We’ve had a few similar discussions in the past, but the field keeps evolving.
I hadn’t been really following this much until Suno came to my attention recently. And there seem to be quite a few rather similar things:
Some of them seem to produce scarily good output from even minimal prompting.
What do you musicians think?
Is it time to hang up the instruments and go play shuffleboard in the Old Folks Home?
No. It’s not threatening the same way “Parappa the Rapper” (a 1996 Playstation music/rhythm game) wasn’t threatening. It’s just not the same thing. Live music is not ever going to be replaced by a smartphone app.
I like this.
[quote=“DCnDC, post:2, topic:1024988, full:true”]
No. It’s not threatening the same way “Parappa the Rapper” (a 1996 Playstation music/rhythm game) wasn’t threatening. It’s just not the same thing. Live music is not ever going to be replaced by a smartphone app.
[/quote]
(Does ‘copy quote’ not work?)
I personally lose interest if I kno it has no ‘soul’. 
I don’t have any real interest in AI generated music but I could imagine scenarios where I wouldn’t really care. Electronic club beats or background music loops in a video games, for example. Music where I’m invested in the lyrics would probably result in me being disinterested in the song. This is sort of interesting to me because I have no issue with AI image making but the best answer I can think of is that I interact with them differently. I don’t typically spend 3-5 minutes looking at a picture but a song has a longer time commitment in addition to me potentially learning the lyrics, singing along, singing it to myself even when it’s not playing, etc. It’s just a vastly different sort of attachment for me versus even the most compelling visual art.
The potential to see an artist live and enjoyment of live music definitely helps but there’s also many songs I like where I have zero chance of ever seeing the musician live. Maybe they’re dead or retired or maybe they’re just some small-time performer in another country who isn’t popular enough to be touring near me. But it would be a factor if I heard a song and learned it was AI.
I’m reasonably confident that I haven’t heard any AI songs yet through Spotify, etc. This is mainly due to me mostly sticking to a curated play list and being rather selective with adding new music versus just letting it feed me an algorithmic playlist regularly. But now and then I see a story about some AI song getting a million listens and have no clue what the song is or sounds like. I also resist the urge to take a listen, ‘lest the Mighty Algo decide I’m saying I want more like it.
Absolutely. There’s nothing like the energy of a good live band!
I was thinking more from a composer & songwriter point of view. Some of the recent sites can crank out a fairly convincing demo in a few seconds from just a title. Something that would take me as a musician and writer at least a few hours in my studio with a good DAW and instruments.
Not Mozart or McCartney… but at a competent level.
I have two worries about this. It gives people with little or no songwriting skill or experience the illusion of ‘creating their own song’.
And it raises the bar for young potential musicians and writers to the point where they may think: “What’s the point of studying instruments and my craft? I can’t even produce something as good as the damn computer right now”…
I’ll never play the guitar like Marty Friedman but I’m not quitting just because I’ll never sweep pick 64th note arpeggios. David Gilmour is never going to win a head-cutting competition against Steve Vai, but he’s no less a guitarist or musician. I make music that I feel, that came from me. What was spit out by a computer, or anyone else, at any quality, is irrelevant to my own creations.
I’m totally with you. But it’s seductively easy to fire up one of these applications and see what they produce. Technically at least, some of it seems scarily good.
I almost feel, like George Harrison after the ‘My Sweet Lord’ fiasco, that I should stop listening to other music, or at least nowadays AI generated music altogether….?
What sort of style(s) do you write in?
Me? Blues and riff-based instrumentals. A lot of experiments in layering and multi-tracking, with a lot of bluesy noodling on top.
Sounds like we’re rather on the same page. Have you heard any of my material?… I have posted links in other threads, but could repeat if needed. Any interest in trying any collaborative work?
Maybe. I’m in and out of here. Sort of between recording setups right now.
I made an “album” of songs for my wife’s birthday using ChatGPT and Suno. They are all songs about her and her life. I published it to all the major streaming services. It took a couple of hours and less than $100. We attend a lot of concerts and have a lot of “musical” friends. Several friends asked who the band was and how much did I pay them to record the music. For better or worse, I think the day is rapidly approaching where A.I. music will be indistinguishable from (and probably surpass) human music. :-/
That was a nice gesture. But, and please don’t take this too personally, how much of you, your own heart and soul, was in those songs?
I find them absolutely amazing and an interesting tool for brainstorming ideas. I’ve uploaded one-minute segments of songs I’m working on and asked for variations on it, or extensions of it, or different stylistic takes on it, and it’s been helpful in working through ideas I’ve had festering for years on my hard drive. This is all music for me, so I don’t care, and I play it all in the end, I just sometimes need a spark to inspire me what direction to take an idea or to come up with an interesting arrangement that I had not considered.
I use it in fits and spurts, but I love the technology and it flabbergasts me how any of it is even possible. I’m a fairly technophilic person, but this is still all like magic to me.
In this case quite a lot. I spent quite a bit of time thinking about what I wanted to say (to and about my wife) in each song but, since I’m not a song writer I didn’t know how to express it lyrically. I had ChatGPT tweak the lyrics over and over till I got what I wanted. When I played the songs at her birthday, with the lyrics on a big screen, she cried, her two daughters cried, a couple of the female guests cried. I’m not in any way saying that computer systems should replace humans in artistic endeavors but, from my personal experiences I see a day where it will be really hard for humans to compete. There will always be a place for humans to fill up a venue like Red Rocks but I suspect more and more of what we listen to via streaming and other audio only methods will come from A.I. 
I would avoid jingle writing as a career.
As others pointed out, I’m sure AI music generators will be fine generating music that is largely meant to be trivial and soulless anyway. Commercial jingles, ambient background music for videogames or corporate trade shows. Stuff like that.
Or maybe in a decade or so, ten thousand women at a live concert screaming along with Taylor Swift to one of her ex-boyfriend songs will seem as quaint and archaic as big band music or a harpsichord seems to us now?