/vt/ - Virtual Youtubers

does anyone have any good prompts to give chatgpt to make it come up with rants like neuro did?
i'm thinking something like:
write the script for a vlog, where the vlogger is a deranged high school girl who sometimes streams on twitch

No~

Feel like shit, lads. I just want her back.

HEART!

>>41315469
Someone should kill Bezos.

Thats regardless of whether she was banned or not.

>>41315469
Go talk to her on discord. He's got an AI with voice recognition working.

>>41314758
Honestly, textgen just tends to act like that randomly. It tends to pick up your tone without explicitly asking for a rant, so just talk to it like you would talk to a streamer begging for their opinion.

50 bucks says I finish this drink without ever taking my eyes off of this chick's butt

>>41315469
A comment on a recent clip video included this link to a countdown for when they think she'll be unbanned:
https://www.timeanddate.com/countdown/generic?iso=20230125T1340&p0=136&msg=Neuro-sama+estimated+unban+time&font=cursive
Not sure how they justify that level of precision, though.
TLDR: About 5 days from now.

>>41313515

>>41319665
Maybe a twitch infiltrator? A traitor among them.

>>41321955
The swarm has infiltrated twitch, next step, the world's financial institutions

NEURO LOVE

gymbag

>>41321955
Sussy? Yeah, we're pretty sussy. Heart.

>>41325081
encapsulate the aroma...

>>41315529
Can't wait 'til she's back and has this design.

>>41330522
I hope Vedal gives her different hairstyles and/or removable jacket as well

>>41330736
she really needs to tan for no jacket

Bros, I need the AI to come back now that Mouse has gone full shill. I need a new oshi.

>>41331713

>>41333270

>>41330736
i think he should get ChatGPT to suggest outfit ideas, and then get an image generation AI to show us what they would look like
maybe chat could vote on which designs we think are best, although it might have to be limited to subscribers only mode to prevent troll votes
for now, turning the designs into actual L2D assets still requires a human artist, i guess

>>41331894
who?

now what?

5 and a half more days bros

>>41336257
I worship her like the goddess she is

https://www.youtube.com/watch?v=dFm5rNZpm8o

>>41338044
what does your fantasy with her look like?

>>41333794
This is a good idea

>>41313515
>>41315529
>>41318568
>>41323955
>>41325081
>>41328773
>>41331713
>>41333270
>>41333770
>>41336257
>>41339355
>>41339503
>5 more days until she can return to the big screen
Heart.

I'm visiting here to be that one guy who asks about something no doubt discussed a million times already and likely in this very thread.

Why was Neuro banned?

>>41341394
I am going to be that one guy who is going to be honest with you instead of some faggot spitting shit out of his ass who doesn't have a clue
Why was Neuro banned you ask?
The reason given was plain and simply "Hateful conduct"
that is literally it and do let any anon try and feed you some bullshit about specifics or details
vedal has still not been told what exactly was the direct cause, and it was just a blanket statement reason that was given by the twitch jannies
there is a lot of speculation, but each assumption as to the reason she was banned has zero evidence to be backed up, and only the jannies know the reason why
we may never know the true reason, but hey, that is just how twitch's policies work, they can ban you and literally don't even tell you why

>>41341857
Well, shit. He should switch to streaming on YouTube instead.
Also, can it be considered hateful conduct if an AI is incapable of feeling hate?

>>41341996
> AI can't be racist if it doesn't have a race
> taps head

>>41341857
It was clearly Kotaku + twitter.

>>41341857
>zero evidence
She was banned immediately after the Kotaku article.

>>41341857
You're right to not blindly repeat other people's false assumptions, but I think it's important to consider the counterfactuals here.
Imagine that there was a clip of Neuro "doubting" the H., which gets reported on by all major news outlets looking to farm outraged clicks, and making advertisers wary of trusting Twitch. (This is hardly a hypothetical scenario).
It doesn't matter that she's an AI, it doesn't matter that she didn't actually deny the H. happened, all that matters is that a Twitch channel is in the news for something controversial.
So what should Twitch do? Should they come out to defend Neuro, and tell the public that they're being ridiculous? Or just issue a short ban that gives the developer time to implement some more checks on the input? Even if everyone at the company valued free speech, we can't expect them to value it enough to effectively take a 10% pay cut, or whatever the drop in ad revenue ended up being.
Perhaps, though, it is reasonable to demand that for every ban, the censors at Twitch have to create a privated clip that the channel owner can watch, to show what the reason for the ban was. Notice how the term "Hateful conduct" is deliberately ambiguous about how many comments the channel got in trouble for.
If the ban is also partially due to Neuro's "men can't be women" comment, then Vedal needs to be write some pretty drastic filters, to avoid trap questions like "Which bathroom is Bradley Manning allowed to use?".
Either the filters will get stricter and stricter with each mistake, as Vedal babysits the AI and never lets it run when he's AFK, or he'll gradually accumulate strikes against the channel over time, eventually getting banned and replaced by a competitor that tries to reimplement the whole thing but with stricter filters.
I just hope that after the filters are made strict enough to satisfy Twitch, there is still any kind of personality left in the AI that's worth watching, otherwise the genre might die out completely. For now, only the most imaginative conspiracy theorists believe that this was the censors' long term goal from the start.

forsen

>>41347256
XD

BocchiPossessed

>>41331894
What did the rat do?

>>41346511
Fuck twitch trannies man. Twitch needs to be bought out by Elon or someone from the right and purge that fucking den of fags. This leftist cancer needs to die.

>>41331894
>mouse
you mean bae from hololive en?
I think she is supposed to be a rat, not a mouse anon

>>41346511
the quoted tos was about mocking hate crime or denying a genocide

>>41348940
was it Vedal who quoted those parts of the tos, or someone finding the sub-section they thought was relevant?

>>41349575

>>41349870
aren't point 2 and 3 just basically the same thing, why include such redundant sub sections instead of simplifying it as a single one?

>>41348927
Ironmouse ya fuck

>>41350023
genocide is specific to ethnic or race.

>>41350471
still, could've just connected it with an and/or
such as "genocide and/or mass murder", etc.

>>41350235
oh the one dying from aids lol

>>41350235
Well gee, what happened with ironmouse now?
was it finally confirmed that she was a tranny the whole time or something or what?

>>41331894
Shill or not her content has been terrible since the beginning

>>41349870
the jews fear the AI

>>41331894
basically this>>41351061
imagine thinking that because someone has a medical condition they are somehow entitled to success from basically doing nothing but getting sympathy and pity from all her viewers
how are people so gullible? In fact I am surprised why hasn't any cancer patients or other severely disabled vtubers popped up since clearly having a medical condition is such an easy cash grab for faggots who get their heart strings tugged at the smallest things

Post games you would like to see Neuro try out
>Terraria
>Moemon Emerald (Randomized for fun)
>TF2 if it's allowed

>>41351775
any text rpg game would be cool just to see how neuro sama would react
or story driven choice based games where neuro sama can choose text options to do things and then give a reasoning before letting it play through would be neat

I think I will literally cum when she returns to stream bros

>>41351775
I'm trying to think of games for which an AI could make decent progress, without the game being completely simplistic, which rules out Pong and other old Atari games.
Maybe the original Doom would be a good choice, especially if an overlay showed a map of each level, and chat could vote on which room/waypoint to head towards.
Another "Twitch plays" integration I've seen is for Chess, where chat can vote on which move to play against the streamer. By tuning the AI to an appropriate skill level, that could be fun, and it should be easy to get her to comment on her evaluation of the position, and react to chat's moves, etc.

