Supposedly, Suno Made My Skill as a Music Producer Irrelevant

and why I still don't give a fuck and do it every single day.

Suno AI has basically made my skill as a music producer irrelevant. But I continue to make music because it is the quickest way that I can enter flow, and it still brings me a childlike sense of wonder.

And unlike most fear-mongering musicians who are trying to outlaw and ban AI, I embrace it. This has opened up a whole new era of creative possibilities that just simply weren’t possible before.

I’m finishing music that needed a very expensive artist or session musician to finish. Or at least seeing an idea through close enough to completion to FINALLY decide: leave it in the hard drive grave of archived music, or finish it, release it, and genuinely be at PEACE with all of this.

When it comes to release and the brand press kit, I have insane, unlimited creative possibilities for visuals and graphics to help with the marketing of the song with the use of tools like Higgsfield & Grok.

My creativity now is the only limit... And I can see why that might be terrifying to other artists and musicians who hyper-specialize in one little area of the music making process. I can also see where their fear comes from... and I believe it mostly stems from ignorance. I think the people who fear it the most are the ones who least understand it.

Anyone that has tried to 1-prompt any music or AI-generated art/graphic/writing first understands that it will require several attempts to iterate a decent enough version to release to the public. And even still, once you have gotten an iteration that may be close to the final version you were looking for, you will still realize that it just doesn’t sound quite like the pros.

Go Try It for a Few Hours

I beg any artist who hates and fears this technology to go in and actually try and use Suno for a few hours. You’ll quickly realize how inconsistent and horrible it fucking sounds to try and generate anything remotely decent.

But you’ll also quickly find that sometimes it will give you some incredible ideas built off of your own input, which could be something as simple as an 8-bar loop you made a few years ago or some chord & melody ideas you created yourself but decided to see where Suno could take it. You may even find moments of peak euphoria from hearing what could be sonically possible with your song by just experimenting a little.

But then you’ll come across the main problem. You have a version that sounds incredible and close to the finish line. But to try and go in and fix little sections without completely re-writing the song now turns into a new game of patience, praying and hoping it doesn’t shit itself and turn a synth-pop-sounding record into an indie-folk-sounding song, because it has a hard time staying consistent between iterations.

Inside My Process

And so this is how my workflow goes when I use Suno.

I have an 8-bar loop, or maybe even a full beat already laid out from intro, verse, pre-chorus, chorus, break, verse, chorus, bridge, chorus outro already arranged out for my song... but I don’t have a vocalist/artist that I personally know that I think would sound good over the song...

And so I run it through Suno. I ask it to add vocals that have a certain tone and style that I think would sound good over it. I may even ask Claude to help me specify exactly what I want the vocalist to sound like by talking to it through Wispr Flow. It gives me a decent prompt to bring back into Suno. I then put it in, tweak a few settings... and begin to generate...

Now, there’s been a few times that it genuinely WOW’d me... and made me go... OMG this is incredible. But more often than not, I am there regenerating until I find something even remotely good that makes me go... mmm, OK, I have seen this idea through... and I am bored and not interested anymore. What is the next idea?

And so for those 2-3 versions out of 100 that made me go WOW... I then decide... OK... there is no way I am going to get a clean version of these Suno-generated vocals, even with the “remove FX” option and “stem separation” tools in Suno. However... I know now that this song has immense potential, and so I begin to look for artists on platforms like soundbetter.com to see if I can hire a real human to help me refine this idea even more and finally get this song over the finish line.

THIS is how I use AI in my workflow.

AI Is the Co-Producer, Not the Producer

It starts with a human-made idea. I use AI as a co-producer in the studio with me that can help me see ideas through. QUICKLY.

Once I land on something I genuinely like and that gets me fucking amped, I then proceed to take that idea back into my hands, make the final tweaks that make it sound PREMIUM and not so thin and tiny, which is what Suno generally makes music sound like; to my ears at least.

From there I send it off to an artist who I believe will be able to take the vision and inspiration from this collaboration of human taste and an AI-generated vocal & melodic performance, and I get my new collaborator to brainstorm more ideas with me, re-record, re-arrange and continue to remix it into a new direction I couldn’t have even imagined when I started from my original 8-bar loop...

Honestly, as I’m writing about it I am getting excited, because in this process I am making new friends through the job board soundbetter.com as well as through the power of social media, and I am creating music I never would have thought imaginable.

Who We’re Really Making It For

Now, I can only speak from my experience with how I use AI. I am sure there are plenty of lazy producers trying to just 1-prompt AI to make them a hit. And there are others who literally just export whatever slop Suno makes for them and they consider it the final version and upload straight to their DistroKid without a second thought...

And so to those who fear AI, let me just put this out there straight and clear. Those people do not concern us... because we can almost always tell when something has been made with that methodology. It sounds weak, half-assed, generic, soulless... like nothing truly spectacular that makes us go... woah... and I think the general population and masses can sense that too.

These fake artists (whether it’s the producer prompting it or the AI itself performing it) are irrelevant to our mission. But was that ever the point anyway? Aren’t we making music FOR OURSELVES first? And the audience comes last? At least that’s what Rick Rubin and I believe.

But either way, that shouldn’t stop you from embracing this tech and finding new ways to use it to inspire new music you never would have thought possible.

At least this is my naive and optimistic take.

I could be wrong, and yes, I am well aware they are most definitely “robbing” my sound by training their models on the original inputs I give it... but oh well... I don’t want to make the same music twice anyways.

I want to explore and keep experimenting with new genres, sounds, styles, textures, vocal performances and arrangements. And AI is just a tool that makes it possible much faster and for a lot cheaper.

AI is not coming to take my joy away from making music. It has actually done the opposite and has made it more exciting than ever before.

— Dan V