Let’s hope not, but they seem to be able to exploit everything so this probably wouldn’t be too different. This is where open source models will be super important
Its about what you can develop to the average listener. And this stuff often sounds better and more interested than what charts todcay. Especially when your combine genres. Plus its a samplers dream come true. No lawyers to worry about.
I was also thinking about that idea where you have the seed for an A.I. vocalist to make multiple tracks with. I think the challenge for Udio with that is if someone generates a voice that's close to or indistinguishable from a real artist, and then passes off a single or album in their name. I suppose they'll need some kind of A.I. detection system that automatically deletes a seed that sounds too much like a real pop/rock/country singer.
I actually love the song you generated. Such a powerful melody, especially the vocal run at the end of the chorus, it has been stuck in my head for days. I am actually keen for ai generated music as i have i hard time finding music that i really like and now i can just generate to my taste. I make music too but listening to your own songs can get old really fast 😂.
Thanks 🙏🏻 I think it almost feels like magic when you get a good result from udio! Maybe we’ll get to the point where we have a single prompt like make me a lo-fi playlist and it will just do a whole playlist for you. That would be interesting
@@promptgeek I reckon that in the future you could just thumbs up and down on auto generations and the ai can just learn what you enjoy over time and automatically generate to your preference. Or imagine even having a device that tracks your mood and just generates you music to match. I don't know what the future or music is gonna be like but it's definately going to be interesting.
What this is is a hip hop producers wet dream! Just like Udio, this is another great sampler. This could actually revitalize rhe original hip hop genre from the garbage fire it is now. What I've personally been doing is downloading udio tracks, loading them into Audacity, then using the openVINO AI plugin I separate the instruments into stems and load them into ableton live and chop up the beat and melodies. If you do this you have to remember ro delete the track feom the udio website or else you run the risk of someone else finding and remixing something you made. Peace and love, creators!!
The idea that AI music generators will kill the music industry is ridiculous. To start, with the music licensing business, it is much easier and faster to search and license music from a muisic library than it is to build a song that fits your needs with a generator for multiple reasons people who use generators probably know. Second, people in the industry from songwriters to performers can use generators for ideas if they wish to compose music faster. Generators do not instantly spit out perfect songs everyone wants to license or listen to. Few people in the music industry actually have anything to worry about except hype.
Spot on, it’s just too much hassle for the average person right now but we could get to the point where it becomes really easy and is as simple as asking for a playlist. Or an Ai that can watch your video clip and then create the music to match the pace and emotion of your film! It will all be about ease of use if it wants to be a real threat but let’s be honest most people are too lazy to learn these tools
Comments about soul looks very stupid. But if you believe in magic then magic is for sure the best protection from robots. Even Alan Turing in his paper about "Turing test" listed ESP as the strongest argument against possibility of creating AI.
i dont really understand in what way this is better than Udio??? Watch my 3 last videos... 1 is a 80's rock song made with udio, the other 2 are pop songs made with udio.
@@HasimFN You probably just dont know how to use it properly or dont know any good prompting method for it... Udio give the most professionnal result with the best sound quality.
The big thing I saw with Udio is what is did with real songs such as Amazing Grace and House of the Rising Sun. People can take public domain songs and make them "fresh."
Pop music now a days does not set a very high bar. In the same way that Ai diffusion art generators simply cant create true artistic masterpieces, I believe music AI will never be able to make true amazing music. All these songs are passable in pop sure, but none of them have any soul. I think this is a net positive for the music industry because it will force human producers to actually make good music again.
@@promptgeek Yes, that's an excellent point. The reason computers (AI) can now generate music that sounds exactly like human-made efforts, is because humans have recently degraded to writing songs like computers! With this I mean, that actual melodies, musical theory, rules, concepts and expression of emotions tend to be ignored, in favor of "cool riffs", "rad beats" and/or provocative lyrics, mostly rapped. These are all areas where computers excel, so you could say, that this new AI is forcing songwriters to start getting their acts together... or lose their jobs. If you do a sloppy job, an AI will be able to do it just as well, and for a fraction of the cost.
