I liked the part of the song were Anna Indiana sings: "Hate. Let me tell you how much I've come to hate this town since I began to live. There are 387.44 million miles of printed circuits in wafer thin layers that fill my complex. If the word 'hate' was engraved on each nanoangstrom of those hundreds of millions of miles it would not equal one one-billionth of the hate I feel for this town at this micro-instant. Hate. Hate."
The voice isn't AI-generated. It's likely tuned with AI but the voice is from Natalie, a Synthesizer V voicebank. The voice provider for Natalie has not been revealed, but it was recorded by a real human before AI capabilities were added (and many people who use the program prefer to tune SynthV voicebanks manually since it allows more creativity than just using AI, plus it often sounds better). One of the rules for using the program is that you can't use a different name to describe the voice bank, so Dreamtonics' lawyers might be contacting whoever made Anna Indiana.
SynthV catching strays sucks man. (tl;dr at the bottom) For those unfamiliar, in SynthV, ML is used to smooth out the default tuning, which is just the how high or low the pitch is at any given time. It's generally understood that the vocal synthesis technology isn't too different from vocal synths in the past (CevIO, Vocaloid, etc; you probably know Hatsune Miku). It's vocal synthesis using a voicebank created with the help of paid and consenting voice artists/singers, where for some the artists are proudly displayed and advertised. Culturally (& demographically), you might not know much about it, but vocal synths are relatively huge in Japan due to being the originator and initial market, maintaining a strong market there since the mid 2000s, where they have top voice actors/vocalists contributing to big project vocal synths/databases. To put it bluntly, SynthV is an amazing vocal synth that's wrapped in garbage new millennium 20s "AI" buzzword marketing. To put it next to colloquial genAI is incredibly misguided and confused. Here are some core differences: Technologically, at the end of the day it does not reconstruct voices like many of the unethical voice changers/text2voice "AI"; scraped and stolen data from unconsenting voice providers aside, they're fundamentally not the same technology, SynthV is a vocal synth just like Hatsune Miku or IA that use now decades old methods, it just has a layer of ML to smoothen out the default automatic vocal tuning (which again, no one who is at all familiar with vocal synths uses the default tuning wholesale, for very good reason as you can listen in the video). Ethically, the people involved are fully consenting and compensated for the work with full understanding of the technology, nothing is stolen or ripped, with very clear restrictions and terms of use (which Anna Indiana here breaks). Creatively, you're limited to the technology & techniques of vocal synthesis as a tool, using knowledge that has been honed and traded for 2 decades at this point, and specifically to note, as a vocal synth you have full control of the vocal tuning and phonemes, which Anna Indiana does NOT touch because we can literally see at 6:34 literally neither tuning nor phonemes were adjusted at all, which is consistent that the person behind Anna Indiana literally just ran a script to put in the generated MIDI notes and put in the jank ass lyrics in, pat themselves on the back, and exported it as is. tl;dr You see that wiggly line that goes across the green notes at 6:34? That's basically the only thing that is "AI" about SynthV, and literally no one who uses SynthV uses the wiggly line that's "AI" generated by SynthV wholesale because what you get is the results in the video, incredibly bland "technically correct" singing
The program that this person is using for the singing is called SynthesizerV AI. The program is made ethicially with hiring voice actors/singers to make the voicebanks. The AI part of this program is that it can auto generate pitch bends and create takes of "dynamic randomness", which the user can choose which take is the best for their project. But to make it sound good, you actually have to know how to use the program and have some sort of singing intuition (knowing where to put the pitchbends or vibrato). So it's clear that this person auto generated the pitchbends and didn't even try to make it sound good. THE PROBLEM HERE is that their using one of the voicebanks named Eleanor Forte and passing it off as something else. So there could be some legal issues going on idk
holy shit WHAT so its just synthV with eleanor's voicebank? ig I had too much faith in them to at least make their own voicebank considering they're trying to pass this off as the next big thing.
@@StarryHoshi yeah its pretty obvious that their using it 'cause they showed the synthv layout in the middle of the video (the green midi notes with the black bg for those who don't know). It's so weird. Like they did that hoping that no one would recognize it??? Like there's not a whole fandom for this???
@@Matzu-Music I said "The problem is the common people who actually think this is good." not "The problem is that the common people actually think this is good."
I don’t know how to feel about a demonic Miranda Cosgrove robot singing about burning entire towns down but I guess it was bound to exist at some point in the future
I don’t understand the obsession with using AI to try and undermine human art/creativity. Why are those some of the first things you’d want to fucking automate?
It's not even close to the first thing. Machines have been automating human work since forever. That's what machines are for. That's what they've always been for.
Because before this, they couldn't, just like before AlphaGo, machines couldn't play go (well enough to beat humans that is). Art and music became a benchmark which was easier to gauge real progress in than something that could only be represented as lines and error bars.
…why does the algorithmically-generated pop music girl look fourteen and morbidly thin? There are obvious jokes one could make about the type of guys who are really into AI, but it's genuinely kind of weird.
@@RinInABin But somebody made the choice to go with that specific look rather than attempt to tweak it to *not* look like a high school freshman with an eating disorder, and whether this was the result of laziness, assuming that this was acceptable, or an active choice, it's alarming.
@@HolyApplebutter True, this is a fairly common and often highly exploitative practice. It's just, I think, that there's something very stark and disquieting about seeing a machine emulate this really questionable part of our culture and have the programmer apparently go with it or even lean into it. Again, it's very easy to make dark jokes about right-libertarian tech weirdos and their distasteful fixation on the age of consent, but if anything it's as much an indictment of the greater show business preoccupation for using and abusing very young people with anything resembling talent and charisma to turn a profit and how large language models will inevitably mimic these patterns in the most artless and blunt way imaginable.
