Hi 👋 hey, how you doin’? Before proceeding, did you know the image above is not just an image? Click it, see what happens, I dare you. Or don’t, I’m not going to bother using reverse-psychology on you, it wouldn’t work anyway, you’re too smart for that nonsense. Moving on…
My friend directed me toward a new “AI” service where you’re able to create music. This video is an example of something I was able to create, using my own lyrics. Yes, this boy can lyric, thank you very much—is it that surprising from a writer with a love for music? From the moment my friend sent me the link to this service to now I’ve been experimenting with it, trying different combinations of genres and sitting and writing lyrics. It’s absolutely amazing from a technical innovation point-of-view, though I do have some concerns about what the future holds for true artists.
Rise of the Machines AI Advanced Machine Learning Algorithms in Music Creation
AI (I should let you know my disdain for this term, it’s woefully inaccurate in my humble opinion, technically) has been creeping its way into many facets of our lives. I’m sure you would be shocked to realise how prevalent it is in our day-to-day lives, most of which is not visible, it’s managing things behind the scenes. It is steadily coming into view of the general public. I was one of those nerds who was using OpenAI tools before ChatGPT exploded, waaay back in 2020/2021. I thought it was pretty novel, though at the time it didn’t dawn on me how groundbreaking it would become, much like how Bitcoin exploded—let’s not go down that road. I dabbled here and there but I never really put serious effort in, my usage was constantly surface-level and I used it for helping with day-to-day tasks, much the same as I do now. As it developed, I found the transcription side of things really intriguing and it was very useful. It had it’s hic-ups and wasn’t perfect, now services have pretty great transcriptions abilities, among others. Take Descript for example, I used them to create the video for the song.
Creating music using AI (from a pleb’s perspective—read, me) hasn’t really been very dazzling, it was basically the same as DALL-E during its inception (and still now in many cases), whereby it creates weird artifacts and, well, you can tell that it’s been created by AI. We’ve reached a new frontier with music.
If you didn’t know the attached song was created using AI, would you have been able to tell? I doubt it. With conformation bias I think it’s a moot point even posing the question. Put in the comments how you think you could tell it was created using AI.
The Good, the Bad, and the In-between?
I’ll give you my opinion. I’ve been listening to music since I was a wee bitty lad, and I have to admit, I love it. Always have, always will, the end. I have inhaled many different types of music through my ears likely from before I was born, thanks to my parents. To the point where I have a chicken-egg problem with tinnitus, I don’t remember a time where I didn’t have tinnitus—and subsequently I don’t know if it was from listening to music to loud when I was young or if it’s something I was born with. Is that even possible? I tried learning music and it was difficult for me, mainly due to concentration issues and other ADHD-ly things—if that topic interests you, I have this post, and that (← this one—ahem, that one is the best so far IMO) post, and even another one over here. So I know some about music, formally but I generally don’t say that out loud because it’s not much—let’s just say I definitely have an ear for music. I know what I like, and what I don’t. Working on the attached song, I knew instantly if the version was rubbish or not. I’ll give you two fors and two againsts for what I think are the biggest implications of AI in the music industry. Let’s begin.
Eventual Influentialness (Or: How I Learned to Stop Traditional-ing and Love the AI)
Implications
There are two big implications I see stemming from the mass-adoption of AI within the music industry, firstly,
Nature Music Preservation
Do you have any idea how many traditional music and traditional music styles have been lost to the wind? Neither do I. Isn’t that crazy? It’s humbling to think about how many of anything has simply disappeared from existence—species, technologies / inventions, other civilizations, dinosaurs we never knew about, methods for cooking, recipes—I’m the list could go on. Consider for a moment, the ways in which we could preserve this for future generations. They could have a digital archive of how music sounded centuries ago. Amazing.
Exploring New Frontiers of Music
Computers are efficient at doing certain tasks really well. One of those is taking inputs and calculating other combinations / permutations (I always have to look up the difference between the two). One of the downsides of the tool I used to create the attached song was that it struggles combining certain genres together. This is because it’s taking input from what already exists in the world. This would change with advancements to the technology. In the future you would be able to mix and match any genres with any other genres. If you want to see a video of me creating a song from scratch, fill the below poll or drop it in the comments and I’ll make it happen—you can even help me choose the genres / theme, etc.
