by Jnmarseille13 (January 4th, 2024)
by Jnmarseille13 (January 4th, 2024)
During the last 2 year, we have all seen the astonishing evolution of artificial intelligence (AI), especially generative AI, which has brought many questions regarding ethics and safety. When it comes to us creatives (such as digital artists and music producers/composers), a lot of people have been debating on whether it is a threat, or something that we should embrace and use to our advantage. Many comments have also been made about the way a lot of these tools have been brought to life, retrieving data from millions of artists’ work without their consent to train their models. Will AI one day replace all creative people because it will surpass us? Will AI destroy art and music, the only things that we all thought would remain reserved for humans only? And if so, how long do we have before this happen? Are there any actions taken to prevent such scenario from happening? Or do we just have to accept this and give up on our passion?
So, I’m gonna try to give an answer to those questions, but I’ll stay as neutral as I can. Of course, like everyone, I have an opinion on this topic, but the goal of this article is to show you how AI actually works (a lot of us don’t really know about that), what are its pros and cons, and how it can (or will) affect our future as creatives. And before anyone asks, yes, this article is completely written by my human hand, just like every piece of content you’ll find on this website! Wait, I actually typed it on computer keyboard… whatever, you know what I mean!
So, first of all, let me give you a simple definition of artificial intelligence: it’s a concept in which a machine can emulate human intelligence and/or intelligent behavior, in a situation. That’s about it.
So, as we can see, AI is a very wide and vague term, and it includes a lot of things. And believe it or not, it’s been there for a very long time. When I say, “long time”, I mean before many of us were even born. And chances are, many of us have been using AI for years without even realizing it (automatic language detection, Siri, non-playable characters in video games, YouTube Content-ID and more). But the one AI technology that has made us all go crazy lately was machine learning, and more precisely: generative AI.
Machine learning (ML) refers to a system that will learn from an existing set of data provided, and then having the system execute tasks and make predictions and logical decisions based on all the data that it learned from, instead of giving it specific instructions through a program. Some people will argue that ML isn’t really part of artificial intelligence, as it can only produce results based on what it learns and cannot do much outside of the context, we trained the system on (but if you don’t mind, I’ll keep referring to ML and AI as the same thing, to make things easier).
So, let’s take an example. You often use YouTube, and you spend more time watching videos about beatmaking and music production than anything else. YouTube’s algorithm studies your behavior on the platform and notices some keywords that pop up often in your research, lile “FL Studio”, “mixing” and “producers”. After learning about your behavior on the platform, YouTube’s algorithm will then determine that you are interested in music production, and randomly pick some of the most popular videos that matches this interest of yours and suggest it to you next time you open YouTube. This is a classic example of machine learning.
Now, when it comes to generative AI, it’s a technique that uses ML to not just execute tasks and make predictions or decisions, but also to generate new content based on the data provided. The more data it has, the more the results will start to look as good as the data that was provided for it to train on. And that’s exactly why the digital art world has started to feel threatened these last 2 years, with tools like Midjourney and Dall-E. Because even though they’re not exactly intelligent, these systems have kept generating better looking art that sometimes looks almost just as good as human made digital art, making it pretty much indistinguishable from human work; the only difference is that, where it would take hours of work for a human to perform such artwork, it only takes up to about a minute for a computer to generate it, after simply being given an instruction (called a prompt). And of course, as many of us know, a lot of progress has been made in the music area as well, making many of us artists, beatmakers, composers and producers, also feel threatened in a way.
Many petitions have been made online by countless artists and creatives of all types, claiming AI development is bad for all creative industries and will leave many out of work, while also killing the spirit of art by suppressing all the human emotion and connection that has always made art meaningful. Others, however, believe that the rise of AI in this field is part of the evolution of human technology, claiming that, just like others tech revolutions have killed jobs and created many new ones (example: cars replacing horse riders, but also creating new mechanics jobs), AI will simply change artists work, but not destroy them.
