GeekWire’s quest for an AI theme song opened our ears to the debate rattling the music industry – GeekWire

For a while now, we’ve been doing an occasional GeekWire Podcast segment called “My AI,” about the latest AI tools grabbing my attention and changing the way I work.

But we’ve been missing a fun musical intro or jingle for the “My AI” segment that would really solidify its place on the show. My colleague John Cook’s attempts to improvise a “My AI” theme song don’t really count, as he would readily admit.

So naturally, we asked ourselves, could AI solve this problem?

We quickly discovered that the answer was yes. The results from our experiments with two AI-based text-to-music generators were pretty amazing, at least in some cases. You can hear for yourself on this week’s show. The services Udio and Suno generated some truly catchy songs, as well as a few duds.

But the incredible thing was that all of the tunes, plus some very nuanced lyrics, were generated from a simple prompt: “Can you create a short introductory jingle for a GeekWire Podcast segment called ‘My AI’?”

Here is one of the songs, an electronic pop tune, generated by Suno. I liked this one because of the structure, which would work well as the intro and outro for a podcast segment (which was the assignment, after all).

In a world of code and screens
Where tech is more than just a dream
My AI, My AI
The future’s here it’s in the sky
My AI, My AI
Tech that makes us reach so high

Just for fun, in some cases, I asked the AI music generators to use specific genres. Here, for example, is a “My AI” country music theme that sounds like it could have come from a Merle Haggard set list.

Talkin’ ’bout my AI
Geek news, set your mind a-fly
Join us on this wild ride
Where tech and truth collide

From the latest in machines
To the code that paints our dreams

We’re breaking down the bytes
Exploring all the sights
In a world that’s so pristine
Where the future’s evergreen

John was partial to this Udio tune, in the spirit of German electronic band Kraftwerk.

My A-I, join the crew
My A-I, look what’s new

My A-I, here to stay
My A-I, leading the way

Tune in to GeekWire, it’s My AI day!

Those are just a few examples. We play more of them on the show, with reactions from John and our longtime colleague, GeekWire Editor Taylor Soper, and discuss the underlying issues raised by these AI music tools.

Even when the music was awful, the technology was pretty incredible.

But when it comes to how the songs are made, how they should be used, and what this means for human creators, the answers are a lot more complicated.

One indication: when the services are asked to create a song in the style of a particular artist, they respond by saying that they can’t do that, but they suggest a series of genres and styles in the spirit of the requested artist.

From a licensing perspective, the services maintain that we’d be within our rights to use the resulting songs for a segment on our podcast. This is from Udio’s FAQ:

Can I use the content I generate using Udio for commercial purposes?

For example, can I distribute the music I generate on music streaming platforms? Can I include it in a monetized video on YouTube? Can I license it for use in TV/movies/advertisements/etc.? Yes, as long as the content does not contain copyrighted material that you do not own or have explicit permission to use, and as long as you properly indicate that the content was generated using Udio

But the reason these services were on my radar to begin with was because they are the targets of litigation from the Recording Industry Association of America.

Here are excerpts from the RIAA’s complaints against Udio and Suno:

“Building and operating [these services] requires at the outset copying and ingesting massive amounts of data to “train” a software “model” to generate outputs. For [these services], this process involved copying decades worth of the world’s most popular sound recordings and then ingesting those copies [to] generate outputs that imitate the qualities of genuine human sound recordings.”

“When those who develop such [services] steal copyrighted sound recordings, the [services’] synthetic musical outputs could saturate the market with machine-generated content that will directly compete with, cheapen, and ultimately drown out the genuine sound recordings on which the [services were] built.”

Here is part of Suno’s response:

Like a human musician, Suno did not develop its capabilities in a vacuum. It is the product of extensive analysis and study of the building blocks of music: what various genres and styles sound like; how songs in those genres and styles are harmonized and structured; the characteristic timbres of the instruments and vocalizations in those genres and styles; and so on.

Those genres and styles—the recognizable sounds of opera, or jazz, or rap music—are not something that anyone owns. Our intellectual property laws have always been carefully calibrated to avoid allowing anyone to monopolize a form of artistic expression, whether a sonnet or a pop song. IP rights can attach to a particular recorded rendition of a song in one of those genres or styles. But not to the genre or style itself.

After using AI to generate the songs, talking with John and Taylor, reading the RIAA’s complaint, and giving it more thought, I’m conflicted about actually using any of the resulting music for the “My AI” segment.

Bottom line, I’m holding off for now, and this project might not be over yet. Stay tuned. Meanwhile, I’ll be trying to get these AI-generated earworms out of my head.

Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.