AI disqualifies anyone as a musician! It's like playback.

[guitarzan]

You could do that already with tools in Logic Pro X. But it means you have to do a little bit more creative thinking that writing prompts in a text window.

So the canned AI bandmates in Logic are that much more challenging? You know that if you write the other guy’s part it’s not jamming, right?

[D-Fusion]

Shadow producers in 2026.
Human = Expensive but you get what you want out of a song.

AI = Cheap but you will waste a lot of time and credits in hope that it will make something that you like.

I do let AI make songs when i am bored but I won't renew my subscription when my subscription is over in October.

Reason for that is that Suno has met a wall on their latest v5.5 model so it plays the same sounds and chords and melody no matter what genre i use now.

So there is no variation there anymore and it outputs the same 2-3 songs over and over again.

Same chord progression, same sounds and over the top sound fx and other crap.
It also play in the same key progression in 95% of the songs now so it looks like the Algorithm is starting to go in circles and is slowly eating itself up.

[BONES]

If you enjoy writing songs and are confident in your songwriting, I don't see why you would ever turn that whole process over to AI.

I can think of half-a-dozen reasons off the top of my head:

1. I don't see songwriting as a worthwhile pursuit, in and of itself. (I also do not see it as something to "enjoy" any more than cleaning the house or putting out the bins.) To me it's just an administrative task you have to undertake if you want to have songs to perform on stage. I am every bit as happy playing covers as I am doing my own original songs.

2. It's not about me, it's about music and if I can have better songs to perform by using AI, and I absolutely can, then that's what I'll use.

3. It's been a great way of pushing us outside of our comfort zone, of trying things we wouldn't have thought to try on our own.

4. It's just a tool, like a guitar or a piano or a synthesiser.

5. It's been a far more exciting and enjoyable experience/process for us than working in the way we have been working for most of the last 30 years.

6. You have to be willing to try new things if you want to learn and grow as an artist.

But AI as a creative partner can help you take things in very interesting directions you might not have considered.

So why not let it? It's only your ego that prevents you from going all-in.

I'm never tempted to use the AI's ideas without significantly changing them.

Again, that's just ego and I'd suggest it is holding you back.

Also, AI could be helpful to create instrumental parts for songs you've already written completely.

We go the other way, we add parts to what the AI has already generated to flesh out the arrangements, make them sound more like us and add a bit of extra interest to proceedings.

... (which essentially gives Suno the rights to your song).

So what? Do you think you're such a songwriting genius that your music is worth anything to anyone? For a third time, you're letting your ego get in the way of logic and reality.

Making music is fun.

Maybe to you but not to me. The "fun", if you want to call it that, is in getting up on stage and performing the music. Where the music comes from, how it comes about, is of no importance to me at all.

I can't imagine using AI to generate music for me. What would the point of that be?

I dunno, maybe you'd do something worth sharing? If you don't try, you'll never know.

Pulling from the void and life's inspirations to write melodies, harmonies, interesting progressions, etc.

I've been writing songs for 45 years and I can't say I've ever done any of that bullshit. Melodies? Harmonies? Progressions? WTF!?!

... in the end create a real sense of accomplishment and artistic fulfillment.

Again, maybe for you but I've never found that to be the case. In fact, I definitely get greater a greater sense of accomplishment from working up a MIDI file into a cover version I am happy with than I ever got from writing my own songs the old fashioned way. It feels like more of a challenge to me, to do right by something you like, created by someone you rate. But I/we have really enjoyed working on/with the AI generated songs we've made with Tunee. It's been the most enjoyable album-making experience of my "career" and, weirdly, it feels like the one where I've had to make the least compromises to get the result I/we have been chasing. It has freed us, to a degree, from the constraints of our tiny minds, allowing us to paint with a much broader palette. I honestly don't understand why you'd let your ego dictate that you have to do it all yourself. It's madness.

What's great about that is through those jam sessions your understanding, even if just intuitively, will improve and you can then further explore writing those ideas in ways that fit your personal artistic identity.

That's exactly what you get by writing prompts for AI. In the hands of someone who knows what they are doing, and has the bucketloads of patience required, what the AI creates can be as pure an expression of the prompter's "personal artistic identity" as anything else. How can you not see that?

You could do that already with tools in Logic Pro X. But it means you have to do a little bit more creative thinking that writing prompts in a text window.

If you honestly believe that, then you should stay away from AI because you'll do a shithouse job of it. Getting the AI to do what you want it to require a level of creative thinking you probably don't possess. I don't think I do, either, but my bandmate is an absolute f**king ninja with it.

Human = Expensive but you get what you want out of a song.

You reckon? That's not been my experience working with producers. You have to be lucky to find someone who can really help you realise your artistic vision.

AI = Cheap but you will waste a lot of time and credits in hope that it will make something that you like.

That's where the hard work comes in. It's not wasted time or credits if the results are worthwhile.

Reason for that is that Suno has met a wall on their latest v5.5 model so it plays the same sounds and chords and melody no matter what genre i use now.

We got to the same point with Tunee before they changed to different AI providers. The new models they have access to are worthless, although we did manage to get one song out of them. It seems we might have lucked out with our timing.

[ksandvik]

You could do that already with tools in Logic Pro X. But it means you have to do a little bit more creative thinking that writing prompts in a text window.

So the canned AI bandmates in Logic are that much more challenging? You know that if you write the other guy’s part it’s not jamming, right?

Yes, you need to think as a musician, rather than prompt that anyone who would write could do.

[zerocrossing]

Shadow producers in 2026.
Human = Expensive but you get what you want out of a song.

AI = Cheap but you will waste a lot of time and credits in hope that it will make something that you like.

This was my experience using a generative video AI site called OpenArt AI. I probably could have eventually gotten something good out of it, but no way would it have happened as easily and cheaply as the paid influencer videos made it seem. Why did I have to pay to correct a floating, disembodied head above a park bench, when the prompt read something like, "<character> sits on a park bench..." How could I have rewritten that prompt to ensure it wasn't a floating head? Sure, that only cost me a few dollars, but what if I had been doing something important?

Reason for that is that Suno has met a wall on their latest v5.5 model so it plays the same sounds and chords and melody no matter what genre i use now.

So there is no variation there anymore and it outputs the same 2-3 songs over and over again.

Same chord progression, same sounds and over the top sound fx and other crap.
It also play in the same key progression in 95% of the songs now so it looks like the Algorithm is starting to go in circles and is slowly eating itself up.

I'm seeing this a lot in other GenAI. The same look and elements over and over. The art director (not really an art director) at the last gig I had told me to use Midjourney to get some quick concept art of different potential biomes for new levels of the game we were working on. (his excuse was it had to be done in a day and we didn't have the time to have our concept artist do three finished images.) Months later, I literally saw an almost identical image randomly posted by someone on social media. I don't know if this is a problem with the model, or a problem with the fact that people are tending to accept the same results and the model is learning that it's a good result that should be replicated, over and over. That sure seems to be a dead end. I saw an art exhibit that used a lot of AI in it, and while the artist did a really good job of using it to it's best (it was an exhibition about future technologies), it was super clear to me that it was all AI, except occasional live media that was being overlayed in short bursts on some of the pieces.

This all leads me to think that what you're experiencing is fairly universal, and we're close to seeing the end of this fad. That's not to say that I think AI isn't great for some things, like removing elements from images. I used to spend hours doing this. I had friends who worked on films where they had to go frame by frame and remove wires and other elements. I don't think that kind of work is a particularly good use of human hours, but I hope the result of having these tools is more humans doing the fun and creative work, and relegating the kind of grunt work to the machines.