- This topic has 13 replies, 7 voices, and was last updated 1 year, 8 months ago by .
-
Posts
-
“Democratizing music production” is a euphemism for putting skilled, knowledgable composers out of work, or at the very least making it not worth the effort to continue. The fact that anyone with a smartphone can crank out a cue, whether they are musically literate or not, is what is driving the value of music down, not setting low prices on your tracks.
One of the most revolutionary advances of AI music tech is that it democratises music creation for everyone, enabling non-musicians to generate studio-quality music, and in doing so – unlocks the potential for any creator, app developer or tech company to programmatically generate original music for their specific use-case, media or Metaverse environment.
I don’t know. Most of the AI stuff I have messed with is pretty horrible. Here is one. I find it more trouble than it’s worth. https://www.aiva.ai/
I think that it’s pretty clear that AI is going to take out the low end. The “hold one finger down and create a cue” composers.
Beyond that level, it’s going to take a LOT of intelligence to start competing at a pro level of composition. Especially those composers who utilize a good amount of real musicians.
But no doubt – the low end consumers of music are going to be very happy fiddling around with some janky software trying to get something useful out of it.
at the very least making it not worth the effort to continue
This is very perceptive Michael. It may not be AI alone right now, but all factors combined that can make it not worth the effort to continue. AI plus shrinking percentages, lack of upfront money, lack of or no sharing of sync fees, subscription services, TV and radio streaming paying pennies, libraries that once earned either out of business or no loner earning. I’m sure I’ve missed some. I hope there are some able to navigate through this and succeed. As you know, I’m personally on a break from submitting exactly because the decline in income just isn’t making it worth the effort.
This is part of the reason I try to build as much “humanity” into my tracks as possible. I think AI will do a scarily good job within just a few years for cranking out stuff like you hear under HGTV (and we know how much they care about the composers making the music currently, anyway). I saw a deepfake of Obama next to a real video. I guess wrong which one was real. If there’s money to be made, or not spent by the networks, they’re going to be using this stuff left and right for the lower-end shows, and of course plenty of non-broadcast usage.
I’d like to think that we as artists/composers/humans will always find a way around obstacles. The old phrase “adapt or die” comes to mind. We somehow find a way of doing just that.
I personally just don’t see a market for AI music any time soon. Even if it can create music of the same quality as stock libraries, I don’t see why someone wouldn’t just continue to use stock libraries. Stock music is very saturated. Why mess with a computer program (or pay someone to mess with a computer program) to create a track that can already be licensed for $3?
The same thing is already happening with AI images. It’s certainly fun to mess around with AI art. But if I need an image of a rolling landscape, there are thousands of existing stock photos available for free or for pennies. I don’t see people using AI images commercially, as far as I can tell anyway.
Production music isn’t very much different. A lot of production music is already being placed with no sync fees. If I’m a music editor, where’s my incentive to create AI music when there are millions of tracks already available that I’m placing at No up-front cost? The networks are paying for PRO licenses anyway. Unless they ditch all their existing programming that has any music written by humans, they won’t save any money by using AI music, because they still have to maintain their PRO licenses. They would have to ditch the PRO model altogether, which would mean ditching every note of music written by humans. I just don’t see that happening en masse any time soon.
Sure, programmers may develop the technical ability to create something resembling stock music, but it doesn’t have a market that isn’t already saturated.
AI music is also far more complex than AI images. An image is a 2D array of pixels, but music is like thousands of 2D images sequenced in a coherent way across a time domain. You can certainly use AI to create some MIDI notes, but that’s very different from a finished composition mixed and mastered by a skilled composer. You could hand over your MIDI data to a non-composer, and tell them exactly which instruments to use, and they wouldn’t be able to replicate the quality of your work.
In short, I think market saturation from more and more composers writing music is a much bigger threat to the income of composers, but I don’t have the right to decide who gets to create music, so all I can do is keep trying to get better.
I think you bring up a good point about saturation, Chris, for music that is “static”–you can’t change it. However, I’ve seen several AI demonstrations where one is able to enter musical changes into a timeline. Imagine a video editor getting their hands on this–it’s like getting someone to score to your picture without paying the fees. Is it here, now? No, but things are moving very, very quickly.
For me, the key is to put as much “humanity” in to my tracks as possible, and make sure they evolve in satisfying ways. Until the AI-helper tools are widespread, it’s all I can think to do.
Well, I hope I’m right, but I have been wrong from time to time in the past. 😉 Maybe I’m whistling past the graveyard.
I created a track using this site to see how the technology works and these are pre-composed tracks where you can “arrange” 4 to 8 bar sequences of the same chord progression…
https://soundraw.io/edit_music?length=180&tempo=low,normal,high&mood=Hopeful
I do not think that AI technology is ready for real higher end scoring, but for 8 bar passages for 15s and 30s and even 60s…it can work. I still think that the video editors will appreciate human composed music. And my “Test” usage of this where I attempted to score a 3 minute video had these issues:
1. It still took 1 hour!
2. It sounds too repetitive so it will not sell.
This site is essentially a “Royalty Free” AI site consisting of already composed ideas and loops.
AI music and computer generated music has been around for 25 years. I used to paste loops and bass lines and horn riffs together using ACID in 1999. Some of that music was cool, but one still had to know how to arrange it, mix it, etc….you still needed “musical skills” to make an impact with the track.
The market will decide, but I do not feel obsoleted by these merging technologies just yet. These technologies do not save time for video editors, they still have to search and hunt, and attempt to “paste together” a music arrangement. I think the video editors will still opt in and buy into real human compositions that are emotive.
Ah, ACID. Remember it well, those were the days!
- You must be logged in to reply to this topic.