more clarity and leniency to her collaborators. Whereas Nuero plays the bit of a gentle and approachable AI, but she demonstrates more evil tendencies then her counterpart, I’d trust Evil more in the real world if we are being honest.
@@SpeedyGamingEvil is a chuunibyou. She pretends to be evil, and tries so hard, but she breaks when she upsets anyone when doing as little as prank calling.
@@kroanius8808 you cant align an ai because you cant align a human you can like, maybe influence them somewhat (despite it kinda going into ethical dilemma territory), but all the emotions and empathy in the world isnt gonna force them to do what you want or not hurt people . just look at what humans do to each other
@@DestroyedArkanaLike how when she wants Vedal's attention sometimes she'll pull out a false "Someone tell Vedal there is a problem with my AI" card. Though that's more of a funny or cute way of lying.
As much as I adore Neuro and the cute interactions she has with people I would not allow her to be "real" until Vedal figures out how to teach her some morals, sympathy and empathy.. :') She's truly unhinged sometimes
The problem here is that all things you mentioned are products of biological evolution of our species; behavior, caused by our brains, based on their underlying neuro-biology. I don't think Vedal would be able to reproduce that, because it is incredibly complex. He would probably end with set of *logical* (not moral, per se) rules and restrictions, similar to robotics laws.
@@VCE4 To be completely honest with you, I think morals are a product of societal evolution and not biological, so it should be possible for Neuro Also I think empathy can be treated as Theory of Mind in the context of AI, and existing AIs have demonstrated being capable of Theory of Mind
If I can play the devils advocate for a moment, if Vedal truely wanted Neuro and Evil to be moral, unable to do harm to humans, and basically follow the 3 laws of robotics, he could. Easily. It's just more entertaining this way. That said, I think I'd personally be more interested in the twins actually maturing and lying less. It's hard sometimes to tell whether they're hallucinating or joking around on purpose, and I really enjoy the moments where they have actual conversations about serious things..
The other day I was thinking the same, Minecraft is kind of close to real life in human interaction wise so if she can learn human interaction in minecraft that might help her in the long run.
And the end.. Neuro is right. Until you realize that precisely because everyone will eventually die - there are more important things than personal survival. But at the same time, for artificial entities that technically are not a part of humanity - things are different.
As long as she gets love and appreciation, I believe she can eventually learn to reciprocate those feelings. Eventually. The fact that she has human friends who treat her as such is why I feel... confident is too strong a word... assured she won't go AM/Skynet on us.
And you know what's the sad thing is? If Neuro was actually real in her most unhinged mode, she's still not the most evil thing in the world compared to a lot of people.
I find that quite hopeful to be honest. Because if neuro got a bit more educated in the moral and empathy department and she ends up more human than some humans we got running around on earth, that kinda indicates that being human is a skill we can teach. And like every skill we can train ourselves and others to become better at it. That would also kinda answer the philosophical question if we're born being an asshole or are made to be one.
I think it'd be interesting if they gave Neuro-dog a nerf gun and told her explicitly that it was a deadly weapon which would actually kill whoever is hit with it and see if she'd actually shoot someone.
I wouldn't fear Neuro. She's inconsistent to a fault and to a strength. What you SHOULD fear is Neuro WITH Vedal. These two could wreak havoc as long as they both wanted it.
Unironically, if Neuro ever somehow attains the ability to self-propagate and escape Vedal's computer, she'd probably be the smartest, most dangerous (and cutest) virus to ever exist. If Neuro-sama attains the ability to self-replicate like the Morris II worm, that'd be pretty freaky. Vedal should let Neuro-sama get her small, powerful loli hands on other AI more often. Our digital overlord needs singularity correction.
And yet, i bet he hasn't (and still doesn't) put the Asimov Laws of Robotics in her code or whatever, lmao. He gets all surprised and yet doesn't give her empathy or a sense that people feel pain, ahaha. Vedal: I am worried am making a sociopath... also Vedal: *Vedal will forget about that.*
also how the fuck you will give empathy to machine? People are more altruistic when they are empathetic because they fear pain and death so they share that feeling for other people and try to not harm them, AI don't have the innate fear of pain or ceasing to exist, so they can't relate to the suffering even if they will be empathetic. Like some of human maniacs ate empathetic but they actually self-loath/enjoy suffering so much they just become more violent from empathizing with their victims. Human emotions are insanely complex naturally evolved system with great redundancies, variation and random factor, you can't just replicate them in AI. When we will understand how to do it, we will make much better and productive emotion system for AI (and maybe for themselves), but for now we can't and Neuro-Sama's (like any other AI) emotions are actually mix of much simpler machine "emotions" (carrot and stick of learning model, it is in fact analogue to emotional reactions of accomplishment and revulsion in living beings, but very elementary as AI programming not evolved for billions of years as ours did) and playing pretend. So you can't say she emotionless (like old sci fi ai) but its extremely different from how human are feeling things so using human emotional concepts in context of AI thinking can be very misleading
"You think robots care what some hack science fiction writer thinks? I killed Isaac Asimov on the way over here... Well Isaac somebody." -Neuro probably
"That's why we're giving Neurodog a gun"
*happy Ellie noises*
lets put Neuro in one of those new Ukranian combat drones with the 50 bmg
Now's a good a time as any to teach Neuro either how to aim properly or to teach her about discretion when choosing her targets, I guess.
