Listen (3:08)
September 7, 2026
I think it’s safe to say that a song can now be written (generated?) with software as its first audience. A potential listener may still be someone driving to work or relaxing on the beach, but the release strategy increasingly begins elsewhere, primarily with a system that measures the first few seconds of not hitting the stop button. Long introductions become a risk. A sudden change in tempo can look like a problem. The song might not be worse for it, but it arrives through the car speakers or earbuds already trained to avoid being passed over.
Artificial plays add further confusion. Thousands of streams can make a track appear active, and the numbers themselves carry a kind of authority. A listener who finds it on a prominent playlist may never know how it got there. They may not care. From the old days of radio programming to record store retail displays, placement has always shaped listening habits. What is new is the ease with which activity can be fabricated and the difficulty of distinguishing a made-up audience from a real one.
This is not just a music dilemma. A growing body of material (yes, I am avoiding the repulsive term “content” here) also seems to exist largely because it can be processed. Songs designed for unobtrusive background listening, articles built from recycled phrasing, and images generated simply to fill a blank space all have a place in the current system. They can also be sorted, copied, reworded, and reused. A person may encounter them, but no particular person seems to have been imagined when they were created. I think that absence is becoming more perceptible, even when the work is technically polished.
People get used to what is placed before them. A playable track appears when they need music, a concise answer when they have a question, and a generic image when a page needs decoration. It saves time. It doesn’t ask too much. That convenience can narrow expectations for art or writing that surprises us. Music becomes revenue-generating when it produces traceable consumption “events.” Streaming increasingly makes music legible to machines before it becomes meaningful to people. It doesn’t need to develop an audience in the ordinary cultural sense.
Throw AI-generated content into the cauldron, and we have a full-blown conundrum from a human cultural perspective and in terms of creativity. Many still think they are choosing what they like, but isn’t the algorithm doing all the heavy lifting for those not actively exploring or searching for new stuff? For original material (yes, that word again) to generate revenue, it needs to get the “attention” of algorithms, which, in turn, draw the attention of other algorithms, layering it into everyone’s feeds and playlists, on repeat in some capacity, to successfully monetize the click, the view, the listen.
Humans are the minority online, particularly on social media, and we’re on a narrowing path toward the entire creator economy devolving into a strictly tech economy. When it comes to taste, culture, and the willingness to value another person’s creativity enough to pay for it, exploration and curiosity are the principles that will keep true fans of art out of those troublesome bubbles and silos.