Posted this comment on Music Lawyer's vid. Think it fits here, too. Just a little about my experience and interaction with Udio AI. *This is long, but broken into short numbered paragraphs. 1) The music industry (business, legal, corporate) has been and is still a heaping pile of rotting shit. This includes streaming services, companies that own artists works, distribution, etc. 2) Music AI companies used, without permission, hundreds of thousands of artists works to train the tech. Just as all other AI's used art work, novels, poems, etc to train their AI. 2a) I've done over 30,000 hrs of writing, somewhere in that, AI probably scraped my stuff for training (hope it wasn't my early stuff, cause that shit was terrible). 3) The people using music AI and only putting in prompts and doing some editing, but then calling the music their own and considering themselves musicians and music makers are, sorry, they are delusional. There is nothing wrong with making songs this way with AI. Enjoy it, have fun, but FFS be honest about what you provided in the process. 4) Music AI is used by a huge percentage of producers, musicians, writers, mixers. That's just a fact. There is an amazing amount to learn from using music AI. The instant marriage of lyrics, instrumentation, genre, emotion opens up worlds of understanding if you are paying attention. 5) I've used music AI (I'm not a musician). It wasn't my first choice. As a serious author with multiple novels, lots of short stories and 6-800 songs/poems, I spent two years talking with, listening to and trying to get singers/musicians to work with some of my stuff. Great experience, but no takers. 5a) I'm older, my health is shit, and I wanted to hear some of my songs with instrumentation-- instead of just the hundreds of my poorly sung A Cappella versions. So, I did about fifty of my songs with AI over the last eight months. I've recently handed off some Lyrics/A Cappellas to some young artists and have mostly stopped the AI stuff, for now. 6) My experience with music AI falls in a grey area. I upload 20-30 seconds of my copyrighted lyrics/melodies via A Cappellas to Udio, and then I generate until the software spits back an approximation of my A Cappella's melody (It also almost perfectly clones your voice). At that point I build out the song. In total, using this approach typically takes me twenty four to forty eight hours to finish a song. (that includes dropping the stems into Garage Band and doing some amateurish mastering). 7) I've spoken online with multiple musicians/mixers who drop their clips into Udio and generate from that. All of them have differing amounts of input to the process. Some end up doing their own singing later. Some use AI to come up with ideas and run with them in a DAW. It really runs the gamut. 8) Where these "hybrid" uses of AI fall legally, I have no idea. I do know that coming up with your own melody, whether instrumental or lyrics, and then working with that in conjunction with the AI is difficult, but very rewarding when the collaboration gets your input and enhances it in a cool way. 9) All that said, I'd still rather be working with live musicians/singers. If you want to listen to some of what I've done with my vocals, lyrics, melodies in conjunction with Udio, you can click my avatar and go to my channel. *Edit* I should add, I've left a number of my original, copyrighted A Cappellas up on my channel. If you're inclined, you can listen to A Cappellas and the versions with music to see what working this way with music AI produces.
At this point, I'm a "former" musician as I stopped writing sync a few years ago after about 20 years because of how rapidly it was declining following the rise of subscription models and so many bands, singer/songwriters and hobbyists entering the field. While I still play and write traditionally, I've thrown myself 100% into AI generated music because the potential is amazing. I think many musicians who are resisting this are eventually going to miss out on some new opportunities. Production companies, ad agencies, and other media outlets using a ton of music for their projects aren't going to have an intern with no musical knowledge to haphazardly generate music on a regular subscription. They're going have an in-house workflow and hire a musician or a music contactor (freelance or small business) who know trends, understands song structure, and are familiar with the traditional side of scoring for visual media and music supervision. I see plenty of business opportunities here that will be far more lucrative than writing sync and doing custom composition here and there. I think Shulman is right on the mark with his observations. Once music is your job, yes, it can be laborious as well as frustrating composing music for clients who want something revised over and over again. It's also not as profitable to only compose what you like. You often have to put something together you think is one step above "this sucks" because it's trending or that's what a client or library wants. You have to keep up on software, VSTi's, etc. And I think he's right about AI music being a community builder for both non-musicians and musicians. Both Udio and Suno have pretty active forums, charts, contests, etc. This is a good thing. It's fun, social, good for networking, etc. Is this going to one day end musicians? No, not at all. Besides the new opportunities, people will continue to perform for the foreseeable future. Might it end the record industry? I think we can see that they're trying their best to make sure that doesn't happen and I doubt it will. They will probably own a large share of the AI generator market because they have the best training data in massive supply, expert music promotion with large budgets, and their artists perform to huge audiences. I don't think they're afraid of this at all. They're just trying to force a settlement that hands over some of Udio and Suno's revenue.