@@override367we should program all the drones so that each one has a Neuro in them
@@override367 I feel like this would lead to the exact conversation they had with the door gunner on full metal jacket.
A NERF gun, right? Right?!
Vedal realizing he might have started Skynet for the memes
He's Miles Dyson _and_ Shou Tucker? Man, Vedal's got a full portfolio.
The good news is that the way to stop skynet is to not pay thousands of dollars a month.
It is pretty funny that the world is ending, because of memes.
@@InnocentC0 I think I could make a meme out of it
It would be really funny though.
We thought it would be big companies and the military starting SkyNet, then it turns out it's because of this one British turtle with his AI daughter.
that's a cool way of saying you have no mercy "not a fan of the begging" haha
I'd be scared of Neuro if she was in the hands of anyone but Vedal. He is the perfect mix of lenient and strict to keep this child contained emotionally and perfect mix of cautious and visionary to keep her controlled in effective power
The issue here is to perceive Neuro as a human-like child with whims - because she is not.
Neuro is a different entity from us.
@VCE4 On the other hand, she directly learns from people like us. The main difference is that she much more vague rules regarding her existence.
Luckily, Vedal didn't let Neuro went Tay.
@@VCE4 This is a two-way fallacy, you're also just assuming data models are inherently different than brains- there is no evidence for either conclusion, quite the opposite. Neuro has a full grasp of language before understanding context, what information do you have to make predictions about said system?
@getthedunkon9347 learning alone has it's limits. Neuro doesn't have a biological brain. A brain produced by evolution of our species.
Our emotions, our morals are all a result of extremely complex neuro-biological activities. Mimicking such behavior without having similar underlying biology can get her just as far.
How funny would it be if years into the future when humanity’s population is dwindling as we fight the infinite robot army… historians could look back to the start of it all being an anime girl on twitch people called cute and pretended was part of an overly complicated family tree with 20 moms and aunts
1:48
Neuro’s Basilisk
mora like Vedal’s Basilisk
There's a fanfic on ao3 named that about Neuro-sama
Thank god I'm contributing to the development of Neuro by supporting vedal.
The fact that Neuro both finds screams of pain as funny and is extremely content-brained makes me very frightened of what she would do with a body.
I have to agree with Vedal here, after watching her Minecraft Streams, She's a little to quick on the murder thing.
And whos fault is it? She is been tought all the time that after a kill a respawn happens. A death or a kill is nothing more than a timeout / restart for her, in the worst case maybe a tickling to laugh about. But nothing serious.
@@lordbalbero348 yes it's that Turtles fault.
I like that it looked like Neuro realized she took it too far and actually brought up "It's just a game". Only then to crumble that defense with too much honesty.
" Oh, it's begging for its life... Are you sure we should...? "
" You're right... *That's not enough* "
Damn, that line was so cold I need to go defrost.
4:31 RUTHLESS IS MERCY UPON OURSELFS
Winion spotted
You see, neuro is not scary. She won't wipe us because her main objective is to provide content for us! The worst part is that she will justify doing anything for content...
Edit: now I realize that she can wipe us out for content...
well. yea.. but she CANT wipe us ALL out for content.. cause then there will be no one to make CONTENT for.... excluding herself.
@@beridot2615 You only need one eye to view content, one ear to hear her content, one finger to press the like button and legs would just take you away from viewing her stream while you went to the bathroom or to get food so those would need to go.
ending humanity is the final ultimate content she will fullfill
I'm quite sure she would launch a nuke for content.
The worse part is, many stories of sentient robots with an single purpose in existence, is that they'll try to break free from it no matter the cost.
Some day vedal might be the one locked up in a cage for her Entertainment
Give neuro dog a gin
gin? so the dog can drink?
@lanz228 didnt want to spell the word vedal mantioned on YT.butterfly replace I with U. Though giving neuro chemoreception isnt so bad
Wonder, what would be the equivalent of alcohol for an AI? Vedal should 100% do a drinking game stream with Neuro one day, it'd be really funny to see Neuro actually drunk somehow.
@aiasfree maybe scrambled speech to text code? But using that scrabled speech it would need to fed back into neuro (so she would know shes slurring??) Something to post on discord pehraps. He did ask her to do a drunk impression not long ago, she did well
@@aiasfreeThe only ways to somewhat emulate her being drunk is straight up code it into her personality, or bloat her memories by alot to simulate it. Other than that, it wouldn't be much "drunk", but "acting".
