In recent months my social media has been inundated with adverts for Suno and people raving with the results that they are achieving. In fact Suno has become one of the most talked‑about AI music tools, and for good reason — it’s fast, surprisingly capable, and pushes right into territory that used to belong only to trained musicians.
Given my interest in all things AI I thought it was worth a deeper dive and I will state from the outset that I am a sceptic, although after reading I am tempted, but only for personal use of course!
Suno is a generative AI music platform that creates full songs — vocals, lyrics, instrumentation, mixing — from simple text prompts. Designed so that anyone, even with zero musical background, can use it you describe the style, mood, or idea, and it produces a finished track instantly. It offers as part of the package multiple modes, vocal styles, and customisation options.
Suno draws upon advanced AI models trained on huge amounts of audio data and while the company doesn’t disclose the exact training set, the process, as with much generative AI starts with your prompt, which ideally should be quite specific for example ‘80s hair metal power ballad or ‘70s prog rock, etc. Using these prompts Suno then generates the lyrics, vocal style (for example male female – which can also be prompted), instrumentation (drums, bass, synths, guitars, etc) and mixing and mastering.
The output can then be fine-tuned with options to change the vocalist’s gender, adjust the style or “persona” and even exclude certain instruments. Thus, with these variables in play, multiple variations can be generated before being downloaded, shared, or used commercially in line with Suno’s T&Cs.
What has surprised me is that some musicians actually like it, as a means to quickly sketch out ideas and provide inspiration. Furthermore, some musicians (according to what I read but I dispute this) say that Suno democratises music by allowing non‑musicians to express ideas. In addition it also provides a lot of assistance with the production which takes a lot of the heavy lifting out of the process. It could even be considered to be a “co‑writer that never gets tired.”
On the flip side there is a question of original source drawn upon (as with generative AI artists use – remember my tiger in an earlier post about AI?), as this could be copyrighted music without permission. More fundamentally it could put session musicians, jingle writers, and composers out of a job. We must remember only very few make it to the top and become mega-rich, while many others make very little or even a loss for their love of the music.
There is then the objection that resonates most strongly with me in that it represents a devaluation of craft because it creates a situation where literally anyone with access to the program can make a song now.
However, Suno is ultimately a tool that can speed up workflows and provide greater flexibility and room for experimentation. Having said that, my gut feeling is that something like music, as with all of the arts, comes from baseline talent that you either have or don’t have. It can be trained, finessed and developed, but that’s as far as it goes. I would also go as far as to say that if you can’t perform it live then don’t bother and that applies to many so-called popular artistes today.
What do you think?
As always stay safe and well my friends.