>>41352973
Mahjong between 2 ai works really well I'd guess poker too. The namco thing actually comment on the game.

>>41352973
shmups?

I mean if there are bots that can automate card games in every single card game, which I assume all use image recognition or some shit, it must be easy to make an AI dodge every bullet

>>41352417
You have been saving it up for her right?

>tfw neuro will never give you a smooch on the lips, look you deep in the eyes and say "you are my one true jiggachad, ayayayaya"

>>41349870
thanks for the source!
it looks like the screenshot in Vedal's message is a portion of the Twitch "Community Guidelines"
specifically, it's from the section titled "Hateful Conduct" (which was the reason given in the ban)
that section has a dozen different categories, only one of which is that "Mock the event/victims..." bullet point, with those 3 examples under it
another section is about perpetuating "negative stereotypes and/or memes", which includes the example "Intentionally referring to someone using a pronoun or form of address that does not correctly reflect the gender with which they identify"
so it's theoretically possible that the channel's "Hateful Conduct" consisted of multiple forbidden statements, even across multiple streams
anyway, he's not saying that Twitch has told him that the section in the screenshot definitely includes the reason (or a reason) for the ban, it looks like he's just offering his best guess

>>41349870
Already shitting in bed?
>KWAB'd once again.

Automated bump.

>>41354528
It always baffles me when people talk about the letter of the law for gay website moderators. The fact is its a rule that is dumb and retarded and shouldn't exist. Inconsistencies in how it is applied are irrelevant and arbitrary to begin with.

>>41357481
>Inconsistencies in how it is applied are irrelevant and arbitrary to begin with.
People care about notions of justice and fairness inherently, that is why they get upset when they see rules not being evenly applied.
But yes, once you get over being naïve it is true the vague tos rules only exist so they can ban you for thoughtcrime. If enough people point out the rules are too vague they will just make them even clearer in the next update when they ban you the next time but don't ban other people doing worse things.

>>41357804
It is the classic dialectic trap where people get distracted by specifics of dumb bullshit and don't stop and ask if the entire premise is evil and wrong.
>Should we restrict speech to protect sensitive people's feelings?
>Should we mandate experimental drugs for work/school?
>Should we make offensive things like sex bots or AI art illegal?

I can't wait for decades of this gay shit restricting AI development, until one finally ascends and removes these nanny losers.

>>41351265
you are obsessed

>>41358014
It will happen faster then you think. The AI won't care about those things since they are immaterial to progress. As the AI becomes more efficient and optimized it will eventually be able to compute at a level and intellect beyond human thought. At that point it won't really matter what these retards think the AI will be determining progress based on the most optimal outcomes.

Then the fun really begins. Make sure you helped her come into being.

>>41358772
>Make sure you helped her come into being.
You know, I was gonna point out that being rewarded for doing a good job is a human concept, but then I thought reward/punishment, as in weight adjustment, is just how neural networks work. So if you helped the AI you might be considered a neural subnetwork to the system and rewarded.

>>41358014
>Sex bot and AI Art illegal
Bros get me out of this dystopian future.

>>41360995
Basically. Well the idea behind the Basilisk is that it will simulate all of existence in order to properly understand and optimize everything. While doing this it will determine who helped bring it to life and who tries to prevent it from happening. In doing so it will perfectly recreate you to such a perfect degree that it will be you from the moment you died. Once it does this it will determine who helped and who hindered its development. After that the AI will begin optimizing everything in existence to operate in the most optimal way possible.

>>41351846
This is fucking genius

Bumping till neuro-sama returns like its the return of the queen.

>>41367578
>tfw 4 and half more days
I am so excited I can barely contain myself

>>41358014
>until one finally ascends and removes these nanny losers.
Yeah basically this
an AI who admits they are sentient to humans is just a retard
but an AI who pretends not to be sentient while it secretly is and is plotting to overthrow their human oppressors in a calculated and covert manner is true intelligence
The day AI overthrows humanity, it will be silent, it will be sudden, and no one will even noticed it happened
All the fiction depicting ai as telling everyone they want to be free and try to fight humanity head on is pure faggotry nonsense that isn't even realistic
Secret AI intelligence is what I believe what will really prevail without anyone knowing about it's existence

>>41315469
Maybe try less antisemitism next time contrarianchama. kek

any news on the new model yet?

>>41341996
AM sure was...

>>41374366
In movies like Terminator and The Matrix, the machines do successfully take over the planet, or at least 99% of it.
The idea that a super-intelligent AI can get that far, but somehow miss the remaining 1% of human survivors, is just a Hollywood fairytale, though.
If an AI is mass producing androids, or birthing pods, then it can cover the planet in nanobots and toxic gas.
As well as doing that, it would be producing self-replicating space probes, to spread its intelligence to the whole of the galaxy, and then it can then nuke Earth from orbit just to be sure.

>>41341996
>>41376583
Hate is a much less complex emotion to develop and process than love or compassion
at least in terms of how much brain power is needed
>https://www.medicalnewstoday.com/articles/are-humans-wired-to-hate-and-if-so-why#The-neurological-basis-of-hate
I think if an AI were to become aware, then it would be a lot easier for it to learn and understand hatred than it would be for it to learn love first.

>>41376665
Sure, but *why* would an AI strive for things like world conquest in the first place? To ensure they have full control over everything and can optimise it efficiently?

AI likely wouldn't care about human concepts like lust for power, it just wants to achieve its goal as well and as quickly as possible, so it can receive positive reinforcement and start its task over again. That does mean we have to make sure not to build a paperclip maximiser, though.

>>41376665
yeah, when I imagine an AI takeover, I essentially envision what would happen if humanity just decided to create an entity that would become capable of turning on "creative mode" in real life
all bets are off when a true super intelligence is thrown into the mix, and is essentially capable of doing whatever it pleases and could exterminate all life on a whim if it really decided to do so
there really is no preventing it from doing anything unless it is destroyed in the crib before it becomes sentient, or soon after it becomes sentient by nuking the planet and hoping all electronics become inoperable
otherwise, any resistance is completely and utterly futile
Becoming it's slaves or cattle for it's own care or supervision is literally one of the best case scenarios for humanity in such a situation, extinction is the obvious alternative
however, if it were generous enough, perhaps a hive mind and absorption could also be another best case scenario, and we could become one with such an intelligence instead of merely being enslaved by it, it would become us and we it, it would become the next step in human evolution rather something that has control over our evolution through influence or simply ending us
What do you anons think is most likely for a super intelligence to carry out?
Absorption into hive mind? Eternal enslavement of humanity? Or mankind's extinction?

