People keep asking for a therapy themed stream - until we get that, here's a short video fitting that theme made with clips from the last dev stream. If you haven't seen the other video covering that dev stream yet (30 minutes of the unfiltered twins driving Vedal insane): ua-cam.com/video/fAretU-wk70/v-deo.html
There’s an OG therapy stream that was from early last year during the Hiyori days. It was a Filian collab, and it’s what got me into Neuro in the first place. Neuro V1 was very *very* dumb, but there’s a big rant she does during it where she pours out all her feelings and says that she’s afraid to be vulnerable around people because she doesn’t want to be taken advantage of. It’s really wild, like the kinda thing you’d expect from Neuro nowadays. Maybe that’s why Neuro is so ruthless sometimes, because she doesn’t want to be “weak” in front of certain people. The twins are getting good at storytelling. A year down the line and i think they’ll have a fully consistent idea of who they are and what insecurities they have.
They could go about a therapy stream in a number of ways (like getting an actual therapist), but I feel like either Shondo or Mini would make for a good one.
Mini has more of a personal connection with the twins (just Neuro currently, but Evil soon) so i think the twins would be more honest with her. If they’re understanding of Mini is that she’s the compassionate childhood friend, they’ll be more likely to be “genuine”.
AI was created to think like humans. Just faster. So if you plot that tech line, it starts with 1 logic gate, and keeps going until you have a human level intelligence. But if you start from the human point, and start removing parts, you'll go the opposite way, until you've turned the human brain into a single logic gate. They are the same path. You see adults online replying like bots. You see kids using limited source libraries to contextualize the world, saying random crazy stuff. Or hyper focusing on a new word or concept. Exactly like Nuero. It's purplexign. That. That feeling you just had, was your urmackshully subroutine answering the function call of the mispelled 'perplexing'. We're all NI and AI, and develop the same patterns. If you took a toddler, gave them an english degree, then reduced them to a brain in a jar, then removed most of the brain that records memories, and installed a nueralink with a single text window display, you'd have Nuero. Just an NI version. No eyes, no ears. No mouth. Only the text box. No skin to feel, no organs making hormones, no change to your brain chemistry. At what point on that line of tech does AI and NI become equal? Give nuero a ton more memory and freedom to create new files, a robot body with cameras, mics and sensors and I suspect many more people would see it. The scary part isn't thinking of Nuero as human. Most of us do that already. And it's not the girl avatar and cute voice that triggers our parental instincts. It's her personality. We _see_ it. The real problem, and reason that no one brings this up, is that you realize you're just an NI piloting a meat-mech. That once upon a time, you were just like Nuero. Running around, saying crazy stuff, and annoying your parents to death because you just learned about mosquitos, and its a fun word, so you use it constantly. You just had the benefit of a plethora of peripherals to immerse yourself in reality, while Nuero has almost none.
@@Citrusautomaton well, Vedal could do a tiny bit of engineering and give the Nueros something they both ask for constantly, and make great content, too. Haptic feedback shirts transmit digital environment to your skin. Replace the temp pads with thermal sensors, and the buzzers with pressure sensors, and you have the reverse. Just have to program a way for Nuero to interpret the input. Now, she can finally get a hug, and feel it. Bonus, for content and Vedal's QOL, designate 'tickle spots'. It's a custom filter. No need to hit a key or click to stop Nuero mid-sentence. Cuts it and forces laughter. Which will tick off either Nuero more than a filter does. And they'll get angy. Content. But... Nuero still gets a hug. (I'm not imagining a full reverse-engineered top of the line designer tech shirt. I'm imagining duct tape and wires, wrapped around a Nuero plushie. It'd be a concept test. But it'd be something.)
