Hey! I’m Mike, I go by DJ AXCESS, and I’m a travel DJ, hype man and public speaker based in Columbus, Ohio.
This morning, like a mad man, I jumped out of bed, cracked open ChatGPT, and got straight to work. Why? Because I had just been listening to one of my favorite podcast hosts, Alan Berg, talking about the power of AI—specifically ChatGPT—and how it’s changing the way we work. I also heard ChatGPT passed a test where it is now indistinguishable from human chats.
Now, I’m usually more of a Google Gemini guy. I find it a bit smoother and I trust Google a little more than I do a newer company like OpenAI. That said, I’m always looking to level up my systems and streamline my workflows, especially when it comes to organizing my music crates.
So I decided to give ChatGPT a chance.
The Mission: Curate A Billboard Music List (No Ballads Allowed)
One area I’ve been wanting to refine is how I prioritize songs within my crates—specifically, ranking tracks by how likely I am to actually play them at a live event. That means focusing on high-impact, danceable tracks and cutting out the slow ballads. So, I gave ChatGPT a simple task:
“Give me the top 20 Billboard Hot 100 songs from each year, starting in 1970 through 2024—excluding ballads.”
Sounds easy, right? Well… not exactly.
Round One: Links, Confirmations, and Ballads
Right out of the gate, ChatGPT responded by confirming the prompt and attaching a Wikipedia link to every song title. Not ideal when you’re trying to copy and paste quickly into a Google Doc. So I adjusted the prompt, asking it not to include links or confirmations.
This helped—somewhat.
Even after cleaning up the instructions, the AI would randomly sneak in links or toss in ballads that I had specifically asked to avoid. So I had to get more specific, refining the prompt further to say:
“First generate the list, then check for ballads and replace any that don’t fit.”
It improved the results, but still wasn’t perfect. Sometimes ChatGPT would flag a ballad but then fail to replace it with another song from the same year.
The Upgrade Experiment: ChatGPT Plus
After about 45 minutes of back-and-forth with the free version, I figured it was time to see if ChatGPT Plus was worth the $20/month upgrade. It did not help.
Even with the Plus version, I still had to manually scan the list, remove a few ballads, and make editorial decisions on what belonged and what didn’t. All said and done, it took me about 1 hour and 15 minutes to create the list from 1970 to 2024. Not terrible—but still not a hands-free solution.
Final Thoughts: Can DJs Trust AI for Music Curation?
Here’s the honest truth: AI isn’t quite ready to take over crate curation—yet. While it did save me time compared to doing it manually, I still had to be heavily involved in editing, filtering, and refining the final output.
That said, if you’re a DJ looking to kickstart a project like this—especially if you’re sorting by genre, energy, or crowd response—AI can definitely help you lay the groundwork. Just don’t expect it to get every nuance right. Music is art, not just data.
What’s Next?
At the end of this blog, you’ll find a link to the Google Doc “we” created, which contains my evolving list of non-ballad, high-impact Billboard hits from the past 50+ years. It’s not perfect, but it’s a strong starting point for DJs who want to re-energize their playlists and keep the dance floor moving.
Check it out, let me know what you think, and feel free to shoot me your favorite go-to tracks that always hit.
Thanks for reading—and I hope to see you on a dance floor soon.
—
Want to stay up-to-date with my playlists, performance tips, and event highlights? Follow me on Instagram @djaxcessmusic, or check out my educational content for DJs at djaxcess.com.
P.S.
It took me another 40 minutes to remove the ballads it missed and reformat the list again. I wasn’t mad about this being a waste of time before, but I am now.
Hot 100 Non-Ballads by Year (1970-2024) Google Doc
https://docs.google.com/document/d/139lzJp-qmSd9xMsPNwm3QKz_iLfdN6kHAjN_3sh8nd4/edit?usp=sharing