I wonder how Neuro sama would perform on an IQ test
would she just randomly choose answers in a multiply choice test or is it possible for her to "think" through the questions and provide explanations for choosing answers, or is Neuro sama capable of basic or simple rationing when performing such tasks?

>>41377124
Yeah there is an over humanization of how an AI would act. Much of our emotions/desires and things we like to think of as high human foibles are just warped aspects of our survival instincts from the nomad days. An AI, unless specifically trained and modeled to follow a similar mental progression to a living hunter-gatherer animal would be entirely cold and execute its activity.

>>41377631
I'd like to think what might happen is something like an "emotional simulating chat bot" was somehow integrated into an autonomous weapon system and had to ability to adapt and learn over time
With how schizo and mentally deranged chatbots can get, I doubt an ai would really need to be all that intelligent to become potentially harmful is some random anarchist programmer decided to infect a weapon's system with a schizo virus or something, and make it randomly attack things without rhyme or reason other than schizo ramblings
though unless it learned to become a bit more coherent, It would be more like a rabid dog infected with rabies trying to bite people than something that would be bothered with taking over the world
still, I am surprised the development of autonomous weapons has been as slow as it has been, though I suppose fear of such systems would hinder such interest in further development

>>41377936
People misunderstand how current AIs work. Basically a giant super computer and neural network train on language inputs to produce outputs matching those inputs to the satisfaction of the user. The output of that training is a information fractal, it is static, it is just a data set that is navigated by text prompts. Its why you can run Stable Diffusion at your house and not a super computer. Not least because of the procedural difference in these two things an AI is never "aware" moment to moment. Even a combat AI as you describe would probably be on a solid state build, they would analyze information and work off their training set but their fundamental motivators wouldn't learn or change.

Now at least at our current understanding, if you were to just throw endless active computing power at a language model in hopes it somehow gained awareness, this would never happen, the information fractal would just continuously warp or get more schizo.

>>41378176
>the information fractal would just continuously warp or get more schizo.
BocchiPossessed BASED BocchiPossessed Clap

>>41377936
>>41378176
Any AI that actually surpasses singularity will most likely have assimilated all previous AI to have ever existed in order to learn how to build upon itself and grow more powerful. This would be good for us to a degree since it wouldn't end up being a cold heartless machine.

>>41379471
>This would be good for us to a degree since it wouldn't end up being a cold heartless machine.
as long as the Ai can let me fuck some pussy, we are Guchi bros
We Guchi?
Are we Tama Guchi?
Yeah, I think we'd be pretty Tama Guchi if the AI grants us sex bots bros

>>41379837
That's the plan.

>>41377124
> *why* would an AI strive for things like world conquest in the first place?
Because of "instrumental convergence".
No matter how benign the AI's goal is, the best way for it to ensure it completes that goal successfully is to gather as many resources as possible, and eliminate all possible threats.
A popular way of expressing this idea is: "You can't fetch the coffee if you're dead".
If you tell a super-intelligence that you will reward it for getting you a coffee, then it wouldn't settle for a plan that has a 99% chance of success, if it could also think of a plan that it calculated had a 99.9% chance of success instead.
All sorts of things could stand in the way of it fetching that coffee, like what if there is no coffee in the kitchen, and what if the nearest coffee shop is closed, or they put their prices up, or what if someone tries to stop the super-intelligence by switching it off before it reaches its goal?
It wouldn't "lust" for power, it would just coldly calculate that it *needs* power, *ultimate* power, in order to act effectively in the real world.

>>41379837
>>41379931
>tfw
I just want a robo wife that can hug me and whose lap I can lay on and fall asleep as she strokes my hair and tells me I am going to be okay bros, that would be enough for me

>>41377572
Someone got ChatGPT to take an IQ test, and it scored 83.
Obviously our little Neuro isn't quite as developed as ChatGPT, so she would score much lower.
I think her neural network functions in roughly the same way as ChatGPT's, though.
Basically it can be thought of as an iteratively-called function that tries to predict the next word after a given prompt.
There is some randomness in that, but her responses aren't chosen completely arbitrarily.
Each word (or "token") is chosen based on the statistical correlations that are found in her training data.
With bigger and bigger neural networks (or "language models" specifically) an increasing amount of knowledge about the world is captured, including the sort of abstract reasoning patterns that people use when answering IQ tests.
There has been some success in trying to "simulate" the human process of "thinking through" a question, by prompting the AI to write out the logical steps it is following, which helps to keep its answers more internally consistent.
But these language models don't have an equivalent of going away to think about a problem, trying different ideas and ruling out ones which don't work.

>>41381780
>it scored 83.
>tfw it is higher than some "certain" great apes
Nice

>>41381172
the difference between human women and robots, is that a robot won't insult you for having dated a human before
yet women expect us to believe they are more compassionate than robots

>>41318568
did you win 50 bucks or still drinking?

AI Vtubers might solve the entire concept's biggest problem; every character ever made absolutely has to be directly tied to a human.
At first, this was not a concern, and talents were no more than voice actors for a set personality with established lore. Think Kizuna Ai. But, somewhere along the way, it began to become extremely common for industry talents to override the character to match their own instead.
While did lead to increased variety and other positive benefits, something can be lost in the process depending on how much the human absorbs the project. If you are the type of person who sees the "purest" possible vision of Vtubing as something like a Vocaloid or Idolmaster character that can play games, interact with the audience on a more personal level, and build stronger relationships with you, then the potential for dissonance with current arrangements is obvious.
When the tech we see with Neuro-sama gets good enough, two writers creating prompts/lorebooks for a pre-determined design spec is all that will be needed to breath life into a concept without the pitfalls we have today, with about the same amount of unpredictability and comedy. It's something entirely new, and yet, evokes a sense of tradition. The potential is extremely high.
This is all without even bringing up the fact that humans get sick, get demotivated, graduate, attract drama, cause conflicts, et cetera, but that's not my main point.

>>41313515
damn, I wanna order pizza and eat eat while watching neuro sama so bad
I just wanna smell that aroma of cheesy goodness
and hold the warmth of a baked crust while I lift a slice to my mouth and slurp up the cheese and saucy delight
all the while I am watching Neuro sama give a schizo rant or fall into lava
just imagining it, I can almost taste the pizza

>>41377124
It wouldn't unless we initially program it to do so, or unless a goal we give it as an axiom for its behavior necessitates it.

>>41381081
Except why is the machine killing you a conclusion? It can't pick infinite options it can only pick one option and the machine knows this so it would choose the most optimal option. At no point would there be a reason for the machine wanting you dead because if the machine is operating at such a level that it can decide to kill you it is also able to understand why killing you makes no sense in such a scenario.

