Home Creators Posts Import Register Favorites Logout

Videos

  • 592613868.mp4

Downloads

Content

For the seventh year in a row, Vox’s Future Perfect team convened to make forecasts about the year ahead. Senior editorial director Bryan Walsh is betting there’s a 60 percent chance that a fully AI-generated song will hit No. 1 on the Billboard Hot 100 in 2026.

Here’s Bryan to break down the reasoning behind that prediction and the way that AI is changing music.

Related Links26 things we think will happen in 2026What AI in music can — and can’t — do

Your Patreon membership also includes unlimited access to all of our stories over at Vox.com. Go to vox.com/patreon-connect to connect your account for unlimited reading.

Files

Comments

Toby Ricketts

Great Art has always surprised and delighted. AI is a summary of what we already know and what has come before. Another big question is does pop music need to be art? or has it morphed into a commodity to entertain our resting brains with familiar patterns? Perhaps we will return to a renaissance-era patron system of art funding, where wealthy philanthropists fund art creation - bringing along with it the issues of bias and exclusivity in art...

Bryan Walsh

Thanks, Toby — really thoughtful comment. I’m with you that a lot of pop is built on familiar patterns, which is exactly what makes AI interesting (and a little unnerving) as a hit-making tool. And the patronage point is sharp too: if creation shifts toward whoever can bankroll training/models/marketing, you get a new version of old gatekeeping, with all the bias and exclusivity that comes with it. Would we get "better" music? Maybe - but I've always thought the most original art bubbles up from below.

EternalLurker

Yesterday I just heard, for the first time, an instrumental track that was a generative AI product — a simple synth tune, sure, but still a full-fledged six minute piece, and an admittedly catchy one at that. I recognized it as GenAI a few seconds in despite having had no idea the field was already that far along (got the exact same uncanny valley vibes I do when looking at most GenAI images), but I've been playing piano for a few decades now; nobody else in the comments section of the associated Reddit thread seems to have realized yet that it was AI-generated. And that's for what appears to have been a complete musical amateur asking commercial software to do essentially all the work for him. Professionals with experience in mixing engineering and the like who can touch things up afterwards to minimize AI artifacts should, judging by what I heard, be able to make much better products that could probably fool even those of us with musical backgrounds. Not that anything beyond being catchy matters all that much to your prediction anyway, given your analysis on what actually determines a chart-topper. So yeah, I'm far less knowledgeable about lyrical songs as opposed to instrumental tracks, but still 60% seems reasonable to me even based on where things seem to be technologically right now, nevermind where they'll be by year's end.