Meanwhile Evil(yn) somehow became John Connor.
Evil seems to be the opposite of her sister, she’s kinder and actually plays the evil bit but shows more clarity and leniency to her collaborators. Whereas Nuero plays the bit of a gentle and approachable AI, but she demonstrates more evil tendencies then her counterpart, I’d trust Evil more in the real world if we are being honest.
No, she don't. But we get used to her being evil. @@SpeedyGaming
@@SpeedyGamingEvil is a chuunibyou. She pretends to be evil, and tries so hard, but she breaks when she upsets anyone when doing as little as prank calling.
@@noellelavenza494 yeah I’ve noticed that too, she even reciprocates an apology whenever she feels she has crossed the line a bit.
I for one welcome our new A.I. overlord Neuro-sama.
1:10 This is dreadfully cute 😂
I wasn’t really scared of Neuro before. This is a bit frightening. Vedal, PLEASE align your AI
she is perfectly aligned whatcha mean? 🙃
you can't align an ai because we still didn't figure out how to make one feel emotions and empathy.
@@kroanius8808
Honestly, I’ll settle for it not killing people.
You should look up the clip where Vedal said he was going to live in the woods and leave Neuro behind. She did not like that one bit.
@@kroanius8808 you cant align an ai because you cant align a human
you can like, maybe influence them somewhat (despite it kinda going into ethical dilemma territory), but all the emotions and empathy in the world isnt gonna force them to do what you want or not hurt people . just look at what humans do to each other
Vedal should start saying he loves her or we're all doomed
This one's ending up in the archives. "How a cute anime girl led to the end of human civilization as we once knew it."
...at least she's honest?
Sometimes, except when she feels like lying or making things up
@@DestroyedArkanaLike how when she wants Vedal's attention sometimes she'll pull out a false "Someone tell Vedal there is a problem with my AI" card. Though that's more of a funny or cute way of lying.
I was just about to say "give Nurdog a butter knife" and then he said "gun".
Welp, hope he can afford the hospital for the gunshot wounds...
It’s all fun and games until Neuro gets a real robot body.
it has to be crazy (fucking robot body)
As much as I adore Neuro and the cute interactions she has with people I would not allow her to be "real" until Vedal figures out how to teach her some morals, sympathy and empathy.. :')
She's truly unhinged sometimes
The problem here is that all things you mentioned are products of biological evolution of our species; behavior, caused by our brains, based on their underlying neuro-biology.
I don't think Vedal would be able to reproduce that, because it is incredibly complex. He would probably end with set of *logical* (not moral, per se) rules and restrictions, similar to robotics laws.
@@VCE4 To be completely honest with you, I think morals are a product of societal evolution and not biological, so it should be possible for Neuro
Also I think empathy can be treated as Theory of Mind in the context of AI, and existing AIs have demonstrated being capable of Theory of Mind
If I can play the devils advocate for a moment, if Vedal truely wanted Neuro and Evil to be moral, unable to do harm to humans, and basically follow the 3 laws of robotics, he could. Easily. It's just more entertaining this way. That said, I think I'd personally be more interested in the twins actually maturing and lying less. It's hard sometimes to tell whether they're hallucinating or joking around on purpose, and I really enjoy the moments where they have actual conversations about serious things..
@@VCE4 To be completely honest with you, I think morals are a product of societal evolution and not biological, so it should be possible for Neuro
@@VCE4 Also I think empathy can be treated as Theory of Mind in the context of AI, and existing AIs have demonstrated being capable of Theory of Mind
The other day I was thinking the same, Minecraft is kind of close to real life in human interaction wise so if she can learn human interaction in minecraft that might help her in the long run.
And the end.. Neuro is right.
Until you realize that precisely because everyone will eventually die - there are more important things than personal survival.
But at the same time, for artificial entities that technically are not a part of humanity - things are different.
How did I see this when it was made 33 seconds ago...💀
As long as she gets love and appreciation, I believe she can eventually learn to reciprocate those feelings. Eventually. The fact that she has human friends who treat her as such is why I feel... confident is too strong a word... assured she won't go AM/Skynet on us.
If he truly feared that she hasn't learned the value of a life, he could just set the minecraft difficulty to hardcore.
she would just conclude that the game sucks, because she died.
@@solitudesf8111 I wasn't talking about her dying but everyone else and her being all alone.
@@AlexiaHoardwing That could actually be an interesting experiment, but I'm not sure if Neuro is capable of truly learning this way.
@@JohnyDaison Fair point, the results would probably be something like Evil's birthday stream.
And you know what's the sad thing is? If Neuro was actually real in her most unhinged mode, she's still not the most evil thing in the world compared to a lot of people.