If we focus on music, it is true that the multiple innovations in music technology have made music production somewhat “easier” and more accessible to everyone who takes an interest. Back in the 70’s, every role was very distinct and every person involved in a music project was an expert in their role: an audio engineer would master all the recording and FX tools in the studio, composers would take their time to write songs, performing artists were great at singing and/or playing their instruments, executive producers were great supervisors who understood every part of the creative and business aspects (I personally believe that this is why 70’s music is the best music we ever had). But as time went on, technology became easier to use more accessible, and more and more people being able to have their hands on every part of their creative process. In the 80’s, we were starting to see a few artists who could almost master every part of the creative process, such as Roger Troutman, Gerald Albright, Prince, Jean-Michel Jarre and Michael Jackson in some way. Nowadays, it’s very common to see artists who have full hands on the entire process, from songwriting to recording to releasing to promoting their releases on their own.
While some will use this statement to remind everyone that these evolutions, which were largely criticized at first, have actually had a positive impact for artists and musicians, others will argue that AI is a different breed, because, unlike those previous evolutions, it almost completely wipes out any human input, leaving it pretty much all to the computer. And some of the neutral ones will argue that we still make music on our own, while using some AI-generated content to help with the process.
But there’s one thing that is unfair, and that many people on both sides agree on: the artists’ feeling of being “robbed” in some way. Remember how ML and generative AI works? Remember how it needs to be trained on existing data in order to generate results? Yep, you see where I’m going. These systems that generate new content that looks just as good as human work, in a matter of seconds, have been trained on an unimaginable amount existing artwork available online. So, as they should, they probably made sure to only retrieve those images on copyright-free stock images websites, right? Well… they didn’t. They took billions of non-copyright-free images on well-known art display websites, without getting the original artists’ consent or compensating them in any way. And, with no surprise, the same thing has been done with music too.
And once the cat is out of the bag, there’s no way to get back in time. These systems are already well evolved by now, so only time will tell if they have a negative or positive impact on creatives. But in the meantime, artists have made sure that measures get taken by authorities to protect them (by filing real lawsuits).
While this is a very complex territory that I’m not an expert in, I will try to summarize what I understood about the regulations concerning the release of AI content and the way copyrights work (in the European Union and the United States of America, but I believe most regions operate the same way, when it comes to copyrights).
You, as a human, are allowed to create anything you want, using anything you want (in most case, a computer). Until recently, using a computer to create art and music still always required a lot of human input, and did not mean telling the computer what to do then sitting back and letting it do its thing (and it still doesn’t mean that). So, you automatically own the copyrights of your work.
But with generative AI, while you still have to use your imagination to create a specific prompt, you don’t have as much control on the result that the computer will generate. In this case, the law considers that there wasn’t enough human input in your work, which means you cannot claim full copyrights.
Then there’s the case where you put a lot of human input in your work, but also used AI in some parts of your work, either to speed up the process, or to start with a base and work over that. In that case, the law grants you full copyrights on your work, considering that the human input was significant.
Yes, AI is a serious topic for us musicians. Digital graphic artists were the first to be really impacted by this, but it’s happening in music as well. It looks like lawmakers have taken some decisions in favor of the creatives, but the way this technology is evolving makes me think that we don’t know what’s gonna be next. Those of you who are old enough to remember how Napster changed the music industry forever can probably understand even better. Major labels tried to stop Napster, but it led to a remodeling of the industry. AI has also been criticized by major labels and could lead to changes in the industry.
Yes, I said that I was gonna be neutral throughout this whole article (and I think I did well, lol). Being a tech guy, I’ve always been interested in new technologies (I’m the developer of this website), but as a music guy, I must admit that this one topic hits a bit different. But, as you all know, once the cat is out the bag, it can’t go back. It’s here now, and we’ll have to work with it, whether we like it or not.
So yes, I tried to stay as neutral as I could, but if you know me, or you look at the “About” page, then you will probably be able to guess what’s my opinion on this. Sometimes, I wonder what will happen if things like this make the music industry collapse one day. Because yes, music existed before the “record industry”, before the 20th century. Those times with no music recording technology, where people would only listen to music occasionally when people were playing an instrument in concerts and parties, or traditional events. So, what if AI destroys the music industry and make music become something non-commercial like it used to be back then?