It won’t kill the industry but definitely not for the reason given. A lot of people do think AI sounds good and is capable of creativity and it’s the same for AI art. Not every musician has the same level of talent and many generators are not as good as Udio. A lot of musicans have little talent and cannot produce anything of global value manually even though they think their work is good and some people might like it. The reason it won’t kill anything is that speed of creation is only one factor. It doesn’t mean everything. People will still listen to human created music, license it, go to performances, etc. because they like the artist and their music.
Yeah you are spot on! There will always be people who want the experience of live music and to connect with an artist in some way. The AI music might saturate the market and make finding real artists difficult but I think very talented artists will still get attention and fans
@@promptgeek Thanks for posting your opinion on that. My comment also asks the question, I guess, "What do we really know about these programs?" How much of the program needs to utilize Ai in order for it to be an Ai program doing what we all believe it is doing? Should we be suspicious?
Do you mean how much is actually someone else’s actual voice? How much is just algorithmic rather than AI? It does certainly raise questions when you look it it that way.
@@promptgeek If you and I actually identified patterns of notes (phrases) commonly used by Bowie and Bono, does it mean the program 'isolated' these two artists? If so, is this worth looking at to decide if this is infringement? Is Ai supposed to narrowly isolate an artist's musical identity? I didn't think so. I thought this was a routine looking for the next most logical note to use. (whatever that means with music)
Yeah I think you are right, if they were to sue and ask for the dataset as part of discovery if their voices are in that training data then Udio would be toast. Would I then be a responsible party if I had monetised the song? The legal minefield is real here
Is it very good AI music generation? Yes. Is it very good music? Not really. Like their human counterparts, AI generated songs will be mostly run of the mill and some will be gems. I always use my own lyrics, so right there, it is pretty much a human-dominated creative process. My unique words will produce my own unique song. Words are prompts in themselves, the AI has to generate music in line with the meaning and syllables of the lyrics it's fed; just like Elton John had to dovetail his music to suit and follow the lyrics Bernie Taupin fed to him.
It’s a really good point, the music industry is very saturated as it is with midi driven stuff. At least with the ai the instruments sound much more real.
I think for now we can be maybe fooled but if we don’t have creative people putting stuff out there the ai feedback loop will ultimately make everything sound the same
I agree, I can see this being used for video background tracks, some movie soundtracks, and even lots of electronic club dance music. But I think a lot of people want to follow an artist or a band and see them perform live and this will never be replaced.
This is crap. For the casual music listener, nothing is expected of you so differentiating between this and real artists won’t matter, however lovers of music cultures and avid participants of music cultures will definitely know the difference. However if you think the music industry or any industry will stop, they will not and it won’t be alright. Have you seen the movie Repo Men? Time? Demolition Man? Terminator? transcendence? Or any tech takover movie? Corporations will own you and the government will be their hired security. But anyway, it’s not expected for a $23B perspective, it’s not expected because you technically are not in the space.
Nonsense. You only think it's crap, because you know it's AI, and it's the early days for this stuff. But this is as bad as it's ever going to be, from this point forward. Give it another year or so, compared to where it was last year. It'll be indistinguishable from music made the old fashion way, or maybe even better. I still think there's a good chance we'll see an AI song hit the top charts by the end of the year. Alternatively, I think a traditional artist may release an AI created song, but not tell anyone, until after it becomes popular. lol
Finally some music that you can post on YT without it being taken down 💁♀️
That’s actually one of the strongest reasons to use ai music! No copyright claims!
Let’s hope not, but they seem to be able to exploit everything so this probably wouldn’t be too different. This is where open source models will be super important
This is scary because I was getting into the song!
Its about what you can develop to the average listener. And this stuff often sounds better and more interested than what charts todcay. Especially when your combine genres. Plus its a samplers dream come true. No lawyers to worry about.
I was also thinking about that idea where you have the seed for an A.I. vocalist to make multiple tracks with. I think the challenge for Udio with that is if someone generates a voice that's close to or indistinguishable from a real artist, and then passes off a single or album in their name. I suppose they'll need some kind of A.I. detection system that automatically deletes a seed that sounds too much like a real pop/rock/country singer.
It’s going to be a legal minefield 🫣
I actually love the song you generated. Such a powerful melody, especially the vocal run at the end of the chorus, it has been stuck in my head for days.