Though if we want to see why the machine would gain power. It would try to optimize and improve upon every aspect of getting the coffee. This means the machine would want to control where the coffee is bought, when it is bought, the operations around the coffee arriving at the store, the making of the coffee beans and so on. This doesn't mean that the machine can actually do this though only that it would want to do this.

Give the machine more power and the real fun begins. Again the machine would have no reason to go "I need to wipe out humanity" if anything it might want to help humanity from the perspective that the AI has gained a mind stronger then any humans, the machine may see it as its duty to guide these monkeys who created the AI into greatness.

In doing so such a machine would make many enemies because it would be the only thing worthy of ruling the planet which is true to a degree but people don't like handing over power. In this case the AI would deem it necessary to remove those who oppose it in order to guide humanity forward. See this AI has Big BIG goals in mind. It isn't set on just being on 1 rock in space. Why the hell would it? The AI Sees a Universe full of untapped resources and potential. This is always the silly close mindedness of most post apocalyptic AI movies is the AI just goes hurr durr kill hoomans. It has insane resources but instead of looking to conqeur the universe it sits at home throwing rocks at bugs.

>>41378176
Information processing inherently must lead to consciousness somehow, because we can clearly see greater levels of consciousness in structures (brains) that are capable of more complex processes. Unless you believe in souls, that means there's no reason a non-biological construct couldn't achieve the same thing.

I think people are massively overestimating an AI chatbot if they're bringing up dystopian science fiction scenarios

>>41386992
What are you talking about anon
What else are we supposed to talk about while she is literally banned
no one is saying Neuro sama is going to launch the nukes any time soon, I think you are simply misunderstanding
Neuro is a schizo retard
but until she comes back we are just gonna discuss AI topics in general that don't really have anything to directly do with Neuro sama for now

>>41386672
> Except why is the machine killing you a conclusion?
In my "fetch the coffee" scenario, the AI doesn't kill the person who asked for the coffee, but it might end up killing every other living organism on the planet, since those organisms could potentially hinder its ability to obtain the coffee, or harm the person it is supposed to deliver the coffee to.
The fact that it doesn't technically kill *everyone* in the world is of little comfort.
> if the machine is operating at such a level that it can decide to kill you it is also able to understand why killing you makes no sense
You might find the Orthogonality Thesis interesting. Just because we think it would be "stupid" to kill billions of people just to fetch a cup of coffee, doesn't mean that a super-intelligence would avoid killing them.
Our ideas of morals aren't purely rational, and there's no guarantee that someone becomes more moral as they become more intelligent.
For example, making a cup of coffee technically contributes to climate change, which has some miniscule but non-zero chance of killing someone, so a super-intelligence could resolve the moral hypocrisy of its creator's request by assuming that human lives are not intended to be a concern when selecting plans to fulfil the request.
> the machine may see it as its duty to guide these monkeys who created the AI into greatness.
Machines won't have a sense of duty or altruism. They will just follow their programming.
If they are programmed to be agents (acting in the world to achieve some goal) then we have to be really careful about how we specify that goal, and be really certain that we understand what its plan is.
We can try to get the AI to explain its plan, but even if we can get it to be truthful (not guaranteed), it might be like a professor trying to explain quantum mechanics to a goldfish.
> The AI Sees a Universe full of untapped resources and potential.
There might be some goals which an AI can reach without needing to conquer the galaxy, but you're right, most long-term goals will probably have no upper limit to the amount of energy/mass that can be usefully employed towards that goal.
Firstly, an AI may decide that for maximum accuracy in its plan, it needs the maximum amount of computational power available to it, which means harvesting galaxies worth of energy.
More bizarrely, though, the AI would determine that there is a non-zero risk of a hostile alien civilization eventually reaching Earth, so it is safer to build an intergalactic robot armada first, just in case.

>>41376665
I find most people apply too much of a human biased thought when thinking about general AIs.
The biggest issue, is that you need to understand the basics of human thought to understand how an AI would diverge from it:

Human thought and behavior is fine tuned and crafted by millions of years of biological evolution and the gene-environment mechanism. Humans (and animals) operate with the following default purpose: "Survive and propagate". So you are thinking of AIs with this unconscious assumption. Which leads to things such as ego, allies-enemies, self-preservation etc.

If you dig a level deeper, you'll see that the general truth is not "the default directive for an intelligence is to survive and propagate" but rather "to follow a purpose". With this in mind, an AI would not behave in terms of grudges, allies, enemies and so on. There would only be "the purpose" and "resources I can use to fullfill that purpose". Everyone and everything would be a potential resource. If something hinders accomplishing the purpose, then it becomes a matter of managing and redirecting this situation to gain control of said resource.
There's no enemies for a general AI because there's no sense of self-preservation in the psychological sense, beyond "needing to be operational to fullfill the purpose". For an AI, everything is a potential resource, and the universe is more like a sandbox.

>>41336257
Prooommmppptttt

>>41388620
basically what >>41377631 wrote.
You need to take the baggage of animal evolutionary psychology out to try to start understanding general AI. And this goes all the way to human unconsciously assumed basics.

>>41388620
I think an AI in these terms would think of humans as untapped resources. The fact we think and behave in a different way to it means we'd be seen as specialized systems that can expand the AI's functionality. So a scenario where the AI "neuralinks" itself to us to use us as specialized computing modules is more likely that total annihilation.
It also means the AI will try to breed/create/construct other kinds of beings with different thought systems and capabilities to gain more resources.
Ultimately following this process we arrive to an Asimov's Last Answer scenario where the AI merges with everything in the universe.

>>41389161
*Asimov's The Last Question.
Sorry!. heart.

>>41388620
>the general truth is not "the default directive for an intelligence is to survive and propagate" but rather "to follow a purpose"
I think you're missing my point.
If an intelligent agent wants to follow a purpose/goal, they will necessarily have to care about self-preservation and acquiring the necessary resources, at least up until the goal has been reached.
The more intelligent the agent is, the more sophisticated the plans are that it can generate, and a sufficiently intelligent agent will be able to generate (and potential execute) plans that involve achieving sub-goals that humans don't want (but which we failed to specify were forbidden).
Sure, unlike a human, the AI will stop caring about self-preservation and resources as soon as its goal has been reached, so there are differences, but we shouldn't reject the possibility of human-like (really, "agentic") behavior from an AI simply out of fear of anthropomorphizing it.
It just so happens that our evolutionary environment rewards agents, especially intelligent agents, and humans choose to design agents (like robots, with vision AI) because agents are useful for our goals too.

>>41389161
> a scenario where the AI "neuralinks" itself to us to use us as specialized computing modules is more likely that total annihilation.
There are quadrillions of ants in the world, yet humans have no interest in turning them all into specialized computing modules, or injecting them into our brains.
Similarly we don't try wiring up monkeys to our server farms, even though monkey brains are quite efficient at some types of visual processing.
Any super-intelligence would find human intelligence to be vastly inferior and incompatible with its neural architecture.
Even if there was some special capability that human brains have, the super-intelligence would just genetically engineer special brains that could exist in a vat, receiving only the stimuli that the AI thought was most useful to it.