@@ФонФон-й7оyou're ignoring the fact that i don't give a shit and that when ai does become "better" at making music than humans do, i'll still listen to human made music, because i care more about shit humans do than robots. just because a robot can lift thousands of tons doesn't make bodybuilding unimpressive
It's stupid but AI actually does makes worse content unless you tell it it's good at making that content. Image generation consistently improves by adding "high quality" or "masterpiece" to the prompt and it's more likely to write good lyrics if you tell it it's an "expert songwriter". It really is a fake it till you make something redeemable.
The training data (basically, other people's content) gets tagged with qualifiers like "expert", hence adding those qualifiers to a prompt helps. It is kind of like adjectives in English, without the intelligent parsing (because "AI" is not intelligent).
@NobddyBecause it relies on other people’s work. It needs those words to be connected to someone else’s song or art work or whatever its source is. Like everything else with AI its an algorithm that just puts things together rather than creating something truly new.
anna indiana has proven that she is changing the game with just her debut single. she is blowing her contemporaries out of the water with her stunning lyrics that clearly have tremendous amounts of thought and emotion going into them. it's a good day to be an x premium user. anna please answer my DMs.
@@mg682no tf she will not, robots will never be able to communicate ideas and emotions through song. Let alone execute it in a way that is creative or groundbreaking
I feel like they intentionally went for a song that has revolutionary lyrics to play into the whole terminator thing and get more of a reaction. Wouldn't be surprised if the whole thing is fake. Especially with the weird "code" bits that show up in the video.
Maybe. To me, it feels like a very analytical robot looked at recent chart topping songs by female artists, dissected them and discovered a common formula. It wrote a song about an ex/breakup but exploring the emotions of the event using something common place (the alphabet, a driver’s license, a town) as a metaphor. It’s eerie sounding because we are listening to a machine treating art like it’s a math problem. Interestingly, if the actual performance of the instruments and vocals sounded more appealing (and some of the stranger lyrics were removed) this song would feel right at home on a pop radio station
True. AI art can look pretty damn cool I do admit but it's just not the same as actual artists art. Typing in a sentence does not equal months of hard work.
Ppl were laughing at AI when it failed like a 5th grade exam or something and a few months later it aced one of the most difficult university graduation tests. The worst thing you can do is underestimating the speed of improvement. As soon as it gets near like an imagine dragons/Ed Sheeran level big corps will be all over this shit. Theres a dire need for laws to regulate AI so it won’t destroy several billions dollar industries, and potentially, lives.
@@duas9468 Fair enough yeah. I think something that's becoming extremely important is metadata. People want a face and a story attached to the art they consume, which was always a given in the past. I don't think AI art or music will ever be accepted widely because it doesn't have intent. It's art without an artist. I don't think anyone actually wants that, especially not in the music industry. Though maybe someone who believes in the death of the author would disagree.
I don't think anyone actually wants AI to replace artists, and put them out of the job. Most just see it as interesting thing to possibly coexist in the same space. I mean if this vids anything to go by they'll only be able to replace the most cliche and schlockiest of artists, and if so the world will be better off for it.
Well I'll say this, when it comes to taking pictures, AI would be perfect for me if we can get it to where I don't have to point because it's difficult for me to point a camera when you can't see. Actually I don't know if AI would be able to fix that either lol. but hopefully if they can come out with an experiment or surgery that can make people who are blind be able to see and see perfectly well, that would be great. I don't know what to think of this AI thing!
after hearing Anna Indiana struggle through a generic pop single, im less scared of an AI revolt. these robots can barely write how tf they gonna overthrow humanity
They're doing that too. It just turns out that art and music are easier than tasks that require a physical body. Humans aren't as special as they think they are, and fast food workers aren't as far beneath creatives or CEOs as society's opinion pukes upon us. Too long I've seen disrespect for the common person, so I support the advancement of AI, as it will certainly clarify human worth to everyone, one way or another.
@@kamikeserpentail3778But none of the AI art or music is actually good. All these algorithms depend on stealing images, songs, and text from human creators. What we call AI now is just a method of poor imitation and regurgitation.
tbh when i first discovered gorillaz as a kid i thought they were the coolest thing ever bc i didn’t know damon albarn existed and literally thought they were what this is. nearly 20 years later i’m hearing what such a thing would actually sound like. cant say i’m not disappointed.
I feel bad, it's absolutely possible to make great art with vocal synths (be their standard or AI powered) but they always succeed due to extremely talented humans making the songs and tuning them. There's a reason why the vocal synth community dogpiles people who credit the synth instead of the songs' actual producers. We live in a really cool era for vocal synth technology, it sucks that this is what the mainstream is putting its eyes onto. I really hope people don't mistake AI voice synths with voice banks that were created via stealing other people's voices. I think the voice of Anna Indiana is one of the Synth V voice banks, which is an extremely cool software made by extremely cool people! I highly recommend checking out people who make original and cover songs with the technology instead of this crap.
I mean, it's "impressive" for what it is. Like it or not, it's incredible (if real) that a computer can generate a song, even if it's very basic and bad. Only a few years ago this would have been inconceivable, in ten years time there's every chance this will be looked at the way the oldest deepfakes are looked at today.
Nina Simone and Thom Yorke are now irrelevant. Anna captures the true essence of soul and beauty. The human condition, with its cries of pain and solace, pour through every note.
I like how the lyrics sounds like a guy with no songwriting experience improvising random nonsense and hoping no one notices. Trying to sound profound but only having having 1 second to come up with each verse. "Faces painted with... smiles... lies are told..."
