On May 5, 2024, Metro Boomin dropped a tweet. The tweet contained a link to a SoundCloud upload entitled “BBL DRIZZY 150 BPM.mp3” as well as an invitation for artists to put their own spin on top of the track. The beat was infectious, containing an earworm in the form of a driving Motown-style hook that crooned “BBL Drizzy”, a reference to Drake’s rumored brazilian butt lift. Motown samples are a mainstay of hip hop culture. The sound of chipmunk voice-style pitch manipulation over the iconic sound of the fifties and sixties feels so familiar to hip hop fans, it was no surprise that many didn’t stop to wonder: wait, what Motown artist fifty years ago knew what a BBL, or for that matter Drizzy, even is?

That’s because no authentic Motown artist did. The song Metro sampled didn’t exist prior to this year, because the song Metro sampled was an AI generated motown song.

With BBL Drizzy, Metro Boomin unleashed the first AI-generated infectious ear worm into pop culture and showed how AI will rapidly accelerate sampling in hip hop and pop production.

The fact hip hop appears to be the first scene to both embrace AI and spread its creations into culture shouldn’t be a surprise. Hip hop has been the dominant musical influence on pop culture for over a decade now. Hip hop has also been nearly synonymous with the practice of sampling, or repurposing previously recorded music as the basis for new music.

Sampling has historically been a laborious process of first finding the perfect sample then editing or “chopping” it up to a new track. You can get a sense of this process in this excellent breakdown of Jay-Z and Kanye’s song “Otis”. Then, once musicians have perfected their creation, they then need to jump through the legal and financial hoops needed to license or “clear” the sample for distribution and monetization.

For producers and musicians using AI like Metro Boomin, almost all the headaches are taken out of the digging, editing, and (controversially) clearing process for the sample they wish to use for their work. Instead, all a producer needs to do is prompt a service like Udio for the genre, vibe, or even lyrics they want to use in their work, get a song back, and get to work.

How AI generates music can even be likened to the practice of sampling, done at a scale only computers can. Computers are trained to listen to the history of recorded music, learn their genres and vibes, edit or “chop” them down to atomic parts, then are prompted to recombine them into something new based on text prompts in emerging AI music services like Udio. Each atomic part of an AI-generated song is essentially a sampled portion of a vocal or instrument sound that had been previously recorded by an artist.

Where Kanye used an AKAI sample pad in the 2010s to sample instruments and vocals, up and coming teenage producers of the 2020s will use text prompts and AI models.

BBL Drizzy is just the first example of what will be the new norm in pop music production. Producers will not need to laboriously search for the perfect sound, texture, or vibe to start their work. Soon, audiences around the world will be singing along to the melodic and lyrical genius not of one human, but of the sum total of the collective works of humans sampled and served back to us by AI.