Only time will tell what will happen. In the meantime, we gotta create music and release our stuff! 😊
During the last 2 year, we have all seen the astonishing evolution of artificial intelligence (AI), especially generative AI, which has brought many questions regarding ethics and safety. When it comes to us creatives (such as digital artists and music producers/composers), a lot of people have been debating on whether it is a threat, or something that we should embrace and use to our advantage. Many comments have also been made about the way a lot of these tools have been brought to life, retrieving data from millions of artists’ work without their consent to train their models. Will AI one day replace all creative people because it will surpass us? Will AI destroy art and music, the only things that we all thought would remain reserved for humans only? And if so, how long do we have before this happen? Are there any actions taken to prevent such scenario from happening? Or do we just have to accept this and give up on our passion?
So, I’m gonna try to give an answer to those questions, but I’ll stay as neutral as I can. Of course, like everyone, I have an opinion on this topic, but the goal of this article is to show you how AI actually works (a lot of us don’t really know about that), what are its pros and cons, and how it can (or will) affect our future as creatives. And before anyone asks, yes, this article is completely written by my human hand, just like every piece of content you’ll find on this website! Wait, I actually typed it on computer keyboard… whatever, you know what I mean!
So, first of all, let me give you a simple definition of artificial intelligence: it’s a concept in which a machine can emulate human intelligence and/or intelligent behavior, in a situation. That’s about it.
So, as we can see, AI is a very wide and vague term, and it includes a lot of things. And believe it or not, it’s been there for a very long time. When I say, “long time”, I mean before many of us were even born. And chances are, many of us have been using AI for years without even realizing it (automatic language detection, Siri, non-playable characters in video games, YouTube Content-ID and more). But the one AI technology that has made us all go crazy lately was machine learning, and more precisely: generative AI.
Machine learning (ML) refers to a system that will learn from an existing set of data provided, and then having the system execute tasks and make predictions and logical decisions based on all the data that it learned from, instead of giving it specific instructions through a program. Some people will argue that ML isn’t really part of artificial intelligence, as it can only produce results based on what it learns and cannot do much outside of the context, we trained the system on (but if you don’t mind, I’ll keep referring to ML and AI as the same thing, to make things easier).
So, let’s take an example. You often use YouTube, and you spend more time watching videos about beatmaking and music production than anything else. YouTube’s algorithm studies your behavior on the platform and notices some keywords that pop up often in your research, lile “FL Studio”, “mixing” and “producers”. After learning about your behavior on the platform, YouTube’s algorithm will then determine that you are interested in music production, and randomly pick some of the most popular videos that matches this interest of yours and suggest it to you next time you open YouTube. This is a classic example of machine learning.
Now, when it comes to generative AI, it’s a technique that uses ML to not just execute tasks and make predictions or decisions, but also to generate new content based on the data provided. The more data it has, the more the results will start to look as good as the data that was provided for it to train on. And that’s exactly why the digital art world has started to feel threatened these last 2 years, with tools like Midjourney and Dall-E. Because even though they’re not exactly intelligent, these systems have kept generating better looking art that sometimes looks almost just as good as human made digital art, making it pretty much indistinguishable from human work; the only difference is that, where it would take hours of work for a human to perform such artwork, it only takes up to about a minute for a computer to generate it, after simply being given an instruction (called a prompt). And of course, as many of us know, a lot of progress has been made in the music area as well, making many of us artists, beatmakers, composers and producers, also feel threatened in a way.
Many petitions have been made online by countless artists and creatives of all types, claiming AI development is bad for all creative industries and will leave many out of work, while also killing the spirit of art by suppressing all the human emotion and connection that has always made art meaningful. Others, however, believe that the rise of AI in this field is part of the evolution of human technology, claiming that, just like others tech revolutions have killed jobs and created many new ones (example: cars replacing horse riders, but also creating new mechanics jobs), AI will simply change artists work, but not destroy them.