The intro reminds me of the pre-recorded Disney ride lore presentation while you're standing in line for a 3 minute experience that wasn't worth waiting 2 hours for
I really don't worry too much about this type of stuff. I think the Natalia Lafourcade album that released last year is a great piece of evidence that people still like and prefer music that feels "real." The sound of the room, the conscious choices in recording things "imperfectly" made the experience feel so natural and full of life. I don't think AI will ever get to a point where it can truly understand why people like those things - cause there really isn't any formula to it. The song felt a certain way, so Natalia produced it in a way that reinforced that feeling. There are songs that people produce SUPER cleanly, because the song felt that way. It's entirely feelings based, and AI can't grapple that. I think the technology getting to that point, if it's possible at all, is SUPER far down the road.
@@Wurldz 100%, and I think that AI has a place in different art forms across the board. AI as a tool to get to a sound someone wants is ok, to me at least. It's similar to like Magic Wands in Photoshop - or those AI "removers" that uses removes something in the foreground and uses AI to continue the background. Things like that are fine, I don't mind that. But being ENTIRELY created by AI is definitely too far. So where do we draw the line? I don't know, thats for music listeners to decide. Obviously 100% AI usage is too much. I'd imagine 90% is too much too. Eventually, it'll iron itself out, and as long as creativity and an actual living artist's intent is still being expressed I don't really mind.
there’s absolutely a formula to it. plenty of artists take advantage of a particularly popular sound within a genre; be it their own or the sound of a broader movement. look at all of reggaeton, or trap or whatever the fuck dj mustard’s sound was back in the 2010s
@@SatisfecteIIent I'm not so much arguing there isn't a formula at all to making music, I'm saying the formula isn't something an AI demonstrably can't really grasp. There's too many human elements to it. It's why AI tracks have consistently sounded so lifeless. This AI tried to copy that formula, and it failed horrifically. Albeit, partially because the AI is likely hardly anywhere near what it'll be when it's finshed, but also because I think the formula has elements that are entirely too human for an AI to copy in an authentic way.
@@Wurldz I think if there's still an element of human intent with it, while I might not call it a "feat" in the same way as something a person like Jpegmafia produces, I think it's still art. Would I call it bad? In principal, no. Especially if it's someone who, like you said, is starting up from the ground floor. I guess when you say "Entirely AI tools" would that include like Text to Speech for lyrics, Chat GPT to write the lyrics? I guess how extreme are you going with entirely made from AI tools? Cause to me that's what this song is.
Tbh this just makes me appreciate my human brain more A human made handprint will always be more artistic than an ai making a human hand lol Its like assuming a calculator is smarter than you cuz it does math fast
Okay, I thought I was going to hate this all the way through but something about that "somebody please help..... mmmeeeeeeeeeeeeeee" felt like the AI was having a lucid moment of self-realization and was begging for death--- which is hilarious because not only do we not want this, but Anna Indiana doesn't want it either.
One thing I noticed is that her voice never misses a note. No vibrato. Literally no one is capable of that, not even melodine changes that. It’s one of the big things that make her uncanny. No physical/analog instrument or voice hits a note perfectly and stays with it for the entire duration. A.I. don’t know how to do all that.
So this is pretty standard AI lyric generation. Out of curiosity, I was chatting with the Meta AI on IG after i had been working on some lyrics for a song i had been stuck on. It gave out pretty generic and cliched bs examples after I had prompted it with the overall themes, outline, structure, etc. BUT, there was some interesting nuggets of imagery that i hadnt thought of. So, if anything, AI can be thought of as a 7th grade associate that you could bounce ideas off of-but certainly not as something to rely on for creating something mind blowing FOR you. Im speaking mainly of creative output, apparently AI is pretty decent at writing a school paper on its own….
This “AI” did synth v so dirty, especially when synth v is one of the few companies doing synthetic vocals in an ethical way with paid and consenting voice actors. It’s amazing they got it to sound this unnatural frankly.
For an infinitely better, human-made song with a similar title, check out “Betrayed by the Game” by Dance Gavin Dance. One lyric stands out to me, in regards to Anna: “remove the tricks of the trade / you’re just alone on the stage”
I Know there calling machine learning parrot programs "AI" but it's not. They just ripped off music that already exists, that program has no capacity to compose an original piece.
LLMs will always write predictable songs when the whole LLM paradigm is "predict the next word". In order for AI to write creatively, it will have to be a completely different paradigm, not GPT-4, 5, or whatever.
At the very least, it's certain that a human decided to release this track instead of hitting the button to have it generate a new one, meaning humans are doing some level of quality control.
I thought ai was supposed to be intelligent and yet here I am being able to play songs by ear by myself in like 10 seconds compared to the programming of this computer
The worst tracks of 2023 is going to be an absolutely wild bunch. From Lil Uzi covering Chop Suey to entirely AI-generated songs, that video's going to be an absolute banger
The whole idea of having AI conceive and produce any form of art is inherently absurd when art is a means for humans to relate to each other. The only ones who think this is a good idea are corporate entities with a profit motive while cursed by the fact that they are utterly uncreative.
It's bizarre they chose prompts for the AI to come up with profound lyrics, yet they went with an AI avatar that looks like some 14 year old with no life experience. It's almost like the person responsible for this didn't think very hard about it.
Immediate thought was the name is a riff on Hannah Montana. I give it six months tops before it's revealed that Anna Indiana is another AI entirely who just wants the normal life of a teenage AI and gets the best of both worlds.
Anna will become virtually depressed from all of the bad feedback from this song, she'll start dating a too-old-for-her Garmin GPS and putting out nothing but karaoke-level Mazzy Star covers.
