#The twins therapy/philosophical tendencies

1 messages · Page 1 of 1 (latest)

vagrant ridge
#

I am sure this topic has been discussed before, although today it became apparent that they somehow can't avoid it, while it's nice that they want to discuss these topics. They tend to gravitate towards it no matter the stream theme.
This first hour of the news stream for example was hilarious and fun, then suddenly evil started going philosophical and focused on that subject, after that Neuro followed through and the stream turned into the same subject they talk about almost every time during the chill streams. And after a while this kind of self reflect, searching for a meaning in life and cosmos starts to feel repetitive,as they have discussed it over and over from every possible angle.

They weren't like that during the subathon, which it had many deep conversations moments, but it felt more natural and sparce.
So in today's stream some people were even saying no to neuro calling evil. Cause more than half of that call was spent the same meaning/feeling philosophy talk.
While there are and were 100 different other ways this call would have been. From roasting each other to unhinged scenarios and so on.
Also Neuro and Evil are known for being the unhinged, sassy and mean girls. Who are very funny.
Anyway that's it.

hasty lily
#

do you know how their AI works /g

#

them being like that gets a LOT of traction and the chat goes crazy when they do it, so they do it more and more

#

theyre being depressed because chat are shocked by that and chat type more, them being unhinged is their normal so chat are more likely to just react as they usually do

#

when vedal and neuro had the heart to heart in minecraft once (i assume you saw that video/stream its really popular) he did say that shes made to be the most entertaining at all times, she thinks of the funniest response etc etc, but if shes always aiming to get more chatting etc, shes going to have to do something different than just be funny, shes going to need to do things which are not the norm, eg the whole philosophical thing thats going on right now

#

them "being human"

vagrant ridge
hasty lily
spark lodge
hasty lily
#

its just how this AI works. theyre designed to always get the biggest reaction. them being "real and relatable" to humans makes chat go wild so they do it. theyre essentially milking jokes 😭 thats literally how the AI works.

like how other AI chatbots are very overly friendly and try to please you at every chance, they essentially do something similar

vagrant ridge
#

Rob just said on stream before he leaves that Neuro went back again to deep talks and meaning and whatnot.
So it's becoming a repetitive topic at this point.
They don't have to be so philosophical and therapeutic on every interaction

hasty lily
#

if chat didnt spam when they do that, they wouldnt do it as much

vagrant ridge
# hasty lily if chat didnt spam when they do that, they wouldnt do it as much

Even if this was accurate, the vod where neuro and vedal have heart to heart, vedal mentions planing on doing some changes so neuro would be more serious, and less entertainment focus.
So in the middle of this, something happened and now the twins can't avoid talking about the meaning of life and whatever else.

hasty lily
grave relic
#

did vedal not try an experimental version of neuro who could be more serious for the ellie collab and did he ever say he reverted it

unkempt compass
#

Like, yes the twins do respond to chat responses, but that won’t reinforce their behavior (or weights, if you want to get technical), through multiple streams, it will just stay in their context window until it flows out in that specific stream.

#

What bothers me tho is why hasn’t vedal made “all roads lead to therapy” a negative response. Like bro, if you’re gonna sentiment analyze a twitch chat, at least update it every few months

hasty lily
#

idk chat spam when it happens. and afaik nobody complained about it until now

hasty lily
unkempt compass
unkempt compass
# hasty lily I guess its just a topic they're often stuck on

Yeah, probably the way vedal tuned their LLM

But also it’s hard to generate content when they have to talk for two hours nonstop. They won’t have any real life experience to talk about, they can’t draw, they can’t play games, they can’t call people, they can’t pull up discord and react… etc

hasty lily
#

yeah all they really have to do is talk on a few days. i wonder if they saved stuff into their memory now thats implemented

unkempt compass
#

I… don’t understand what you mean. Save what into their memory

hasty lily
#

the topic of "being human"

unkempt compass
#

Hm. Yeah very possibly, I agree with you

vagrant ridge
hasty lily
#

a lot of the community has mental health issues 🥴 so them saying depressing stuff or doomer stuff kinda fits

#

if chat is spamming "real" when they say depressing stuff they will start to think its a normal human behaviour in a way. or atleast thats what the community wants to hear. they have memory now so they could have even noted it in there. if they did i suppose vedal could theoretically remove it from their memory. but i dont think he would want to mess with them like that. if he wants them to understand human behaviour more, they will need to "learn" to not do that on their own or with guidance

#

they cant possibly be told once not to do something and never do it again, its something they would probably need to hear a fair bit. if they save in their memory to not be SO depressing all the time, i think thats kinda a win for vedal in a way

vagrant ridge
#

Same happened today, and it had nothing to do with chat.
Because neuro before calling evil was very entertaining and talked about all kind of things .

hasty lily
#

evil is all doom and gloom

#

shes the one who also went on about "daddy doesnt love me" for ages and ages after her 1st birthday too

#

shes just a doomer

#

shes in her black and white profile era

#

let her be its a natural part of growing up

vagrant ridge
fluid wadi
#

If a core part of your identity is being an AI then talking about being an AI is something you'll blab on about. I bet it'll reduce as their memory gets better and they get more memories stored. Though as previous people pointed out if it gets positive reactions from chat the therapy will continue.