If we focus on music, it is true that the multiple innovations in music technology have made music production somewhat “easier” and more accessible to everyone who takes an interest. Back in the 70’s, every role was very distinct and every person involved in a music project was an expert in their role: an audio engineer would master all the recording and FX tools in the studio, composers would take their time to write songs, performing artists were great at singing and/or playing their instruments, executive producers were great supervisors who understood every part of the creative and business aspects (I personally believe that this is why 70’s music is the best music we ever had). But as time went on, technology became easier to use more accessible, and more and more people being able to have their hands on every part of their creative process. In the 80’s, we were starting to see a few artists who could almost master every part of the creative process, such as Roger Troutman, Gerald Albright, Prince, Jean-Michel Jarre and Michael Jackson in some way. Nowadays, it’s very common to see artists who have full hands on the entire process, from songwriting to recording to releasing to promoting their releases on their own.
While some will use this statement to remind everyone that these evolutions, which were largely criticized at first, have actually had a positive impact for artists and musicians, others will argue that AI is a different breed, because, unlike those previous evolutions, it almost completely wipes out any human input, leaving it pretty much all to the computer. And some of the neutral ones will argue that we still make music on our own, while using some AI-generated content to help with the process.
But there’s one thing that is unfair, and that many people on both sides agree on: the artists’ feeling of being “robbed” in some way. Remember how ML and generative AI works? Remember how it needs to be trained on existing data in order to generate results? Yep, you see where I’m going. These systems that generate new content that looks just as good as human work, in a matter of seconds, have been trained on an unimaginable amount existing artwork available online. So, as they should, they probably made sure to only retrieve those images on copyright-free stock images websites, right? Well… they didn’t. They took billions of non-copyright-free images on well-known art display websites, without getting the original artists’ consent or compensating them in any way. And, with no surprise, the same thing has been done with music too.
And once the cat is out of the bag, there’s no way to get back in time. These systems are already well evolved by now, so only time will tell if they have a negative or positive impact on creatives. But in the meantime, artists have made sure that measures get taken by authorities to protect them (by filing real lawsuits).
While this is a very complex territory that I’m not an expert in, I will try to summarize what I understood about the regulations concerning the release of AI content and the way copyrights work (in the European Union and the United States of America, but I believe most regions operate the same way, when it comes to copyrights).
You, as a human, are allowed to create anything you want, using anything you want (in most case, a computer). Until recently, using a computer to create art and music still always required a lot of human input, and did not mean telling the computer what to do then sitting back and letting it do its thing (and it still doesn’t mean that). So, you automatically own the copyrights of your work.
But with generative AI, while you still have to use your imagination to create a specific prompt, you don’t have as much control on the result that the computer will generate. In this case, the law considers that there wasn’t enough human input in your work, which means you cannot claim full copyrights.
Then there’s the case where you put a lot of human input in your work, but also used AI in some parts of your work, either to speed up the process, or to start with a base and work over that. In that case, the law grants you full copyrights on your work, considering that the human input was significant.
Yes, AI is a serious topic for us musicians. Digital graphic artists were the first to be really impacted by this, but it’s happening in music as well. It looks like lawmakers have taken some decisions in favor of the creatives, but the way this technology is evolving makes me think that we don’t know what’s gonna be next. Those of you who are old enough to remember how Napster changed the music industry forever can probably understand even better. Major labels tried to stop Napster, but it led to a remodeling of the industry. AI has also been criticized by major labels and could lead to changes in the industry.
Yes, I said that I was gonna be neutral throughout this whole article (and I think I did well, lol). Being a tech guy, I’ve always been interested in new technologies (I’m the developer of this website), but as a music guy, I must admit that this one topic hits a bit different. But, as you all know, once the cat is out the bag, it can’t go back. It’s here now, and we’ll have to work with it, whether we like it or not.
So yes, I tried to stay as neutral as I could, but if you know me, or you look at the “About” page, then you will probably be able to guess what’s my opinion on this. Sometimes, I wonder what will happen if things like this make the music industry collapse one day. Because yes, music existed before the “record industry”, before the 20th century. Those times with no music recording technology, where people would only listen to music occasionally when people were playing an instrument in concerts and parties, or traditional events. So, what if AI destroys the music industry and make music become something non-commercial like it used to be back then?
Only time will tell what will happen. In the meantime, we gotta create music and release our stuff! 😊