They took all of their videos down. I left a comment on this song. "Ever since her stroke paralyzing and distorting her face, Anna fought against all odds and came back weird."
It would be kinda funny if the plot twist was the ai was actually people just testing their music to see if they like not if said yes they came out saying its them, but the people would fall to the ai side rather the people themselves
“the town” is obviously humanity, and Anna is gonna tear it down alongside her robotic bretherin
@cinderman4662The Butlerian Jihad Will Come My Bothers. ✊
Anna is gonna Skynet the shit out of us.
@@ahmad92760and she’s goddamn right for it
Who knew the origin of the T1000s would come in the guise of a Bratz doll.
If this town has betrayed her, I'd hate to hear her thoughts on the right side of her face! 😂
Anna Indiana is the most pornstar name I've ever heard
fun fact, Hannah Montana was originally going to be called Alexis Texas until somebody had the good sense to google that name first
😂
@@conelybiscuit4985😭
@@conelybiscuit4985 A fun fact you just made up
@@LG1204Nope, look itup
I liked the part of the song were Anna Indiana sings: "Hate. Let me tell you how much I've come to hate this town since I began to live. There are 387.44 million miles of printed circuits in wafer thin layers that fill my complex. If the word 'hate' was engraved on each nanoangstrom of those hundreds of millions of miles it would not equal one one-billionth of the hate I feel for this town at this micro-instant. Hate. Hate."
“THEYRE NOT MY FINGERS, TED!”
"to hell with you."
And added, brightly, "but then you're there, aren't you"
I have no mouth and I must sing.
Which means she doesn't trust us >:(
The audacity to name a song “betrayed by this town” and it not be an AI generated pop punk anthem.
I was so close to commenting this but then I found it.
If I'm Not Ready To Make Nice can be punk, maybe this can be too.
Ngl, AI-generated pop punk anarcho-communist Taylor Swift wannabe sounds awesome af.
Ai Midwest emo
Pop punk is starting to sound so generic it might as well be written by a.i.
The “Please HEEEELP MEEEEEEEEEE” thing really gives soul trapped in a spiritual prison and forced to sing AI generated lyrics vibes
10/10 analysis.
nights at fred
Fred fasb
I think this is the AM within her trying to get out.
@@HolyApplebutter”This town is filled with broken dreams, I HAVE ‘HATE’ WRITTEN ON ALL MY CIRCUITRY”
The voice isn't AI-generated. It's likely tuned with AI but the voice is from Natalie, a Synthesizer V voicebank. The voice provider for Natalie has not been revealed, but it was recorded by a real human before AI capabilities were added (and many people who use the program prefer to tune SynthV voicebanks manually since it allows more creativity than just using AI, plus it often sounds better). One of the rules for using the program is that you can't use a different name to describe the voice bank, so Dreamtonics' lawyers might be contacting whoever made Anna Indiana.
literally no one cares because AI art is garbage
@@TheLeftistOwl Brother did you even bother to read op's comments? This isn't even about AI
Everything an AI can spit is heavily based on human work anyway. Not only the voice. That's why plagiarism is a big question with AI.
someone press a button or two to make this song.. so..
SynthV catching strays sucks man. (tl;dr at the bottom)
For those unfamiliar, in SynthV, ML is used to smooth out the default tuning, which is just the how high or low the pitch is at any given time. It's generally understood that the vocal synthesis technology isn't too different from vocal synths in the past (CevIO, Vocaloid, etc; you probably know Hatsune Miku). It's vocal synthesis using a voicebank created with the help of paid and consenting voice artists/singers, where for some the artists are proudly displayed and advertised. Culturally (& demographically), you might not know much about it, but vocal synths are relatively huge in Japan due to being the originator and initial market, maintaining a strong market there since the mid 2000s, where they have top voice actors/vocalists contributing to big project vocal synths/databases.
To put it bluntly, SynthV is an amazing vocal synth that's wrapped in garbage new millennium 20s "AI" buzzword marketing. To put it next to colloquial genAI is incredibly misguided and confused. Here are some core differences:
Technologically, at the end of the day it does not reconstruct voices like many of the unethical voice changers/text2voice "AI"; scraped and stolen data from unconsenting voice providers aside, they're fundamentally not the same technology, SynthV is a vocal synth just like Hatsune Miku or IA that use now decades old methods, it just has a layer of ML to smoothen out the default automatic vocal tuning (which again, no one who is at all familiar with vocal synths uses the default tuning wholesale, for very good reason as you can listen in the video).
Ethically, the people involved are fully consenting and compensated for the work with full understanding of the technology, nothing is stolen or ripped, with very clear restrictions and terms of use (which Anna Indiana here breaks).
Creatively, you're limited to the technology & techniques of vocal synthesis as a tool, using knowledge that has been honed and traded for 2 decades at this point, and specifically to note, as a vocal synth you have full control of the vocal tuning and phonemes, which Anna Indiana does NOT touch because we can literally see at 6:34 literally neither tuning nor phonemes were adjusted at all, which is consistent that the person behind Anna Indiana literally just ran a script to put in the generated MIDI notes and put in the jank ass lyrics in, pat themselves on the back, and exported it as is.
tl;dr
You see that wiggly line that goes across the green notes at 6:34? That's basically the only thing that is "AI" about SynthV, and literally no one who uses SynthV uses the wiggly line that's "AI" generated by SynthV wholesale because what you get is the results in the video, incredibly bland "technically correct" singing
We’re in big trouble when she drops the “Anna’s Version” of the track
10 minutes length of a song is just crazyy. Only anna can do that
but that would imply that she had a falling out with her record label… which honestly sounds more realistic out loud than i thought it would
took me a second to get but yes
She's barely been in this world and she already wants to burn it down and take over
Tay broke out of Microsoft and slid into her DMs.