This is the guy that has learning disabilities. It is excellent for me because I write Lyrics. I have written over 2,000 songs, this is since I have been 17 I am now 58, so I am making full blown songs out of everything I wrote. My Cousin Willa, plays the harp, violin, Piano, Bass, Drums, keyboards and Ukulele, and many more instruments her mother forced lesson's, she hates playing music because of this. But this is a dream for me. When I told a friend that I was using "Band in a Box" my friend could not wrap his head around that. He spent many years learning his craft, and he told me he would never write music with me again, he never did. This was in 2012, so yes we never wrote any songs together again. He would not even get together to have coffee and even discuss music again. I ran into him and I told him about the latest technology, and he quite even talking to me about music now. He blocked me on the phone, social media, and even e-mail. I ran into him in Wall-Mart and he passed me up like a dirty shirt. Would not even speak to me, I ran into him at another place and he avoided me like the plague.
Thanks for chiming in, and thanks again for watching and commenting! Your 'friend' sounds like a jerk, his loss, not yours. The fact that you can get the songs out of your head into a musical form on your own is amazing! I bet Shulman would love to hear your story, because this is the sort of thing that makes tech evangelists and creators happy. I don't really care what the haters think, lots of people can't ingest opinion or ideas other than their own, and thats their problem, not ours. Keep on making music!
@@YoPaulieMusic I will, that is gratifying to be able to put all the lyrics to music after all those years of them just being Lyrics. And I can come up with my own tunes, and sing them into the Laptop, and it is 100% on in what I am trying to get out of my head. How can I send you a sample of my song called "You Should Have Been Mine"? I wrote the first version with "Band in a Box", then I put a 2 minute sample and it made my song a whole bunch better. One thing I can do if you will allow this, But I can also put both versions on Unlisted and if I send you the link you will be able to listen to it. I am not sure weather you will allow it for you to listen to, or how an I send you a link for you to listen to. The 2 different version are pretty close to the same, but the one made by suno opened up the song to it's full potential.
Great vid. Love that you take the time to do all the background work. Thanks!
Posted this comment on Music Lawyer's vid. Think it fits here, too. Just a little about my experience and interaction with Udio AI.
*This is long, but broken into short numbered paragraphs.
1) The music industry (business, legal, corporate) has been and is still a heaping pile of rotting shit. This includes streaming services, companies that own artists works, distribution, etc.
2) Music AI companies used, without permission, hundreds of thousands of artists works to train the tech. Just as all other AI's used art work, novels, poems, etc to train their AI.
2a) I've done over 30,000 hrs of writing, somewhere in that, AI probably scraped my stuff for training (hope it wasn't my early stuff, cause that shit was terrible).
3) The people using music AI and only putting in prompts and doing some editing, but then calling the music their own and considering themselves musicians and music makers are, sorry, they are delusional. There is nothing wrong with making songs this way with AI. Enjoy it, have fun, but FFS be honest about what you provided in the process.
4) Music AI is used by a huge percentage of producers, musicians, writers, mixers. That's just a fact. There is an amazing amount to learn from using music AI. The instant marriage of lyrics, instrumentation, genre, emotion opens up worlds of understanding if you are paying attention.
5) I've used music AI (I'm not a musician). It wasn't my first choice. As a serious author with multiple novels, lots of short stories and 6-800 songs/poems, I spent two years talking with, listening to and trying to get singers/musicians to work with some of my stuff. Great experience, but no takers.
5a) I'm older, my health is shit, and I wanted to hear some of my songs with instrumentation-- instead of just the hundreds of my poorly sung A Cappella versions. So, I did about fifty of my songs with AI over the last eight months. I've recently handed off some Lyrics/A Cappellas to some young artists and have mostly stopped the AI stuff, for now.
6) My experience with music AI falls in a grey area. I upload 20-30 seconds of my copyrighted lyrics/melodies via A Cappellas to Udio, and then I generate until the software spits back an approximation of my A Cappella's melody (It also almost perfectly clones your voice). At that point I build out the song. In total, using this approach typically takes me twenty four to forty eight hours to finish a song. (that includes dropping the stems into Garage Band and doing some amateurish mastering).
7) I've spoken online with multiple musicians/mixers who drop their clips into Udio and generate from that. All of them have differing amounts of input to the process. Some end up doing their own singing later. Some use AI to come up with ideas and run with them in a DAW. It really runs the gamut.
