In my quest to bring you the new, new thing, I probably should have written about Hit Radio AI as soon as its creator, veteran programmer/engineer Dave Solomon, brought it to my attention around Labor Day. “All AI-generated hit music from average people — 150 artists and about 400 songs,” he wrote.
By the end of September, there were 50-100 submissions daily. Even then, he wrote, “I have seen [the music] make amazing progress in just a few months, and know that a year from now, it will be incredible. Devices that can create songs based on your mood, etc. Music will become much more personal.”
In recent months, the mainstream hit music pipeline has slowed to a crawl. It’s exciting to CHR programmers that the first apparent hit of 2025, SZA’s “BMF,” emerged within days, not 2-3 weeks. On Jan. 2, 2025, Hit Radio AI added 22 songs. Less than week later, on Jan. 8, it added another 35 titles. Solomon has built a community of creators who promote the station on their social media. He’s also created a podcast.
I knew Solomon through his work in Oldies/Classic Hits, then through an AI/radio Facebook group. Through that time, the controversy among ROR readers was more about the prospect of AI personality in radio. (Hit Radio AI is jockless, but Solomon says AI tracks will start soon.) On broadcast radio, the initial public efforts at AI jocks seemed only to confirm the value of human contact. In 2024, there was surprisingly little to report and the consensus seemed to be that AI would be of more value to radio in other ways.
The issue of AI-generated music has been equally galvanizing among creators. In a companion piece this week, S-Curve Records’ Steve Greenberg suggests that readers think of AI in the way that they came to accept sampling, as a creator’s tool. He also suggests that labels are more willing to allow existing music to inform AI, as long as they’re compensated in the way they would be for sampling.
That it took me a while to write about Hit Radio AI was, in part, trying to wrap my head around both the concept of AI music, and trying to ingest a body of music that sounded mostly like the hit music we know, but was 100% unfamiliar. It took me a while to put it in perspective.
A lot of the songs I heard on Hit Radio AI remind me of hit music from the late ’00s: often with a similar mix of trappy and ethereal. To some extent, it makes sense that AI-generated music would be an extension of bedroom pop, especially if it’s being created there. But I also heard songs that sounded like late-’00s turbo-pop, like late-’80s Eurodance or hair ballads. Its creators’ idea of hit music was wider than what Top 40 accommodates now.
When I first listened to Hit Radio AI in September, it felt like most of the songs could be Shazamed or found in streaming. (There’s a station Spotify playlist of songs from around that time.) When I listened this week, there seemed to be a higher percentage of songs that stumped Shazam or weren’t yet available in streaming, an indicator of how fast the music is proliferating.
Here’s Hit Radio AI just after 11 p.m. on January 7. I’ve tried to convey the best sense of what a dozen songs likely unfamiliar to you sound like without resorting to the A&R or label promo person’s oft-reductive “a cross between Artist A and Artist B.” As with the sound coding of music by radio people — in which no two people make the same decisions about even familiar megahits — the artist might have a very different take, or you would.
- Kuda, “Dive In” — trappy, but melodic male pop
- Josh Ray, “Father of the Revolution” — melodic rap, seemingly Hamilton-inspired, in which the narrator takes the POV of George Washington
- Atom Bites, “Bitcoin Bandit” — reminiscent of ’30s jive or the ’90s swing revival. but with decidedly timely lyrics
- Kade Noir, “Feel Some Way” — pop/R&B ballad definitely reminiscent of the late ’00s
- Morning Star, “Be You” — affirmative female teen pop; a modern-day Radio Disney song
- Voices of Melbourne Florida, “Chasing the Clouds” — female-sung/rapped vocal; ethereal intro. An act that had new music added to the station last week and this
- Amira Uwu, “Skin Care” — R&B ballad, one of the year’s first adds
- Arizto, “Untouchable” — the turbopop throwback and the most melodic, uptempo song in the stretch I heard; I like to think that the resemblance in the act’s name to the lost ’70s hit “Eres Tu” is intentional
- AI.Ka & Inkaz, “Love Is Gone” — the song with the ’80s Euro feel. The act is on Shazam and the artwork self-identifies the act as “AI.KA Digital Singer”
- Angel Warrior, “Last Goodbye” — the late-’80s-style epic hair ballad
- Triad, “Locked Away” — female singer/songwriter pop
- Mr. Bazeman, “Hide and Seek” — ballad with a Country feel and lyric about missing childhood
After several listens, I haven’t heard the song that I would have recognized as blatantly AI-generated out of context. That’s much different from radio’s stilted early jock breaks, although chances are those have improved in the last year too, and we may be hearing some less awkward ones that aren’t being flagged for us. On the station, I heard only a lyric or two that might have had the awkward wording of an AI chatbot-generated lyric. And some artists play up the AI aspect, such as the act currently playing, (AI)n’t Real.
In the stretches of Hit Radio AI I’ve heard, there have been songs that I enjoyed. I haven’t yet encountered the song that I would have taken into a label A&R meeting. But, as happened when I did A&R and listened to random submissions, I have encountered creators whose next song I would want to hear. Also, to be fair, I also might not take a lot of the music coming from TikTok seriously without the built-in cachet that we give it as an industry. When you take a listen, I’ll be interested in your reactions.
Greenberg’s positioning AI as just another in a succession of artist tools that once didn’t exist — the electric guitar, the synthesizer, the sample — makes me more open to its uses. Greenberg is also clear that the tool needs a human POV to validate it, which in turn makes me more dubious about the concept of AI voice-tracking at radio, since it replaces the work of people, instead of augmenting it. In the meantime, Hit Radio AI’s debut means that when we look at Intriguing Stations of 2024 next week, there will definitely be something that truly didn’t exist last year.
This story first appeared on radioinsight.com