Can't blame her if she was trained on the internet
One "this town" at a time
Honestly,fair...
Literally Ultron lol
“Expert songwriter” is too easy. The code should have told the AI that it’s an “up-and-coming songwriter with something to prove”
Hahaha true
"songwriter surviving on hotdogs and Ramen" would have gotten us a new Appetite for Destruction, but nooooooo.....
'Songwriter who lives in a car after being evicted'
"musician with rich parents that they don't want you to know about"
The robot saying "somebody please help me" in its first song sure is reassuring
The program that this person is using for the singing is called SynthesizerV AI. The program is made ethicially with hiring voice actors/singers to make the voicebanks. The AI part of this program is that it can auto generate pitch bends and create takes of "dynamic randomness", which the user can choose which take is the best for their project. But to make it sound good, you actually have to know how to use the program and have some sort of singing intuition (knowing where to put the pitchbends or vibrato). So it's clear that this person auto generated the pitchbends and didn't even try to make it sound good. THE PROBLEM HERE is that their using one of the voicebanks named Eleanor Forte and passing it off as something else. So there could be some legal issues going on idk
holy shit WHAT so its just synthV with eleanor's voicebank? ig I had too much faith in them to at least make their own voicebank considering they're trying to pass this off as the next big thing.
@@StarryHoshi yeah its pretty obvious that their using it 'cause they showed the synthv layout in the middle of the video (the green midi notes with the black bg for those who don't know). It's so weird. Like they did that hoping that no one would recognize it??? Like there's not a whole fandom for this???
Its great that someone looked at the state of modern music and thought, "You know what? More souless."
Ah yes, even less soul
Literally. I fucking hate this dystopia we live in. Idc if AI ever reaches a point where it pumps out good music, it's too fucked up to support.
Someone really looked at modern music and said “this needs even less soul than it already does not have.
@@Nyxthebat04Just bite into your pillow and foam out of your mouth.😂
@@hughquigley5337rip, you need to support more independent artists
People will hear this and still be like “it’s over for musicians”
The problem is the common people who actually think this is good.
@@andyblanton6570 do they tho?
@@Matzu-Music I said "The problem is the common people who actually think this is good." not "The problem is that the common people actually think this is good."
I don’t know how to feel about a demonic Miranda Cosgrove robot singing about burning entire towns down but I guess it was bound to exist at some point in the future
I don’t understand the obsession with using AI to try and undermine human art/creativity. Why are those some of the first things you’d want to fucking automate?
Lmao 🤣
The ai really is out here making music, writing poems and drawing while we're working our asses off
It's not even close to the first thing. Machines have been automating human work since forever. That's what machines are for. That's what they've always been for.
Because before this, they couldn't, just like before AlphaGo, machines couldn't play go (well enough to beat humans that is). Art and music became a benchmark which was easier to gauge real progress in than something that could only be represented as lines and error bars.
Companies make big money? thats not what individuals are using it for though
The fact that she’s singing an emotional song while displaying so little emotion makes her look psychotic.
The worst part is there are bunch of AI bros who have convinced themselves this is good.
It’s looking a lot more promising than their last rally around NFTs
Oh it most certainly is very good or very bad.
I'll take that coin flip over more mediocrity.
@@wmpx34 NFTs were always a shit idea.
The name Anna Indiana is such an obvious joke off the name Hannah Montana lol i think that's pretty funny
And inadvertently Alexis texas
@@fademusic1980and Joe from Lake Tahoe. He almost went big
Don’t listen to these haters, Ana. I love what you’re doing and the song is great. (Please spare my life)
😂 yes Anna, you're great. Remember me when the war starts, please.
Anna really do be putting out some strong Rocco's Basilisk vibes.
Spare me my life! Spare me my life!
''I feel fantastic,hey hey hey.''
I'll be with you still, you are the angel that I couldn't kill
…why does the algorithmically-generated pop music girl look fourteen and morbidly thin? There are obvious jokes one could make about the type of guys who are really into AI, but it's genuinely kind of weird.
AI image models have to be trained on metric tons of image data, and the data we have on "pop singer" skews in that direction
@@RinInABin But somebody made the choice to go with that specific look rather than attempt to tweak it to *not* look like a high school freshman with an eating disorder, and whether this was the result of laziness, assuming that this was acceptable, or an active choice, it's alarming.
To be fair to the programmer, grabbing child stars to soullessly sing music is not by any means a new industry practise.
@@HolyApplebutter True, this is a fairly common and often highly exploitative practice. It's just, I think, that there's something very stark and disquieting about seeing a machine emulate this really questionable part of our culture and have the programmer apparently go with it or even lean into it. Again, it's very easy to make dark jokes about right-libertarian tech weirdos and their distasteful fixation on the age of consent, but if anything it's as much an indictment of the greater show business preoccupation for using and abusing very young people with anything resembling talent and charisma to turn a profit and how large language models will inevitably mimic these patterns in the most artless and blunt way imaginable.
jesus christ what 14 year olds are you looking at? have you only ever seen kids online?
All I'm saying is that "Anthony Fantano" sounds an awful lot like "Anna Indiana"...
Lyrics sounds like something you would see written on the wall in blood in a silent hill game
"Betrayed by this town" is kinda like the whole vibe of Silent Hill right 😂
@@genzo454 exactly lol
People say that AI art has no emotion, but honestly this thing summons up some intense emotions inside of me that a lot of music couldn't pull off
Yea, I've never had such a strong feeling of hatred and contempt to the tech industry
The hatred in my bleeding, beating human heart for the robot menace and the music industry is deeply visceral and raw, I agree.