8) Where these "hybrid" uses of AI fall legally, I have no idea. I do know that coming up with your own melody, whether instrumental or lyrics, and then working with that in conjunction with the AI is difficult, but very rewarding when the collaboration gets your input and enhances it in a cool way.
9) All that said, I'd still rather be working with live musicians/singers. If you want to listen to some of what I've done with my vocals, lyrics, melodies in conjunction with Udio, you can click my avatar and go to my channel. *Edit* I should add, I've left a number of my original, copyrighted A Cappellas up on my channel. If you're inclined, you can listen to A Cappellas and the versions with music to see what working this way with music AI produces.
At this point, I'm a "former" musician as I stopped writing sync a few years ago after about 20 years because of how rapidly it was declining following the rise of subscription models and so many bands, singer/songwriters and hobbyists entering the field. While I still play and write traditionally, I've thrown myself 100% into AI generated music because the potential is amazing. I think many musicians who are resisting this are eventually going to miss out on some new opportunities. Production companies, ad agencies, and other media outlets using a ton of music for their projects aren't going to have an intern with no musical knowledge to haphazardly generate music on a regular subscription. They're going have an in-house workflow and hire a musician or a music contactor (freelance or small business) who know trends, understands song structure, and are familiar with the traditional side of scoring for visual media and music supervision. I see plenty of business opportunities here that will be far more lucrative than writing sync and doing custom composition here and there.
I think Shulman is right on the mark with his observations. Once music is your job, yes, it can be laborious as well as frustrating composing music for clients who want something revised over and over again. It's also not as profitable to only compose what you like. You often have to put something together you think is one step above "this sucks" because it's trending or that's what a client or library wants. You have to keep up on software, VSTi's, etc. And I think he's right about AI music being a community builder for both non-musicians and musicians. Both Udio and Suno have pretty active forums, charts, contests, etc. This is a good thing. It's fun, social, good for networking, etc.
Is this going to one day end musicians? No, not at all. Besides the new opportunities, people will continue to perform for the foreseeable future. Might it end the record industry? I think we can see that they're trying their best to make sure that doesn't happen and I doubt it will. They will probably own a large share of the AI generator market because they have the best training data in massive supply, expert music promotion with large budgets, and their artists perform to huge audiences. I don't think they're afraid of this at all. They're just trying to force a settlement that hands over some of Udio and Suno's revenue.
This is the guy that has learning disabilities. It is excellent for me because I write Lyrics. I have written over 2,000 songs, this is since I have been 17 I am now 58, so I am making full blown songs out of everything I wrote. My Cousin Willa, plays the harp, violin, Piano, Bass, Drums, keyboards and Ukulele, and many more instruments her mother forced lesson's, she hates playing music because of this. But this is a dream for me. When I told a friend that I was using "Band in a Box" my friend could not wrap his head around that. He spent many years learning his craft, and he told me he would never write music with me again, he never did. This was in 2012, so yes we never wrote any songs together again. He would not even get together to have coffee and even discuss music again. I ran into him and I told him about the latest technology, and he quite even talking to me about music now. He blocked me on the phone, social media, and even e-mail. I ran into him in Wall-Mart and he passed me up like a dirty shirt. Would not even speak to me, I ran into him at another place and he avoided me like the plague.
Thanks for chiming in, and thanks again for watching and commenting! Your 'friend' sounds like a jerk, his loss, not yours. The fact that you can get the songs out of your head into a musical form on your own is amazing! I bet Shulman would love to hear your story, because this is the sort of thing that makes tech evangelists and creators happy. I don't really care what the haters think, lots of people can't ingest opinion or ideas other than their own, and thats their problem, not ours. Keep on making music!
@@YoPaulieMusic I will, that is gratifying to be able to put all the lyrics to music after all those years of them just being Lyrics. And I can come up with my own tunes, and sing them into the Laptop, and it is 100% on in what I am trying to get out of my head. How can I send you a sample of my song called "You Should Have Been Mine"? I wrote the first version with "Band in a Box", then I put a 2 minute sample and it made my song a whole bunch better. One thing I can do if you will allow this, But I can also put both versions on Unlisted and if I send you the link you will be able to listen to it. I am not sure weather you will allow it for you to listen to, or how an I send you a link for you to listen to. The 2 different version are pretty close to the same, but the one made by suno opened up the song to it's full potential.
@@YoPaulieMusic I figured it out. I got your e-mail off your TH-cam. I will send you a link to this e-mail address.
@@YoPaulieMusic Sent
It didn't go through. Can I post the links here?