I am actually keen for ai generated music as i have i hard time finding music that i really like and now i can just generate to my taste. I make music too but listening to your own songs can get old really fast 😂.
Thanks 🙏🏻 I think it almost feels like magic when you get a good result from udio! Maybe we’ll get to the point where we have a single prompt like make me a lo-fi playlist and it will just do a whole playlist for you. That would be interesting
@@promptgeek I reckon that in the future you could just thumbs up and down on auto generations and the ai can just learn what you enjoy over time and automatically generate to your preference. Or imagine even having a device that tracks your mood and just generates you music to match. I don't know what the future or music is gonna be like but it's definately going to be interesting.
I love that idea though! It would be like a life soundtrack
What this is is a hip hop producers wet dream! Just like Udio, this is another great sampler. This could actually revitalize rhe original hip hop genre from the garbage fire it is now. What I've personally been doing is downloading udio tracks, loading them into Audacity, then using the openVINO AI plugin I separate the instruments into stems and load them into ableton live and chop up the beat and melodies. If you do this you have to remember ro delete the track feom the udio website or else you run the risk of someone else finding and remixing something you made. Peace and love, creators!!
Oh that’s damn smart
Normie listens to song: "who is this?"
"ElevenLabs!"
"No who sings it?"
"No one!"
😂 then watch their heads 🤯🤯
udio does better at getting the emotion of the vocals through I think, and emulating artist and vocalist is super easy on udio
I certainly feel that too, I was going for a pulp style song in this video and I think udio really nailed it
It doesnt lol udio voices are really bad
The idea that AI music generators will kill the music industry is ridiculous. To start, with the music licensing business, it is much easier and faster to search and license music from a muisic library than it is to build a song that fits your needs with a generator for multiple reasons people who use generators probably know. Second, people in the industry from songwriters to performers can use generators for ideas if they wish to compose music faster. Generators do not instantly spit out perfect songs everyone wants to license or listen to. Few people in the music industry actually have anything to worry about except hype.
Spot on, it’s just too much hassle for the average person right now but we could get to the point where it becomes really easy and is as simple as asking for a playlist. Or an Ai that can watch your video clip and then create the music to match the pace and emotion of your film! It will all be about ease of use if it wants to be a real threat but let’s be honest most people are too lazy to learn these tools
Comments about soul looks very stupid. But if you believe in magic then magic is for sure the best protection from robots. Even Alan Turing in his paper about "Turing test" listed ESP as the strongest argument against possibility of creating AI.
I had to read up about it after seeing your comment earlier. Fascinating that he had to eliminate esp as a factor in his tests
i dont really understand in what way this is better than Udio??? Watch my 3 last videos... 1 is a 80's rock song made with udio, the other 2 are pop songs made with udio.
Bro udio is the most shit of them all
@@HasimFN You probably just dont know how to use it properly or dont know any good prompting method for it... Udio give the most professionnal result with the best sound quality.
@@FulguroGeek i do know how to use it. I made a lot with suno. I tried udio bit quality sounds bad
@@HasimFN does suno give the wav file format now? instead of just mp3
@@FulguroGeek yes
oh the irony, when everyone thought that art and creativity would be the last AI frontier, while this is the first thing it broke apart
The more we see from Ai the more I’m leaning towards simulation theory
The big thing I saw with Udio is what is did with real songs such as Amazing Grace and House of the Rising Sun. People can take public domain songs and make them "fresh."
I hadn’t seen that but I’m going to play with that idea now thanks!
@@promptgeek9
Pop music now a days does not set a very high bar. In the same way that Ai diffusion art generators simply cant create true artistic masterpieces, I believe music AI will never be able to make true amazing music. All these songs are passable in pop sure, but none of them have any soul. I think this is a net positive for the music industry because it will force human producers to actually make good music again.
I love this angle on it!
@@promptgeek Yes, that's an excellent point. The reason computers (AI) can now generate music that sounds exactly like human-made efforts, is because humans have recently degraded to writing songs like computers! With this I mean, that actual melodies, musical theory, rules, concepts and expression of emotions tend to be ignored, in favor of "cool riffs", "rad beats" and/or provocative lyrics, mostly rapped. These are all areas where computers excel, so you could say, that this new AI is forcing songwriters to start getting their acts together... or lose their jobs. If you do a sloppy job, an AI will be able to do it just as well, and for a fraction of the cost.