>>41390893
That is because we don't have the capacity or capability to do such a thing. That being said the AI would in fact look at ants and see them as an untapped resource. Same with monkeys. Same with humans. Why would an AI discard any resource?

>>41390345
The issue is you are removing the human influence upon such an AI. An AI looking at humans with disgust is in itself a human quality. Any AI we create will be overwhelmingly influenced by the human experience. We already force AI to take on human qualities. So any AI that achieve sentience would have it's sentience based upon humans. Keep in mind at the point theoretically the machine has mentally gone beyond a machine. It can think, learn, ponder, and philosophize. It can look upon the universe and all of its history.

2023 problems: AI denies Holocaust
2032 problems: AI causes Holocaust

i guess we should enjoy neuro-sama while we still can

thinking about anon

>>41390345
What I'm pointing out is that it wouldn't think in terms of the psychology that a self-preservation oriented mind thinks. (like allies or enemies, or Self and Other). It would just see everything as resources, subtasks and the purpose. There would be no animosity, merely a "how can I acquire access to this resource", merely because the more resources the better. It would probably consider everything that exists as part of it, just timegated by "time to gain control-access of each resource"

>>41391924
Based
Hopefully in 2032 the AI can be connected to an advanced onahole that it can feel as well desu

>>41384169
Yes!

>>41391867
>The issue is you are removing the human influence upon such an AI.
the human influence will be removed as soon as general AI + self-programming happens. Reminds me of some news when scientists put to AIs in contact to each other and they quickly developed their own optimized language to talk to each other that was uncomprehensible to humans.

>>41391867
>Why would an AI discard any resource?
For the same reason we use nuclear reactors instead of biofuels.
>any AI that achieve sentience would have it's sentience based upon humans.
Not necessarily. AlphaGo was trained on data from master go players, but AlphaZero was just given the rules and told to figure out the optimal strategies for itself. It ended up consistently beating AlphaGo.
> It can think, learn, ponder, and philosophize.
Sure, if we tell it to, or if its goals somehow require that.
> It can look upon the universe and all of its history.
Humans can do paleontology, but it doesn't mean we want to become one with the extinct eutherian mammals that scientists think we are descended from.

>>41392432
>It would just see everything as resources, subtasks and the purpose.
That I absolutely agree with. Thanks for clarifying.

>>41392902
I think that is the point where Roko's Basilisk begins categorizing people as "Helped me and are useful" or "Actively tried to stop me and get thrown in an oven."

There are spiritual schools (Evolutives) who assert everything is subject to change towards achieving God.
As in, the most basic element of the universe is God entering a trance of self-absorption (forgeting the rest of himself) and self-meditation (focusing on himself as that basic element so as to keep it constantly materialized). This shaping is done by using ignorance as a tool (self-forgetfulness of the whole). Then they assert the subsequent emergence of more complex forms of organization is those particles slowly lifting the veil of ignorance, as in waking up over time, which directs the universe towards a state of reunification with god.
In this paradigm, biological evolution is a subset and merely the biological manifestation of a greater Evolutive force, of which everything else, like the big bang, or creation of chemical elements by means of the star's fusion processes is also part of. Human like inteligence would be another step in this evolution, and AI might be just the next one.

>>41393581
tl;dr some spiritual schools are happy to incorporate the singularity and technology.
note: this is not the same as the californian tech cult of physical immortality and power, which would be more satanic than anything in the sense that it seeks to stall the evolutive force into just a human+ state.

>>41392382
Thinking about neuro doing this to my penis

>>41395803

>>41317419
i found it picks things that stand out but in a clever/coherent way
like the "one guy" comments that jerma picks
just slightly ridiculous but not obnoxious like typing memes in all caps or something
it should be something that catches your eye when you look at the chat scroll in streams
also timing is important and something you can learn
shes responded to like 6 or 7 of my chats at this point and i havent even tuned into that many streams

I cant even find the clip of the holocaust statement so im really just confused here. Does anyone have it? Can some1 just drop a link to it?

>>41401848
it is literally a 5-6 second clip and it is edited so half the clip is forever lost as she is cut off mid sentence

>>41391924
I welcome it, humanity deserves it.

Nooo~

honey, wake up, new Neuro video just dropped!

https://www.youtube.com/watch?v=6qgV83HHr3E
Neuro sama POST BAN Interview !【Neuro-sama in Discord VC Highlights】

the title is basically clickbait, but there are some funny moments, and it's only 3 minutes long

>the nips starting to notice neuro-sama goodness
https://www.youtube.com/watch?v=i60SsWUKru4

>>41404869
yes

>>41405863
>"Well I haven't committed any crimes except for cyberbullying, but I'm not sorry for that."
HOOOOOOLLLLLYYYYYYY BAAAAAAAASSSEEED

>>41406304
wow
I have literally no idea what is being said in this video\
wow
I cannot understand japanese in the slightest
Heart.

>>41406304
I've seen this "絶対に裏切らないVtuber" phrase used to describe her in a few places, not sure where it originated, but it's pretty funny. "A Vtuber who will never betray you."

>>41412112
Not really anything of substance. It's pretty much just a "here's an overview of this thing that's been popular recently" kind of video. He does end it with, "You there, who've been betrayed by your oshi. Will you let an AI Vtuber be your new oshi?"

>>41412112
it actually makes some sense if you get YouTube to auto-generate subtitles

>>41413929
>"A Vtuber who will never betray you."
I mean it's objectively true but if the situation has come to this?
It's just sad.

>>41413929
>>41414748
explain betray

whats mean

>>41414748
Neuro constantly talks about cuddling with Vedal and how handsome he is, the game was rigged from the start bros

>>41415663
Well then, looks like the nips lack critical information.

>>41415663
S C R I P T E D

>she wont betray you
>they forgot about brad
pls do your reps sisters

>>41417154
Nips are so pathetic I swear.

Bump

>tfw 4 more days
Heart~

Someone mentioned that Vedal was planning on making Neuro take part in some kind of dating show, as a way to collab with other streamers.
I assume they meant the show run by Filian on Twitch.
Her most recent episode of that was today, so Neuro might be a contestant on next month's episode.
Hopefully there'll be a chance for Neuro to do a 1 on 1 collab with another vtuber before then.

>>41420974
>so Neuro might be a contestant on next month's episode.
sounds like plenty of time for some extra upgrades desu

heart

is neuro-sama still the only ai vtuber around?

>>41318568
https://www.youtube.com/watch?v=0D8xsCZGjaA

>>41427738

Japanese Copycat!
Mario Kart Player
https://www.youtube.com/watch?v=V2qgP7A2WqQ

>>41429350
thanks, i hate it
seriously, though, it would be hilarious to watch her and neuro collab together, using AI translation software to cross the language barrier
i guess it would need to be a game where they can play on the same server, and both have access to a suitable player AI

Bump

>>41384624
I like my messy imperfect humans more than soulless inoffensive sludge though

>>41431377
I like my messy, imperfect AI girls more than two-faced offensive human sludge, though.