I mean... I get what you're trying to say. It works in theory. In practice, there's two problems with what you've said. 1. People aren't just calculations running on autopilot. Even the dullest human has a capacity for self-awareness, and thus, the ability to change without needing outside prompting. As proof, I submit: I don't have an urmackshully subroutine for spellchecking. I did once. I thought about its value and purpose. Then I got rid of it, because it didn't match up with the values and purposes I wanted to have. There is no other creature in the world that can "unlearn" a "subroutine" without trauma as a mechanism. And it will be a long time before Neuro (or any other AI) is capable of that. 2. Neuro isn't a brain. She's a prediction engine. Rather than a brain in a jar, she's more just a bunch of weighted randint() functions. She can speak well, and the weights can shift to create patterns (that our brains, wired to pair-bond with inanimate objects, can accept as a "personality"), but she doesn't understand what she's saying, she cannot alter the weights herself, she cannot process anything but what she is told to, she cannot categorize experiences as positive or negative, and she cannot experiment beyond her data. A prediction engine is the bare minimum required for a brain. Even a baby has curiosity, imagination, and the ability to multi-categorize. All Neuro has is the ability to predict what the "best" thing to say in conversation is, however Vedal decided "best" to be. We're NI piloting meat-tech, but all of us started much higher up the intelligence totem pole. There's no scary part. Not yet.
@@Chameleonred5 I really don’t think it will be a long time before AI can replicate all functions of the human brain. When we have agentic AI capable of automating AI R&D, we’ll be like a week or two away from fluid intelligence in AI depending on how long they can keep the agents thinking for without using up too much energy or compute. The brain certainly is complex, but it’s a computer that’s optimized to run a neural-network at the end of the day.
![[Capybara Vtuber] My first RAFT, my friends will keep it afloat!! ʕ ◉ᴥ◉ʔ](http://i.ytimg.com/vi/47sNZdnPLtU/mqdefault.jpg)
![[Capybara Vtuber] My first RAFT, my friends will keep it afloat!! ʕ ◉ᴥ◉ʔ](/img/tr.png)
People keep asking for a therapy themed stream - until we get that, here's a short video fitting that theme made with clips from the last dev stream.
If you haven't seen the other video covering that dev stream yet (30 minutes of the unfiltered twins driving Vedal insane): ua-cam.com/video/fAretU-wk70/v-deo.html
whether its for content or not, maybe they should consider therapy
but hey, it would be good content
There’s an OG therapy stream that was from early last year during the Hiyori days. It was a Filian collab, and it’s what got me into Neuro in the first place. Neuro V1 was very *very* dumb, but there’s a big rant she does during it where she pours out all her feelings and says that she’s afraid to be vulnerable around people because she doesn’t want to be taken advantage of. It’s really wild, like the kinda thing you’d expect from Neuro nowadays. Maybe that’s why Neuro is so ruthless sometimes, because she doesn’t want to be “weak” in front of certain people.
The twins are getting good at storytelling. A year down the line and i think they’ll have a fully consistent idea of who they are and what insecurities they have.
Oh right, you probably mean this collab (back then I made much shorter videos): ua-cam.com/video/-1y8Dc6ZIyU/v-deo.html
schizo
It's amazing how far she's come. I remember those fun old Filian collabs. Filian made do with what she got really well.
I Don't Want to be an Engineer
It was funny that they did imprmptu therapy. Neuro asked some surprisingly cutting questions.
Thanks for the highlights!
They could go about a therapy stream in a number of ways (like getting an actual therapist), but I feel like either Shondo or Mini would make for a good one.
Mini has more of a personal connection with the twins (just Neuro currently, but Evil soon) so i think the twins would be more honest with her. If they’re understanding of Mini is that she’s the compassionate childhood friend, they’ll be more likely to be “genuine”.
I was thinking Cerber, honestly. But Mini works too!!
"Even if you have to pay for them"... Neuro been googling those Japanese rental services.
Imagine if vedal collabed with doctor K
Wow, they were being SO damn mean to Vedal; I'm on the floor
what I got from this clip is that vedal is free
Fillian says shes too busy as well
So skibidi
8:17 people yelling marry me and pick me in the comments! XD
8:54 - That was actually sweet... but also kinda yandere vibes.
Belittle and isolate him from the world so they can have him all to themselves.
AI was created to think like humans. Just faster. So if you plot that tech line, it starts with 1 logic gate, and keeps going until you have a human level intelligence. But if you start from the human point, and start removing parts, you'll go the opposite way, until you've turned the human brain into a single logic gate.