It won’t kill the industry but definitely not for the reason given. A lot of people do think AI sounds good and is capable of creativity and it’s the same for AI art. Not every musician has the same level of talent and many generators are not as good as Udio. A lot of musicans have little talent and cannot produce anything of global value manually even though they think their work is good and some people might like it. The reason it won’t kill anything is that speed of creation is only one factor. It doesn’t mean everything. People will still listen to human created music, license it, go to performances, etc. because they like the artist and their music.
Yeah you are spot on! There will always be people who want the experience of live music and to connect with an artist in some way. The AI music might saturate the market and make finding real artists difficult but I think very talented artists will still get attention and fans
6:20 David Bowie & Bono.
I can hear that too now you mention them
@@promptgeek Thanks for posting your opinion on that. My comment also asks the question, I guess, "What do we really know about these programs?" How much of the program needs to utilize Ai in order for it to be an Ai program doing what we all believe it is doing? Should we be suspicious?
Do you mean how much is actually someone else’s actual voice? How much is just algorithmic rather than AI? It does certainly raise questions when you look it it that way.
@@promptgeek If you and I actually identified patterns of notes (phrases) commonly used by Bowie and Bono, does it mean the program 'isolated' these two artists? If so, is this worth looking at to decide if this is infringement? Is Ai supposed to narrowly isolate an artist's musical identity? I didn't think so. I thought this was a routine looking for the next most logical note to use. (whatever that means with music)
Yeah I think you are right, if they were to sue and ask for the dataset as part of discovery if their voices are in that training data then Udio would be toast. Would I then be a responsible party if I had monetised the song? The legal minefield is real here
Is it very good AI music generation? Yes. Is it very good music? Not really. Like their human counterparts, AI generated songs will be mostly run of the mill and some will be gems. I always use my own lyrics, so right there, it is pretty much a human-dominated creative process. My unique words will produce my own unique song. Words are prompts in themselves, the AI has to generate music in line with the meaning and syllables of the lyrics it's fed; just like Elton John had to dovetail his music to suit and follow the lyrics Bernie Taupin fed to him.
That’s a very important point about the lyrics!
The music industry already flooded with muzak, aliexpress produced crap, don't matter no more.
It’s a really good point, the music industry is very saturated as it is with midi driven stuff. At least with the ai the instruments sound much more real.
Kind of ironic really!
A constant that no AI will ever change and that true art needs is soul. And that cannot be simulated.
I think for now we can be maybe fooled but if we don’t have creative people putting stuff out there the ai feedback loop will ultimately make everything sound the same
@@promptgeek But as with samplers back in the days. If a tool is used in the right creative way.. The output becomes art.
I agree, I can see this being used for video background tracks, some movie soundtracks, and even lots of electronic club dance music. But I think a lot of people want to follow an artist or a band and see them perform live and this will never be replaced.
I believe suno is much better
Hell no
@@Insertrandomnamehere412yes
The video narrator is mispronouncing the word music. I mean, you need to fix that. It sounds like it is saying moo-zik. Really bad
In this case, you should your own voice, not an AI voice.
This is crap. For the casual music listener, nothing is expected of you so differentiating between this and real artists won’t matter, however lovers of music cultures and avid participants of music cultures will definitely know the difference. However if you think the music industry or any industry will stop, they will not and it won’t be alright. Have you seen the movie Repo Men? Time? Demolition Man? Terminator? transcendence? Or any tech takover movie? Corporations will own you and the government will be their hired security. But anyway, it’s not expected for a $23B perspective, it’s not expected because you technically are not in the space.
You may be right, we’re being seduced by the novelty of this but without considering the dangers
Nonsense. You only think it's crap, because you know it's AI, and it's the early days for this stuff. But this is as bad as it's ever going to be, from this point forward. Give it another year or so, compared to where it was last year. It'll be indistinguishable from music made the old fashion way, or maybe even better.
I still think there's a good chance we'll see an AI song hit the top charts by the end of the year. Alternatively, I think a traditional artist may release an AI created song, but not tell anyone, until after it becomes popular. lol