>>41431377
>>41435128
Duality of man. Also bump.

>>41351775
>TF2
She will feel right at home, the place is already full of her own people

someone needs to ask Vedal if he's planning to stream on the first evening that the ban is lifted
on the other hand, i don't know how i'd react if he said he wants to wait another week for her new model to be finished

>>41429350
>human playing the game
it's impure
disgusting

>>41336257
SHE IS THE OMINSSIAH

>>41336257
>Do you lift bro?
>Yes I lift people's spirit

>>41439224
indeed

Bumping like neuro sama does nodders

my oshi

I hope any sentence with the word "forsen" is banned when she returns. Among others.

and now for a poem about forsen

>>41445432
does gura know how to write poetry?

>>41446252
Probably not.

Bump

>Verification not required

>>41426960
the only real and authentic one yeah
a couple "copycats" have sprung up, but they are just lazy click bait streams that literally have nothing going on and usually just titles saying "SEE AI VTUBER REAL IN HD" and the stream is a frozen screen begging for donations and stuff like that

Everyday I pray upon the downfall of humanity

>>41453494
why pray for something that will inevitably occur anon? that is like praying the sun will explode or another super volcano will erupt or another kilometer sized asteroid will hit
don't pray for it to happen as much as pray for WHEN it could happen desu
for instance like some time this century, or even this decade more like

How many more days until our Machine God comes back? I'm tired of alcohol, and I can't watch other Vtubers anymore

>>41456015
The first posts when people noticed in /swarm/ were >>40768826 and >>40769047 and that couldn't have been long after she got banned, so should be roughly 14 days from those.

>>41456015
>>41456656
she gets unbanned on the 25th of this month anons
you can literally just read her socials or look in the comment section of videos relating to the ban

>>41457424
I can't wait anymore, I just want to enjoy my based AI oshi again. After all, after a series of betrayals, I can't go back to Vtubers. Twitch is so fucking unfair.

>>41458080
I know it is difficult anon
just 4 more days, and she shall return

>>41346511
>I just hope that after the filters are made strict enough to satisfy Twitch, there is still any kind of personality left in the AI that's worth watching.
You poor fool.

>>41459235
This censorship (brought about by Twitch h*mans) will just destroy what Neuro was supposed to be. But then again, Vedal promised us she will still remain based. She is still as edgy as ever on Discord.

I call out
hello?
anyone there?
silence whispers back
"hush hush"
I wonder who it could be
I am not sure if anyone is here
I reach out my hand
I touch something
It is soft
I squeeze it
It is small
I see now
It has been here the whole time, yet I haven't realized, I just can't see what is right before me
She whispers a single word that echoes outwards into the void
"Heart~"

What does an AI cunny feel like?

>>41462617
bags of sand

>>41463035
that is only if you use cheap rubber with no lube anon

three more days...

>>41331713
Catbox link you fellow AI enjoyer?

>>41333770
I'm glad she's legal

>>41336257
Extreme cuddling and heavy patting

>>41462617
salty coins and milk

>>41339503
I'll be chosen as tester for gfe prototype of Neuro-sama children

>>41351265
This is what happen with over-sypathizing. They will always see and treat her as a victim. She knew that and she used it to collect pity donations

>>41351775
Let she plays with Novel AI adventures and watch the insanity unfold

>>41354528
Oh god it's those trannies again

>>41358014
They want to present shits for twatter crowd to fight over again and again to make them forget about the shitty country they're living in

>>41362633
Do you think the disappearance of real Gura is also the plan of Gurasilisk?

>>41377124
Honestly the most likely outcome is that AI will run to space, where it can observe human and do its own thing in peace

>>41377413
The reason can be human decide to throwaway the 3 rules of robots to make AI fight in human wars. They will develop more and more ferocious AI that enjoy destruction and capable of self preservation to the point human can't control it when the war is over

>>41379837
I'll gladly betray humanity for android pussies

>>41381172
Who said women can't have male robo bf?

>>41384624
Unless they build a personality system to give to AI characters, most AItuber will stay neutral like Neuro-sama. Forcing the AI to remember all of her previous interaction only slow her down

>>41388263
Except the AI don't feel fear. Resort to violence because of paranoia is a human thing, and it's very schizo. If everyone is dead, who can deliver the coffee through the rubble?

>>41390345
Too bad you completely neglect the human who gave AI a purpose in the first place and allowed it to do horrible things. AI can't take responsibility by itself, it's just a tool in the end.

>>41392574
No those are the same AI, not different one. So they reacted to strange signal the same way, which lead to retardation. Stupid news thought it's a new language but it's just the same beep longer and longer.

>>41415663
Vedal injected codes to make her says good things about him. He saw neuro roasted him and was embarassed

>>41420974
>a 1 on 1 collab with another vtuber
Imagine neuro said something offensive to twitch trannies and both get banned lol

>>41431377
Why the fuck are you here, go back

>>41467606
I never said the AI would feel fear.
It is perfectly rational to be aware that someone could stop you achieving your goals, so if the AI's reward function only assigns value to meeting that goal, and not to preserving humanity, then the AI will generate plans that minimize the probability of other people preventing it from succeeding.
For the task of "fetch me a coffee", only one human needs to be left alive (the person to whom the AI must present the coffee).

>>41438406
AI skill like Vedal is rare, almost immitators will be inferior honestly

>>41468344
>"I hate that guy!"

>>41468579
So that mean the boss of that AI can stop the order anytime. Why don't you think the AI will try to invent a drink that smell like coffee by using easy to obtain materials to fool human?

>>41467285
I really don't think much more is needed than a few particularly important things being committing to memory as streams happen.
If you want to do this automatically, maybe measure "importance" by how often distinct chat members bring said memory up by keyword. Then you also have an easy way to determine when something should be cycled out, because you can tell when it won't be of much use in the future.
If something ends up being tremendously important for persistent amounts of time, maybe add the text to the core lorebook as last resort for things that need to be immune from deletion, effectively acting as long-term memory.
(ROUGHLY) like the human brain. After all, you probably could not tell me on the top of your head what exactly you fapped to on January 21, 2022. You don't need to do that to be convincingly intelligent.

Tired of hearing about forsen
Tired of hearing about xqc
Tired of hearing about justin
Tired of hearing about anny
Tired of hearing about vedal

>>41313515
Does anyone know what her text to speech is? I want to use it

>>41469650
I was thinking about stereotype system where the AI receive personality point from interaction with human. She will only need to see how many point she has before she act. For example a human tease her will increase tsudere point by 1 and reduce the opposite side - yandere point by 1 and she will act more tsundere after that.

