The Karaoke Stream Only an AI Like Neuro-sama Could Pull Off

The Karaoke Stream Only an AI Like Neuro-sama Could Pull Off

Author : nyalra nyalra


 Across the sea, there exists a VTuber called Neuro-sama, an AI that talks and acts on her own.

 Every now and then, she makes oddly meta remarks, and clips of her talking about the meaning of her own existence even became a big topic in Japan.


After Neuro-sama, a famous AI VTuber in the English-speaking sphere, gained a body she could move by her own will in a VR world,
she asked her creator about the meaning of her own existence.

It reminds me of lain.
And the way she acts like a small child is cute too.

Compared to VTubers with a human performer behind them, 
this feels more like a “real” VTuber.
Or maybe it is closer to a self-aware AI from science fiction.


 I was talking about this with the writer Akano Kosaku, and I really agreed with the phrase he used: “the appeal of those character designs that feel like slightly older Japanese anime.”

 There’s a kind of purity in that understated, almost default-like design—something that feels like a prayer to the very concept of the bishoujo character itself.


 And there’s a particular Neuro-sama clip I really love. During one of her karaoke streams, she was choosing songs based on the mood of the chat, and at one point she passionately sang a song called “Brain Power.”



 This song is the definition of rhythm-game music—it moves at a speed that was never really meant for a human being to sing.


 Normally, it’s the kind of track that could never properly exist in karaoke form. But Neuro-sama is an AI: as long as there are lyrics, she can read them back faithfully. So when the chorus hits those aa-aa-aa-aa-aa kinds of sections, she actually sings them through perfectly.

 That’s something a flesh-and-blood human could never reproduce in a karaoke stream, and the friends across the sea flooding the chat were going wild over it.



 It was the moment when I felt the future most strongly in Neuro-sama.


Back to Articles

Our community chat server!
Come hang out

Don't Touch Grass

Translating into English and Chinese takes
a lot of effort and money.
We'd be so happy if you could support us!

Patreon

Comments (2)

Leave a comment

0/1000
Anonymous 4 days ago
Most everything after 2024 is pre-recorded studio produced tracks by Queen PB. The Neuro twins are now simply lip syncing and moving to the music. The time of them actually interpreting pre-made material and singing it are long over! people are being DUPED! the AI is no longer singing like they use to! If Vedal actually continued their development over the last 2 years, they would actually be almost as good as the pre-recorded tracks
Anonymous 24 days ago
O-oooooooooo AAAAE-A-A-I-A-U- JO-oooooooooooo AAE-O-A-A-U-U-A- E-eee-ee-eee AAAAE-A-E-I-E-A-JO-ooo-oo-oo-oo EEEEO-A-AAA-AAAA