Beyond singing along to Neil Sedaka in the car, my friend has never been musically inclined. And yet, now he's a published songwriter.
He sent me a link and there it was, a Christmas song. It wasn't him singing, but it was a fully composed song. He wrote the lyrics and a prompt, and an online AI service built his musical vision.
I gotta try this, I thought. Finally, a chance to repair my relationship with AI. Between bad headshots and warped action figures, my AI experiments over the past couple of years, while fun, would probably not land in the win column.
I've always wanted to record music, but I was impeded by my musical limitations. Writing lyrics isn't the problem, and I could carry a tune in a bucket in my younger days. But my drumming is elementary at best, my keyboard work is weak on a good day, and I'm guessing at least two guitar manufacturers would gladly pay me to admit publicly that I don't use their gear. So, to make my rock star dreams a reality, I'd need a band, some sort of recording studio, and a substantial budget to pay for all of that.
Scratch that plan. Enter AI.
First, I needed a song. I've had a song title floating around my subconscious for a while, so I sat down and wrote "Rockin' Chair" in about 20 minutes. If you study the lyrics, you can pick up the subtext of getting older and the loss of youth. But mostly it's an invitation to get busy on the front porch. (Hey, this is rock 'n' roll, not a sewing circle.)
Then, on a lazy Saturday, I went to the Suno website on my personal laptop. It took a few minutes to register, a few minutes more to write a prompt, and maybe two minutes to type in the lyrics. So, let's say I invested about 30 minutes in the process.
Suno spit out several versions of my song in less than a minute. And they were good.
Oh, the cadence was off, and none of them sounded exactly like how I "heard" the song in my head. However, every version had a good beat, background vocals, solid instrument work, and all the spit and polish you'd expect from a hit single.
I literally closed my browser and walked away. Didn't even look at it for more than a week. Honestly, I needed to process what had just happened. In seconds, AI had made my song a reality. I hadn't even plugged in my amp.
At this stage of my life, I have no illusions of becoming a musical sensation. If everything goes right, maybe I'll make tens of cents in royalties from "Rockin' Chair" over the next decade, which will not even cover the cost to use the service for a month. Oh, you can get a taste for free, but if you want to tweak your song and avoid copyright issues, you'll need to spend some money.
But that's not the point, is it? The point is a guy who can only play a few chords on a guitar could publish a toe-tapping tune in less time than it took to write the lyrics.
That said, this is AI, which means it has to learn how to do all this, no doubt using data (read: songs) produced by copyrighted artists. Remember, AI is not about soul, it's about statistical patterns. Essentially, my song is a kernel of creativity wrapped in a cocoon of predictive chunks and synthesized production.
So, is Suno a shortcut? You bet. Does that make it less of a song? I'm not sure. After all, we've been listening to songs built around music samples for decades. One could argue that AI is the next step in that sort of musical evolution. And I have it on good authority that Nashville music pros are using Suno to quickly produce demos (and save a lot of money in the process).
I'm not ready to equate machine learning to inspiration. But I'll say this: I think "Rockin' Chair" is a good song—and without AI, it would have never existed beyond my notepad. Rock on.