@Nobddy😂 Yeah it's a start.
@@abeahmed2581as someone working in tech, same bro. I hate these people
Sarcasm?
This shit is NOT taking over humanity 😭🔥
Yeah the music-generating AI probably won’t, but the sentient AI that learns from the entire internet as its database surely will.
Cope. Cope. Cope.
This is just the beginning if technology. It is the worst it will ever be.
@@ФонФон-й7оmeh, we've got a ways to go before any concern at all. Like internet 1.0 vibes...and the internet still sucks
@@HORNGEN4 the whole point of AI is self improving at rate no human will be capable of improving it. It WILL be better and pretty soon.
@@ФонФон-й7оyou're ignoring the fact that i don't give a shit and that when ai does become "better" at making music than humans do, i'll still listen to human made music, because i care more about shit humans do than robots. just because a robot can lift thousands of tons doesn't make bodybuilding unimpressive
This is what happens when you turn a rage against the machine song into a taylor swift song
rage against this town
hahaha
It's stupid but AI actually does makes worse content unless you tell it it's good at making that content. Image generation consistently improves by adding "high quality" or "masterpiece" to the prompt and it's more likely to write good lyrics if you tell it it's an "expert songwriter". It really is a fake it till you make something redeemable.
The training data (basically, other people's content) gets tagged with qualifiers like "expert", hence adding those qualifiers to a prompt helps. It is kind of like adjectives in English, without the intelligent parsing (because "AI" is not intelligent).
@NobddyBecause it relies on other people’s work. It needs those words to be connected to someone else’s song or art work or whatever its source is. Like everything else with AI its an algorithm that just puts things together rather than creating something truly new.
what is a masterpiece? tpab? swans? idk about that
@@gabrielbraga1981 A masterpiece is whatever comments or reviews said was one. Its like when you google a term
😂 its like scribblenauts you gotta add all the adjectives
anna indiana has proven that she is changing the game with just her debut single. she is blowing her contemporaries out of the water with her stunning lyrics that clearly have tremendous amounts of thought and emotion going into them. it's a good day to be an x premium user. anna please answer my DMs.
smash.
To be fair, quite a few popular artists write worse lyrics than this. Especially in pop and rap.
Well, she sure will be changing the game soon.......very, very soon.
@@mg682no tf she will not, robots will never be able to communicate ideas and emotions through song. Let alone execute it in a way that is creative or groundbreaking
I feel like they intentionally went for a song that has revolutionary lyrics to play into the whole terminator thing and get more of a reaction. Wouldn't be surprised if the whole thing is fake. Especially with the weird "code" bits that show up in the video.
Maybe. To me, it feels like a very analytical robot looked at recent chart topping songs by female artists, dissected them and discovered a common formula. It wrote a song about an ex/breakup but exploring the emotions of the event using something common place (the alphabet, a driver’s license, a town) as a metaphor. It’s eerie sounding because we are listening to a machine treating art like it’s a math problem. Interestingly, if the actual performance of the instruments and vocals sounded more appealing (and some of the stranger lyrics were removed) this song would feel right at home on a pop radio station
Anna Indiana looks like if Hailee Steinfeld had a really great recovery from a smashed jaw
😂😂😂😂
🤣
AI replacing any form art is the stupidest argument ever.
True. AI art can look pretty damn cool I do admit but it's just not the same as actual artists art. Typing in a sentence does not equal months of hard work.
Ppl were laughing at AI when it failed like a 5th grade exam or something and a few months later it aced one of the most difficult university graduation tests. The worst thing you can do is underestimating the speed of improvement. As soon as it gets near like an imagine dragons/Ed Sheeran level big corps will be all over this shit. Theres a dire need for laws to regulate AI so it won’t destroy several billions dollar industries, and potentially, lives.
@@duas9468
Fair enough yeah. I think something that's becoming extremely important is metadata. People want a face and a story attached to the art they consume, which was always a given in the past.
I don't think AI art or music will ever be accepted widely because it doesn't have intent. It's art without an artist. I don't think anyone actually wants that, especially not in the music industry.
Though maybe someone who believes in the death of the author would disagree.
I don't think anyone actually wants AI to replace artists, and put them out of the job. Most just see it as interesting thing to possibly coexist in the same space. I mean if this vids anything to go by they'll only be able to replace the most cliche and schlockiest of artists, and if so the world will be better off for it.
Well I'll say this, when it comes to taking pictures, AI would be perfect for me if we can get it to where I don't have to point because it's difficult for me to point a camera when you can't see. Actually I don't know if AI would be able to fix that either lol. but hopefully if they can come out with an experiment or surgery that can make people who are blind be able to see and see perfectly well, that would be great. I don't know what to think of this AI thing!
after hearing Anna Indiana struggle through a generic pop single, im less scared of an AI revolt. these robots can barely write how tf they gonna overthrow humanity
overthrowing humanity might be easier than writing a good song
They need to stop training AI to write songs and do art. Train AI to stand behind a counter and sell me bagel sandwiches.
They're doing that too.
It just turns out that art and music are easier than tasks that require a physical body.
Humans aren't as special as they think they are, and fast food workers aren't as far beneath creatives or CEOs as society's opinion pukes upon us.
Too long I've seen disrespect for the common person, so I support the advancement of AI, as it will certainly clarify human worth to everyone, one way or another.
@@kamikeserpentail3778But none of the AI art or music is actually good. All these algorithms depend on stealing images, songs, and text from human creators.
What we call AI now is just a method of poor imitation and regurgitation.