Benefits
Make It Quick!
Creating music using AI tools would cut down production times significantly. Lowering the barrier of entry to music creation enables just about anyone to create music. This can be considered a pro and a con, as I’ll discuss in the drawbacks section. Let’s focus on the positive here! It would be UH-mazing for people like me who aren’t able to put the time into formerly learning all the ins and outs of music theory, music production and so on. Additionally, AI doesn’t need to be used to create the entire songs, it could be used within any part of the process—composition, arrangement, editing, mixing. It’s not like we’re being forced to all use it for the entirety. We are human, we do still have choices—for the moment. And all those repetitive tasks that everyone hates doing can go out the window.
Collab Improvements
Having an AI buddy at our side can really help with creativity and collaboration. It’s like having an enhanced executive function attached to your fingertips—or your ear if you’re using voice functionality, and likely ‘telepathically’ in the future. To be clear, collaboration in this context would be collaboration between a human and AI—though I’m sure you could think of novel ways for the AI to increase collaboration between humans though that doesn’t seem to be a high priority.
B-sides Downsides
Once More With Feeling
If you’ve used ChatGPT or related tools for any length of time, you may have noticed how unfeeling they are. They’ll spit “facts” (using “facts” without quotes would be pretty funny, you hallucinationist, you) all day long, however, they’ll never give you their opinion, they’re not allowed to have opinions. I’m not sure I necessarily want them to, I’m just pointing it out for argument’s sake. There is a definite lack of authenticity and emotional depth. A computer can’t have the same lived experiences that a human does, and even if they did, they wouldn’t have any feeling about them one way or another—they’re weird as [expletive] boiii! Sidenote: As a kid, I used to get creeped out by things that tried to appear human but weren’t—it was so disconcerting. So AI can create “music”, though we may find that humans don’t resonate with what it creates in the same way that would with “pure”, human-made music.
Rest Uneasily on My AI Metaphorically Created Laurels
Many of us humans don’t want to do things the hard way. For good reason. Though there is the argument that struggle builds resilience, allows us the mental fortitude to power through and get things done. Creative thinking stems from boredom, so one could argue that in this day and age our collective creativity is diminishing at an alarming rate due to us being “entertained” 24/7. We don’t get a break, our brains can’t rest, we are constantly actively engaging our brains and they don’t have time to digest all the inputs. I fall victim to this as well, it’s difficult. The algorithms’ tentacles draw you in from every direction with one-way barbs that tear into the fleshy goodness of our creative minds.
When we do things the easy way, there is a risk that we become dependent on whatever is making the task easier. Could you imagine washing your clothes on a washboard instead of in a washing machine—having to sit, bent over for a couple to a few hours washing every item of clothing, one-by-one. No, instead we get to spend around 30 seconds dumping the clothes into the washing machine, put in the powder / liquid and pressing a couple of buttons. That gives us the time to do the same with the dishwasher, then the bread maker, rice cooker, slow cooker—you get the idea.
So for the positives we’ve listed above, the compensating factor in this case is that an overreliance on the technology could potentially lower the creativity of “musicians”—would we call them musicians? I feel like that would be like calling prompt engineers writers, it doesn’t sit right with me. It’s also possible that people’s creativity gets displaced to other areas, it could manifest itself in different ways. Much like as music in its traditional form has evolved and the techniques used to generate different sounds have evolved along with it. I’m still amazed when I find out the unique ways people have created sounds for their music.
Fin
That’s all folks. I was originally planning on simply posting the song, then I gave it some thought and realised I definitely have something to say on this topic. We covered what I think some of the implications of AI interrupting the music scene may look like, as well as the pros and cons—because hey it’s not all bad, and whether we like it or not AI is here to stay and is only going to keep advancing.
That’s it for now, as always, good luck, stay safe and be well.
Share this post