You can make a track now without having much to say. That’s always been true. What’s different in 2026 is how fast a finished sounding piece can appear.
In July, Deezer reported that fully AI generated tracks had passed half of its new music uploads on peak days in June, at roughly 90,000 a day. That figure counts deliveries to one platform. It doesn’t measure what people actually choose to hear, and Deezer says those tracks accounted for just 1 to 3 percent of its streams. To me the gap between the two numbers is as interesting as the headline.
There’s more music available than any one person can follow. So what makes one electronic track feel like it came from somebody?
A polished sound is only the beginning
A strong kick and a wide pad can make an immediate impression. Add a clean rise and you have something that sounds finished. You can also assemble all of that into a track that leaves nothing behind when it ends. I’m not saying that to dismiss craft. I’ve spent enough time with production software to know how much work a few apparently simple bars can contain.
But a finished surface and an artistic point of view are different things. You can hear a point of view in the decision to hold a chord for an extra bar, or to let a rough texture interrupt a beautiful one. You can hear it when a hook comes back quietly even though every instinct says to make it bigger.
Those choices can sound as clean as you like, as long as they feel considered.
The useful question is “why this?”
When I’m working on a track, the temptation to add is always close by. One more layer might thicken the chorus. A new preset might make the lead seem more expensive, and another transition might tell the listener that a big moment is coming.
The harder question is whether the piece needs any of them. Why this melody, in this register? I ask the same thing about the spot where the bass leaves space, and about whether the second return should be different from the first.
A tool can offer possibilities quickly, and a library can give a producer a beautiful starting point. Neither one does the listening for you. Somebody still has to choose, reshape and sometimes discard, and that’s where a producer’s identity becomes audible. I started on FruityLoops in 1998 and have moved through other production environments since (there’s more on that in Building an EDM Sound in Los Angeles). The tools in front of me keep changing, but I still have to decide what belongs in the track.
Being human is no excuse for being careless
I’m wary of turning this into a purity contest. Electronic music has always been made with machines, sequencers and samples, and the assistance keeps getting more sophisticated. Still, there’s a meaningful distinction between using technology in a creative process and asking a system to produce most of the music in your place. The boundary can be complicated in individual cases, but it’s worth describing honestly.
Platforms are drawing different lines. Bandcamp’s January policy says music generated wholly or substantially by AI isn’t permitted there. A policy can’t tell me whether a particular chord progression moves me, but it does tell me what sort of creative community a platform wants to support.
Listeners deserve honesty too. If an artist explains how a record was made, that story should be true, and a human credit should mean something. I want the process to be open enough that curiosity about a track leads to a person and a set of choices, not a fictional biography.
Make one decision nobody asked you to make
Here’s a small exercise for the next session. Find the moment where the track does exactly what its genre has taught it to do, then try a different move. Drop the drums earlier, or leave the vocal line unanswered. Keep the quieter version of the synth. Let the last chord change the meaning of the hook.
The alternative won’t always be better, and that’s fine. You’ll have learned what the predictable version was doing and whether it earned its place. I think a distinct voice gets built through that kind of attention, one decision at a time.
Having this much music around makes discovery harder, but care still counts. If someone spends four minutes with a piece of music I made, I want those four minutes to contain choices I can stand behind. The sound can be as vast as it likes, as long as I know exactly why it’s there.
For more on the listening side of those decisions, read Beyond the Drop, my guide to hearing more in electronic music.