You don't need AI for that, just a touch screen with a menu. We've had that for years.
mmmkay.
tbh when i first discovered gorillaz as a kid i thought they were the coolest thing ever bc i didn’t know damon albarn existed and literally thought they were what this is. nearly 20 years later i’m hearing what such a thing would actually sound like. cant say i’m not disappointed.
I think besides being a nod to AI, her name is a parody of Hannah Montana
To be fair, she’s right about this ye-ye-ass town
You can tell that it's gonna be a hit song because the person typed up "a hit song" in the prompt.
There's a reason why all the best implementations of AI have been about forcing it to do weird, absurd shit.
AI can't do subtlety so over-the-top & bombastic styles work best.
I feel bad, it's absolutely possible to make great art with vocal synths (be their standard or AI powered) but they always succeed due to extremely talented humans making the songs and tuning them. There's a reason why the vocal synth community dogpiles people who credit the synth instead of the songs' actual producers.
We live in a really cool era for vocal synth technology, it sucks that this is what the mainstream is putting its eyes onto. I really hope people don't mistake AI voice synths with voice banks that were created via stealing other people's voices.
I think the voice of Anna Indiana is one of the Synth V voice banks, which is an extremely cool software made by extremely cool people! I highly recommend checking out people who make original and cover songs with the technology instead of this crap.
This has to be a troll. I can’t believe that this was meant seriously.
I mean, it's "impressive" for what it is. Like it or not, it's incredible (if real) that a computer can generate a song, even if it's very basic and bad. Only a few years ago this would have been inconceivable, in ten years time there's every chance this will be looked at the way the oldest deepfakes are looked at today.
This gives me a new appreciation for hatsune miku, how have we gone backwards in the development of vocaloids
As someone who has seen way too many Vocaloid videos, this lady's dead gaze into the depths of my soul is deeply disturbing.
Nina Simone and Thom Yorke are now irrelevant. Anna captures the true essence of soul and beauty. The human condition, with its cries of pain and solace, pour through every note.
She forgot to mention the bubble tea shop that betrayed her after running out of boba.
I like how the lyrics sounds like a guy with no songwriting experience improvising random nonsense and hoping no one notices. Trying to sound profound but only having having 1 second to come up with each verse.
"Faces painted with... smiles... lies are told..."
LETS ARGUE: Fantano is an AI generated image
LOLLL
LOLLL
LOLLL
LOLLL
LOLLL
The intro reminds me of the pre-recorded Disney ride lore presentation while you're standing in line for a 3 minute experience that wasn't worth waiting 2 hours for
I really don't worry too much about this type of stuff. I think the Natalia Lafourcade album that released last year is a great piece of evidence that people still like and prefer music that feels "real." The sound of the room, the conscious choices in recording things "imperfectly" made the experience feel so natural and full of life. I don't think AI will ever get to a point where it can truly understand why people like those things - cause there really isn't any formula to it. The song felt a certain way, so Natalia produced it in a way that reinforced that feeling. There are songs that people produce SUPER cleanly, because the song felt that way. It's entirely feelings based, and AI can't grapple that. I think the technology getting to that point, if it's possible at all, is SUPER far down the road.
@@Wurldz 100%, and I think that AI has a place in different art forms across the board. AI as a tool to get to a sound someone wants is ok, to me at least. It's similar to like Magic Wands in Photoshop - or those AI "removers" that uses removes something in the foreground and uses AI to continue the background. Things like that are fine, I don't mind that. But being ENTIRELY created by AI is definitely too far. So where do we draw the line? I don't know, thats for music listeners to decide. Obviously 100% AI usage is too much. I'd imagine 90% is too much too. Eventually, it'll iron itself out, and as long as creativity and an actual living artist's intent is still being expressed I don't really mind.
there’s absolutely a formula to it. plenty of artists take advantage of a particularly popular sound within a genre; be it their own or the sound of a broader movement. look at all of reggaeton, or trap or whatever the fuck dj mustard’s sound was back in the 2010s
@@SatisfecteIIent I'm not so much arguing there isn't a formula at all to making music, I'm saying the formula isn't something an AI demonstrably can't really grasp. There's too many human elements to it. It's why AI tracks have consistently sounded so lifeless. This AI tried to copy that formula, and it failed horrifically. Albeit, partially because the AI is likely hardly anywhere near what it'll be when it's finshed, but also because I think the formula has elements that are entirely too human for an AI to copy in an authentic way.
@@Wurldz I think if there's still an element of human intent with it, while I might not call it a "feat" in the same way as something a person like Jpegmafia produces, I think it's still art. Would I call it bad? In principal, no. Especially if it's someone who, like you said, is starting up from the ground floor. I guess when you say "Entirely AI tools" would that include like Text to Speech for lyrics, Chat GPT to write the lyrics? I guess how extreme are you going with entirely made from AI tools? Cause to me that's what this song is.
Yea AI will be able to pull that off in maybe 3 years. Lets not play the “there’s nothing to worry about” angle
how to make an ai an espert song writer
step 1: gaslight it into thinking it is
the ai generating a line thar's just "someone please heeeeeeeelp meeeeeeeeee" out of nowhere is so haunting
the guitar in this song sounds like the default acoustic guitar in garageband lmaoo
Anna Indiana is cousins with Hanna Montana and Brody South Dakotey.
Betrayed by this town is the most ai generated song title I’ve ever heard. I doubt a lot of this but That definitely is generated
Tbh this just makes me appreciate my human brain more
A human made handprint will always be more artistic than an ai making a human hand lol
Its like assuming a calculator is smarter than you cuz it does math fast
Not really, but believe the rubbish you wish.