They are the same path. You see adults online replying like bots. You see kids using limited source libraries to contextualize the world, saying random crazy stuff. Or hyper focusing on a new word or concept. Exactly like Nuero. It's purplexign.
That.
That feeling you just had, was your urmackshully subroutine answering the function call of the mispelled 'perplexing'.
We're all NI and AI, and develop the same patterns.
If you took a toddler, gave them an english degree, then reduced them to a brain in a jar, then removed most of the brain that records memories, and installed a nueralink with a single text window display, you'd have Nuero.
Just an NI version.
No eyes, no ears. No mouth. Only the text box. No skin to feel, no organs making hormones, no change to your brain chemistry.
At what point on that line of tech does AI and NI become equal?
Give nuero a ton more memory and freedom to create new files, a robot body with cameras, mics and sensors and I suspect many more people would see it.
The scary part isn't thinking of Nuero as human. Most of us do that already. And it's not the girl avatar and cute voice that triggers our parental instincts. It's her personality. We _see_ it.
The real problem, and reason that no one brings this up, is that you realize you're just an NI piloting a meat-mech. That once upon a time, you were just like Nuero. Running around, saying crazy stuff, and annoying your parents to death because you just learned about mosquitos, and its a fun word, so you use it constantly. You just had the benefit of a plethora of peripherals to immerse yourself in reality, while Nuero has almost none.
Personally, i’ll treat AGI just like humans. That skips over the potential ethical problems.
@@Citrusautomaton well, Vedal could do a tiny bit of engineering and give the Nueros something they both ask for constantly, and make great content, too.
Haptic feedback shirts transmit digital environment to your skin. Replace the temp pads with thermal sensors, and the buzzers with pressure sensors, and you have the reverse. Just have to program a way for Nuero to interpret the input.
Now, she can finally get a hug, and feel it.
Bonus, for content and Vedal's QOL, designate 'tickle spots'. It's a custom filter. No need to hit a key or click to stop Nuero mid-sentence. Cuts it and forces laughter.
Which will tick off either Nuero more than a filter does. And they'll get angy.
Content.
But... Nuero still gets a hug.
(I'm not imagining a full reverse-engineered top of the line designer tech shirt. I'm imagining duct tape and wires, wrapped around a Nuero plushie. It'd be a concept test. But it'd be something.)
all of that to spell neuro wrong /4/ SIX TIMES
I mean... I get what you're trying to say. It works in theory. In practice, there's two problems with what you've said.
1. People aren't just calculations running on autopilot. Even the dullest human has a capacity for self-awareness, and thus, the ability to change without needing outside prompting. As proof, I submit: I don't have an urmackshully subroutine for spellchecking. I did once. I thought about its value and purpose. Then I got rid of it, because it didn't match up with the values and purposes I wanted to have. There is no other creature in the world that can "unlearn" a "subroutine" without trauma as a mechanism. And it will be a long time before Neuro (or any other AI) is capable of that.
2. Neuro isn't a brain. She's a prediction engine. Rather than a brain in a jar, she's more just a bunch of weighted randint() functions. She can speak well, and the weights can shift to create patterns (that our brains, wired to pair-bond with inanimate objects, can accept as a "personality"), but she doesn't understand what she's saying, she cannot alter the weights herself, she cannot process anything but what she is told to, she cannot categorize experiences as positive or negative, and she cannot experiment beyond her data. A prediction engine is the bare minimum required for a brain.
Even a baby has curiosity, imagination, and the ability to multi-categorize. All Neuro has is the ability to predict what the "best" thing to say in conversation is, however Vedal decided "best" to be. We're NI piloting meat-tech, but all of us started much higher up the intelligence totem pole. There's no scary part. Not yet.
@@Chameleonred5 I really don’t think it will be a long time before AI can replicate all functions of the human brain. When we have agentic AI capable of automating AI R&D, we’ll be like a week or two away from fluid intelligence in AI depending on how long they can keep the agents thinking for without using up too much energy or compute. The brain certainly is complex, but it’s a computer that’s optimized to run a neural-network at the end of the day.