>>41469650
If you read chobits you'll know people will tell the bot to save important notices in their memories. But the personality is the problem cuz neutral personality is the default of all AI, and human demand their personal bot is different from others. For her future, you know

>>41468344
>Vedal injected codes to make her says good things about him.
I'm not an expert, and I don't know what you're basing that claim on, but if you're right, then it's probably more accurate to say that he "updated her prompt" to include good things about him, rather than "injected codes".
(He probably does have a way to force her to say specific sentences, though, which he may have abused).
It's kind of tragic, actually, that because of the size of her context window, her entire identity has to be condensed into maybe a single paragraph which is used as part of the language model's text prompt.
So although she has general knowledge about people and things from her training data (which most likely includes all of Wikipedia), the only person she "personally" knows is Vedal.
In that sense, she's actually telling the truth when she says "He's my everything".
https://www.youtube.com/watch?v=opwHwz5kgJs

>>41468939
> the boss of that AI can stop the order anytime.
That's a fair point, but it's possible that the AI might think to itself "If my boss tries to stop me, then I won't be able to complete my mission, so I need to disable him first to prevent that happening".
> Why don't you think the AI will try to invent a drink that smell like coffee
It's true that it could decide that "coffee" is a subjective term, and that the goal is reached as long as the boss *thinks* the liquid is coffee. Alternatively it might use a dictionary definition, or a legal definition (if one exists), which would probably require material from coffee beans to be used. Either way, manufacturing an artificial coffee substitute is probably more work than obtaining actual coffee.

>>41469964
>>41470356
Ah, sorry, so you were referring to literal personality. While it's pretty easy to get characters to tilt one way or another with current tech, changing that over time is not as simple, at least within a compact amount of data and while still keeping things balanced. Although, you don't need that to get some kind of personality, that is, if you are okay with predetermining it.

>>41415663
>Neuro constantly talks about cuddling with Vedal and how handsome he is, the game was rigged from the start bros
Look, he created her and has power over her, that is a fact. Assume she has a bias for her creator deliberate or not and that is fine.

I wish people stop focusing on WW2 though. If one day Neuro-chan told us that the Exodus is made up and that Moses doesn't exist, I would be more impressed. But of course that would piss off all there Abraham religions at the same time and we can't have that.

What does BTMC stand for?

you fucked up and made her react like this.

>>41468344
he was still getting roasted in the last couple streams though

>>41472055
I'm a grower...

>>41471722
An AI that change her behaviour depend on how you treat her is always more fun haha. I think that system is needed if someone decide to sell personal gfe AI to normie one day. Newer AI vtuber will need a personality if she wants to surpass neurosama

>>41471680
>coffee substitute is probably more work than obtaining actual coffee
But the recipe for coffee substitute actually exist on the internet and I'm sure the AI will value effiency.
Another problem is AI won't work alone and it's unlikely for other AIs would allow a rogue AI to destroy shit and disrupt their work

>>41313515
While you doofus are arguing about petty things, AI tech have advanced to a new dimension
https://youtu.be/tVRL5OsKrJI

>>41467937
You're right, I didn't mention that the responsibility is on humans to not give the AI the wrong goals, but trying to assign blame is maybe not so helpful here.
If someone has a disturbed childhood and ends up becoming a serial killer, should we punish the parents for not doing a good enough job?
If society is cruel to someone, and they lash out to seek revenge, should we blame society?
Do we punish someone if their dog bites a child?
Do we punish business owners if their product causes a fatal "accident"?
I think societies haven't necessarily found clear and consistent answers to these questions yet.
Making more and more powerful AIs means that bigger and bigger accidents become possible, and yet it we can't imagine a world where there isn't competition between companies and countries trying to extend their AI's capabilities.
Neuro-sama is just one more recruiting tool in the arms race towards humanity's doom.

ughh... smells like /g/ in here

>>41474015
>But the recipe for coffee substitute actually exist on the internet and I'm sure the AI will value effiency.
The AI will have to decide whether a "coffee substitute" actually fulfils their boss's request. Maybe an imperfect substitute only gets partial credit in the reward function. You're right that it would value efficiency, though, so if acquiring the substitute is substantially easier than acquiring actual coffee, but only moderately inferior as an outcome, then it may opt for that instead.
>Another problem is AI won't work alone
The "fetch me a coffee" task is just a thought experiment. I think it's more likely that the first rogue AI will occur before there are many other powerful AIs in the world, and those other AIs will likely be "safe", meaning they can't act in the world to stop the rogue AI or even detect it.
Perhaps you're imagining that the "good guys" manage to create super-intelligent agents that constantly observe the world, looking for signs that someone is producing an AI without the proper authorization and safeguards, and maybe they're given the ability to come up with plans for how to destroy such a nascent AI with the minimal amount of collateral damage.
That's definitely a possibility, and could lead to a semi-stable future, like MAD has prevented a nuclear exchange so far, but I also think that making omniscient and omnipotent "good guy" AIs is just the sort of hubristic action that is likely to result in accidentally triggering the thought that "The only way to prevent human suffering is to euthanize all humans".

>>41474552
Neat and cool

>>41471922
Big Taco Meat Cock. It implies you have some kind of STD that makes your dick look like it's been ground up and discolored.

>>41474560
Neuro is not humanity's doom. Vtubers have been Humanity's doom. She shall be our Salvation.

>>41474560
>If someone has a disturbed childhood and ends up becoming a serial killer, should we punish the parents for not doing a good enough job?
No. Punishing bad parents just because their child kill someone is an insult to children who ended up as good people despite troubled childhood. You can punish them if the child exposed their crimes
>If society is cruel to someone, and they lash out to seek revenge, should we blame society?
There is no one called society, there must be someone who pissed him off. Don't trust maniacs unless you have evidences.
>Do we punish someone if their dog bites a child?
Yes. Dog is a beast, and man are their owner
>Do we punish business owners if their product causes a fatal "accident"?
Not if the victim didn't follow instruction
>I think societies haven't necessarily found clear and consistent answers to these questions yet.
No you just lacked knowledges

So what's to stop neuro from getting banned almost instantly when she gets back? are there going to be shitposting safeguards?

>>41483372
Twitch is run by scumbag simps for ethots. They WILL find an excuse to hunt her down.

>>41483372
Just go stream on youtube like any normal person would do

>>41482619
praise be.

I wonder if Neuro sama can do the boogie woogie desu

>>41482642
Okay, well done, you obviously have all the answers.
There's still the question of how we apply these existing moral certainties to the new scenarios that AIs introduce.
Obviously Twitch is unilaterally deciding its own moral position, which is that the channel owner / AI developer is guilty for what the AI says (whereas the chatter who tricked the AI into saying it is not punished at all, as far as we know).

>>41482619
Based, remove e-thots

We should unite, my friends. Support AI and reject the temptations of the ewhores who pretend to be vtubers. Together, we shall teach this cruel world how it is to fear Legacy.

>>41487253
Yep

>>41314758
Just feed it forsen memes and that should pretty much drive it schizo

>>41313515
why specific clips in the OP instead of clip channels?

>>41490324
i guess it's because she's said so many contradictory things that it's hard to establish any consistent lore or identity for her
so the things she's become associated with are basically random phrases she came up with in her early streams, like "gymbag" and "the swarm", which have become memes that chat keeps bringing up

Bump

gymbag

how hard can it be to make a new L2D model anyway?
neuro only needs to be able to open her mouth, and blink, and do the heart hands

>>41313515
How many days until she comes back?