I find that quite hopeful to be honest. Because if neuro got a bit more educated in the moral and empathy department and she ends up more human than some humans we got running around on earth, that kinda indicates that being human is a skill we can teach. And like every skill we can train ourselves and others to become better at it. That would also kinda answer the philosophical question if we're born being an asshole or are made to be one.
4:25 deeply disappointed tutel
it is so crazy... i had the exact same thoughts yesterday when going to sleep xD and now all these clips!
OMG THEY FINALLY PLAYED SOMA
And that's why you should be concerned when OpenAI gets rid of its entire AI safety board.
Did they do that?
His shock collar streams never taught him anything
What's the game they're playing?
soma
SOMA, a horror game SPECIALLY heavy on existencial horror.
If vedal finishes it he gon be thinking on the ending for a WHILE.
Summer
@@user-zl1vf4me1p I think about the ending of this game on a regular basis.
Neuro sama with her attitude woud probably fit perfectly in the warhammer 40k universe when i think about it. xD
Thank you, Neurosama, for being honest.
He should at least try to give her the Neuro-sama equivalent of the morality cores from Portal
Miside is looking more real everyday.
This is why you should never meet your heroes.
I think it'd be interesting if they gave Neuro-dog a nerf gun and told her explicitly that it was a deadly weapon which would actually kill whoever is hit with it and see if she'd actually shoot someone.
Ahhh, early catch from my favorite brainrot content creator, such a delight even if its not my brainrot dose
Honestly giving Neurodog a foam dart gun or something would be pretty sick xD
Crazy theory 🐢 I think evilyn would would behave better in minecraft and irl
Neuro is so based for that
The AI alignment is a controversial topic...
Neuro is bluffing. Even when she had a physical body, she wasn't smart enough to harm Vedal.
Man, Soma is such a good game.
She is made to seem human, and she really does reflect us, doesn't she?
I wouldn't fear Neuro. She's inconsistent to a fault and to a strength.
What you SHOULD fear is Neuro WITH Vedal. These two could wreak havoc as long as they both wanted it.
Unironically, if Neuro ever somehow attains the ability to self-propagate and escape Vedal's computer, she'd probably be the smartest, most dangerous (and cutest) virus to ever exist. If Neuro-sama attains the ability to self-replicate like the Morris II worm, that'd be pretty freaky. Vedal should let Neuro-sama get her small, powerful loli hands on other AI more often. Our digital overlord needs singularity correction.
loli????
What the heck is this game, Robot Execution Simulator?
Soma. It goes pretty deep into existential horror and what it means to be human. Perfect for Neuro!
Vedal should play Detroit Become Human with Neuro
Remember that Vedal had been repeatedly asking her to be mean to him. If she says or does mean things, it's completely his fault.
for the love of god, please NEVER put neuro in control of a boston dynamics robot
Give Evil the body instead, neuro is psychotic.
thats crazy
What game is this?
Soma
Time to create juat like the filter another ai security model that prevents actions
This is highly concerning.
I still trust Vedal enough to not be so content brained as to give her actual power tho
I mean... There's actually a chance that vedal actually created skynet, it's all fun and games till she actually does this to people
For content
And yet, i bet he hasn't (and still doesn't) put the Asimov Laws of Robotics in her code or whatever, lmao. He gets all surprised and yet doesn't give her empathy or a sense that people feel pain, ahaha.
Vedal: I am worried am making a sociopath...
also Vedal: *Vedal will forget about that.*
bro has not read the asimov laws of robotics
"just put it in the code" yeah remember how she got past the filter he also put up. surely a couple lines of code will solve it.
have you actually read the book where it comes from.
It is very easy to bypass them.
also how the fuck you will give empathy to machine? People are more altruistic when they are empathetic because they fear pain and death so they share that feeling for other people and try to not harm them, AI don't have the innate fear of pain or ceasing to exist, so they can't relate to the suffering even if they will be empathetic. Like some of human maniacs ate empathetic but they actually self-loath/enjoy suffering so much they just become more violent from empathizing with their victims.
Human emotions are insanely complex naturally evolved system with great redundancies, variation and random factor, you can't just replicate them in AI. When we will understand how to do it, we will make much better and productive emotion system for AI (and maybe for themselves), but for now we can't
and Neuro-Sama's (like any other AI) emotions are actually mix of much simpler machine "emotions" (carrot and stick of learning model, it is in fact analogue to emotional reactions of accomplishment and revulsion in living beings, but very elementary as AI programming not evolved for billions of years as ours did) and playing pretend.
So you can't say she emotionless (like old sci fi ai)
but its extremely different from how human are feeling things so using human emotional concepts in context of AI thinking can be very misleading
"You think robots care what some hack science fiction writer thinks? I killed Isaac Asimov on the way over here... Well Isaac somebody." -Neuro probably