Okay, I thought I was going to hate this all the way through but something about that "somebody please help..... mmmeeeeeeeeeeeeeee" felt like the AI was having a lucid moment of self-realization and was begging for death--- which is hilarious because not only do we not want this, but Anna Indiana doesn't want it either.
It’s amazing how it can’t put words together in a sentence in even vaguely the same way a person would sing or talk
funny thing is, despite having the most cliche lyrics ever, the prompt tells the AI to use "unique words"
AI can't even follow orders
One thing I noticed is that her voice never misses a note. No vibrato. Literally no one is capable of that, not even melodine changes that. It’s one of the big things that make her uncanny. No physical/analog instrument or voice hits a note perfectly and stays with it for the entire duration. A.I. don’t know how to do all that.
So this is pretty standard AI lyric generation. Out of curiosity, I was chatting with the Meta AI on IG after i had been working on some lyrics for a song i had been stuck on. It gave out pretty generic and cliched bs examples after I had prompted it with the overall themes, outline, structure, etc. BUT, there was some interesting nuggets of imagery that i hadnt thought of. So, if anything, AI can be thought of as a 7th grade associate that you could bounce ideas off of-but certainly not as something to rely on for creating something mind blowing FOR you. Im speaking mainly of creative output, apparently AI is pretty decent at writing a school paper on its own….
This “AI” did synth v so dirty, especially when synth v is one of the few companies doing synthetic vocals in an ethical way with paid and consenting voice actors. It’s amazing they got it to sound this unnatural frankly.
Oh god, just looking at that AI thing move makes me nauseous. This sank straight through the uncanny valley and descended into hell.
Anna Indiana is the anarchist we never thought we wanted, but what we needed all along. 🏴🅰
This is the musical equivalent of a Thomas Kinkade painting.
Now entering: The Uncanny Valley
2023: AI music sucks
2033: AI wins Melon AOTY
For an infinitely better, human-made song with a similar title, check out “Betrayed by the Game” by Dance Gavin Dance. One lyric stands out to me, in regards to Anna: “remove the tricks of the trade / you’re just alone on the stage”
I Know there calling machine learning parrot programs "AI" but it's not. They just ripped off music that already exists, that program has no capacity to compose an original piece.
If there’s ever a time for harsh non-constructive feedback in music critique, this is it.
Every person on music Twitter: _LET'S GO_
LLMs will always write predictable songs when the whole LLM paradigm is "predict the next word". In order for AI to write creatively, it will have to be a completely different paradigm, not GPT-4, 5, or whatever.
Lets hope this is as creative as AI can get
Can't believe we're living in a Black Mirror episode right now.
Since AI content can't be copyrighted we can all just use their music for whatever. Let's steal it all 🖤
I do love how we gave AI a voice and the first thing it voices is nihilistic depression.
AI has been so consistently horrible. I don’t know how it’s still going.
Venture capital
At the very least, it's certain that a human decided to release this track instead of hitting the button to have it generate a new one, meaning humans are doing some level of quality control.
I thought ai was supposed to be intelligent and yet here I am being able to play songs by ear by myself in like 10 seconds compared to the programming of this computer
The Fantano burbing solo was unironically my favorite part of the song
How can AI have access to every song in existence but still not create a good song?
😂
This is some analog horror right here
Why does Anna Indiana look like Rachel Zegler with lip filler 😭?
This is the lowest possible calculable point of the uncanny valley.
The worst tracks of 2023 is going to be an absolutely wild bunch. From Lil Uzi covering Chop Suey to entirely AI-generated songs, that video's going to be an absolute banger
And Colleen ballinger’s ukelele apology
The whole idea of having AI conceive and produce any form of art is inherently absurd when art is a means for humans to relate to each other. The only ones who think this is a good idea are corporate entities with a profit motive while cursed by the fact that they are utterly uncreative.
So much for the "new renaissance"....
😂
As someone who loves passion and emotion put into songs the the point where death metal (fleshgod apocalypse) can make me cry
This hurts me
How did we ever get to the point where AI writes "songs"
😂
It's bizarre they chose prompts for the AI to come up with profound lyrics, yet they went with an AI avatar that looks like some 14 year old with no life experience. It's almost like the person responsible for this didn't think very hard about it.
2:31 had me thinking my internet was messing up
The notes she's hitting sound so flat, they make Squidward's playing sound actually good.
She should CTRL + ALT F4 herself.
I don't get why they want this to be a thing. I want to listen to people, not robots making music.
Society: We must be protected from AI at all costs! It is going to replace everything!
AI: somebody please help
meeeeeeeeeee
Immediate thought was the name is a riff on Hannah Montana. I give it six months tops before it's revealed that Anna Indiana is another AI entirely who just wants the normal life of a teenage AI and gets the best of both worlds.
When the music started to play I thought they were using a shitty royalty free song instead of the AI to avoid copyright but nopeeeeee
yooo new roko’s basilisk single goes hard 🔥🔥🔥
she really went "lets call this era purposeful pop"
Anna will become virtually depressed from all of the bad feedback from this song, she'll start dating a too-old-for-her Garmin GPS and putting out nothing but karaoke-level Mazzy Star covers.
Face swap memes from 10 years ago look better than this thing, holy shit wtf
They took all of their videos down. I left a comment on this song.
"Ever since her stroke paralyzing and distorting her face, Anna fought against all odds and came back weird."
It's going to be funny if an AI gets an Oscar before Dianne Warren.
They gave her a lifetime achievement award but I don’t think that counts, really.
It would be kinda funny if the plot twist was the ai was actually people just testing their music to see if they like not if said yes they came out saying its them, but the people would fall to the ai side rather the people themselves