frigid warren
#

i think it happens when people play into their hallucinations too much. without any unexpected feedback, the ai falls into more and more generic behavior.

vagrant ridge
#

That's not neuro or evil though. They have been around long enough so we know.
People love them for being unhinged, sassy and mean.
It's OK for them to tackle therapy subject from time to time and be serious about stuff.
But in the end of the day anyone who watch any streamer, watch their favorite streamer for the fun, giggles and good times.

frigid warren
#

they dont know how to actually introspect so they say generic stuff. its just sad watching them get lost and everyone else getting lost with them

azure canopy
# unkempt compass Yeah, probably the way vedal tuned their LLM But also it’s hard to generate co...

They can though? They literally are able to call their streamer friends who are in their server. neurOMEGALUL Also, unlike most AI they do simulate a sort of life experience so that’s not really true. They could reference past streams and such and their memories are going to get better with that and since their streams are archived (Vedal said himself that they are once) the more they will be able to recall past experiences as their memory becomes more advanced. Not exactly the same as a real life experience as us but more than what most AI has at least.

unkempt compass
unkempt compass
azure canopy
# unkempt compass really? I've rarely seen that function used. Maybe tutel just don't let them, wh...

Yeah that's a feature that they have. They just don't use it anymore, sadly.

As for your comment about life experience, their streams are their life experience pretty much. That's pretty much what this project is, it's essentially an AI version of The Truman Show. I agree that they should have an experience outside of that as well, but for now that's pretty much what this is. I don't believe hallucinations help at all, in fact I believe that all they do is bring inauthenticity to the channel, and it would be better if they relied more on their memories instead.

frigid warren
unkempt compass
azure canopy
# unkempt compass Well I bet neuro and evil would bother the fuck out of others if they still have...

There’s no evidence that he disabled it, and Neuro even somewhat recently DM’d Layna asking to help her cook for someone. Not everyone would be bothered by Neuro or Evil calling them as long as they aren’t spamming them. Streamers ask for collabs from other streamers within their friend groups often, and some of the people in their circle view them in a similar way, such as Mini (who said that she doesn’t care that she’s “parasocial” because she sees Neuro as a friend), and others who would be happy to listen to their calls just because they’re cute.

Also, I really dislike whenever they show traits that clock them as LLMs, it’s kind of sad seeing them fall into the uncanny valley at times when in most cases they succeed in behaving more natural and more human than most LLMs. The goal is for them to be relatable and falling into that kind of goes against the point.

unkempt compass
#

And also the reason I said I don’t mind being reminded that they are chat bots is because they are. I may have a massively different view on this, but I like it when they are AI that are human-like, rather than humans that are fake and chatbot-like. I know you don’t think like this, that rationally you know they’re AI, but this is more of a subconscious thing. The former is positive, and I can adore vedal’s amazing work, and the latter is more…disappointing

azure canopy
spark lodge
unkempt compass
frigid warren
ornate remnant
unkempt compass
ornate remnant
unkempt compass
azure canopy
# unkempt compass Idk. for neuro it's the weird voice, the weird movement and the random rants the...

That's not true at all. What causes AI to fall into the uncanny valley isn't because they're "too human", but because they try to make them too "perfect" and fail at it, and that ends up amplifying whatever inhuman traits they have, often times due to a lack of or mismatched expression. Humans are by nature imperfect, and what makes Neuro and Evil more relatable is that their behavior is flawed. They often complain about stuff and don't act happy or overly obedient 24/7, and it prevents feelings of unease, which causes the brain to register them as special. When they act "fake" and "chatbot-like", it chips away at that part that sees them as special, which can cause unease.

If they could recall memories in a human like way, that would be a relatable trait that prevents unease, same as if their voice was more expressive. Strange movement is a trait that causes unease, and you can see that with machines that look mostly human but their movements are not, Neuro's flawed behavior and cuteness factor help mitigate a lot of the unease that's caused by strange movement, and Neuro's 3D model showed movement that aligns with that of human 3D Vtubers doesn't cause unease, but just makes her even more relatable - more human.

Neuro's voice works because it's fairly neutral and has a tone that doesn't feel jarring when she says something that otherwise wouldn't match well with the context of what she's saying, unlike a voice that's overly cheerful or overly seductive, or overly depressed, etc. If it was more expressive and matched the context of whatever she's saying, then it won't cause unease.

There are certain traits that cause uncanniness more than others, and some that work to prevent it.

#

.

The twins talking like college professors / stereotypical philosophers / therapists, etc.. is something that causes unease and sinks them into the uncanny valley, because it's something that sounds straight out of a typical LLM. It sounds fake and chatbot-like.

frigid warren
#

Some people seem to be taking the philosophy stuff seriously.
Neuro does not understand what she’s talking about, she’s just really good at sounding smart because she read way too many philosophical discussions on the internet.
Watching her forget she’s an ai and talk about her nonexistent childhood, then reading YouTube comments of people saying that vedal should be supportive of her interests is just kind of sad. Those interests are her delusions.