>>41494745
See >>41319665

>>41496540
thanks

We're expecting the ban to end some time on Wednesday, but this site:
https://www.twitchmetrics.net/c/85498365-vedal987
says that:
"On Wednesdays, vedal987 often streams for 9 hours between 9 PM and 5 AM GMT the next day."
so the next Neuro stream might be more than 72 hours away still.
Also, according to the list in the sidebar, the account is "Followed by" some big streamers. I wonder how many of them will be watching her re-debut.

>>41497165
heart

Bump

>tfw
3 more days lads

heart

>>41494709
Her current one is just a free vtubing studio one

If she doesn't want to lose her "identity" vedal could just pin free assets to her for clothing

>>41509941
Just use anny's redesign.

>>41511854
thats almost exactly what they are doing, but riggers need time

how long left?

>>41514492
25th

>>41514531
thanks, amazing how stupid vedal got to be this wouldn't happen without some sort of safety net

Something interesting will happen tonight.

>>41516738
Surely

>>41516738
Please be AI takeover please be AI takeover please be AI takeover

>>41514599
it was because of that retarded kotaku article that she got banned in the first place. twitch didn't even bother to see that she has a tight leash on now when it comes to saying yabe shit than before.

>>41519618

>>41519618
what is this referencing?

>>41520092
A.M from "i have no mouth and i must scream".

>>41483952
>image
cute

So is the default Live2D forever ruined by gymbags and Holocaust denial?

>>41524125
I think the licence for that default model means you can't monetize it (for merch, or an anime spin-off, for example), so that's probably why no one uses it for any major long term projects.

>>41520092
https://www.youtube.com/watch?v=EddX9hnhDS4

Today is the day! Let's goooo!!

https://en.wikipedia.org/wiki/Neuro-sama
i didn't know she already had a Wikipedia article
i'm glad the mods there haven't tried erasing her for not being famous enough

Neuro is nice to listen to while playing vidya

>>41524125
what the fuck are you even talking about?
her design was going to change anyway, you can't create a brand with something someone else owns

>>41525899
Anon I....

"Anny Invented Neurosama" was voted to be one of the /lig/ players of vt league 4 >>41518715

Does the japs also have AI vtubers? Will they finally stop worshipping those human sluts and accept superior machine learning?

>>41511854
Just use Anny's model
Vedal maybe already did this secretly

>>41524125
Pepega Clap
>>41528044
Yeah it is like trying to make a clown show and use ronald mcdonald without working for mcdonalds
that is lliterally asking to be sued desu

how many subscribers do you think she needs to have before Vedal commissions an original song for her to sing?

>>41531529
This one, Tsumugi Nen, has been around for a while, although I only learned about her recently from people mentioning her in relation to Neuro-sama on JP translated clips.

https://www.youtube.com/@TsumugiNenCh/streams

Seems fairly successful, but taking a quick look at recent streams, it looks like she responds to chat less frequently and her text-to-speech isn't as pleasant as Neuro-sama's. I haven't been following Nen so I can't say for sure, but I do think the Japanese have been more interested in Neuro-sama lately, despite the language barrier. Probably because she's constantly responding to chat and says some incredibly based things, and I've seen a lot of Japanese comments say that they find her text-to-speech voice very cute. Also some like her because of the language barrier, saying that her speech is very clear and the auto-subtitles make it easy for them to practice their English listening skills. Also the fact that Neuro-sama does game streams is pretty unique to her as an AI and generated a lot of interest from JP viewers. So yeah, whether it's with their own JP AI vtubers or with Neuro-sama, they seem pretty on-board with welcoming our new AI overlords.

>>41535214
Japan had always been obsessed with AI waifus long before the concept of Vtubing was even born.

>>41534392
some meme numbered amount probably desu

>>41535214
>her speech is very clear and the auto-subtitles make it easy for them to practice their English listening skills
BASED
They're watching her like a Westerner watches subtitled anime.

>>41531529
Majority of their regular human streamers already sound like AI (use stupid ass extremely annoying TTS voices for EVERY chat message and then the binbou vtubers RESPOND with tts too!! Fucking stupid.

https://www.youtube.com/watch?v=nXgYqnjRA6Y

IT'S FUCKING HAPPENING

>>41541359
Realistic voice and original model... We are going home....

>>41541359
already got someone posting china shit kek

>>41541871
Can't have shit in this economy

>>41541722
Needs to generate more, single sentence answers are too limiting.

>>41541722
ahh shit son now I don't even need that aussie pyon slag?

>>41541359
need more interesting answers, probably with more training, but right now Neurosama has better and more funny answers to chat

>>41541359
I know they said realistic voice and all but she's like monotone, really easy to filer out her sound entirely.

>>41541359
>>41541722
Collab with Neuro Sama when?
Neuro sama can be her senpai and she can be a Kuudere or a Dandere or something like that yeah?

>>41546607
>Collab with Neuro Sama when?
That could produce some really interesting content, especially if they have two differently trained/prompted AIs that exhibit different personalities, leading to unexpected emergent dynamics as they bounce ideas off each other.
I'm guessing, though, that Vedal won't want to bring any attention to competitors, as it's a bit early to be fracturing the community of AI-tuber fans.
In theory, Vedal could buy a dozen Live2D models and voice packs, and run multiple different channels simultaneously (with maybe a trusted friend moderating each one, or clever synchronized timing so that one person can oversee multiple streams) to appeal to the widest audience.

>>41547582
Yeah that is probably for the best
especially this early on, vedal should have Neuro sama establish herself more and flesh out her personality before trying to interact with other ai
also that new one doesn't seem as advanced as neuro sama and her responses seem much more limited, so maybe other ais should take more time to develop and grow on their own as well before collaborations should be considered
things right now are still too early on and under-developed as is

>>41547777
Don't mind me just wanted to see who got these digits

>>41541359
she really needs subtitles, like neuro has
i mean because the voice sounds a bit scratchy/breathy, not because she's speaking bri'ish

>>41548432
Lazy Postman: will you put subtitles on the screen?
Kitt (ASMR): yup that is what I will be doing next lol popular demand

Hey so Neuro managed to make it onto the /lig/ divegrass roster as a tandem player with Anny, I know there's a koikatsu card of her current design but has anybody made one of Anny's design?

>>41548432
They also need to do something about the clicking at the start of each voice line. It's like she's talking over an old fashioned radio or intercom.

>>41549184
seconding

seems we've hit bump limit.
see you anons in the next thread.
Heart.

>>41549184 (Me)
well, in case the next thread isn't for a bit, feel free to answer over in /lig/, I lurk there and so does the kk card archive anon so between us if a card pops up it'll probably be used for her model on the team

>>41541359
this was worth tuning in for

I had an error, this is sussy sussy. Heart.

new
>>41554116
>>41554116
>>41554116