Exclusive: Disclosure’s Bronze Editions Break New Ground In AI Music

BY Lorepunk

December 12, 2023

When we set aside considerations of price and popularity, one of the greatest joys of our space is discovery: appreciating, for the first time, what a new technology can do—and what this can mean for the future. That sense of discovery is what comes to mind as Disclosure: Bronze Editions, the groundbreaking collaboration between the celebrated electronic music duo Disclosure and the cutting-edge AI technology of Bronze, opens minting today on next-generation digital marketplace Beatport.io.

While it is not the first time that AI and musicians have collaborated on a collection, Disclosure: Bronze Editions takes a novel approach. Disclosure worked directly with Bronze AI as part of their production process, remixing their hit song “Simply Won’t Do” into 1000 unique variations that were generated directly from its individual elements, or stems.

Every edition of the lush, catchy song–available at Beatport.io as an NFT on Polkadot, for $20USD–will also come with a unique work of visual art. In collaboration with German art collective Grotesk Group, Bronze translated each remix into RGB color values, analyzing the audio spectrum of each track to create the elegant, sculptural pieces, digitally wrought in materials like marble, glass and bronze.

Working With Bronze AI

To a layperson, collaborating with an AI might mean entering a prompt into a tool like Midjourney or DALLE. For Howard and Guy Lawrence—the award-winning English duo who make up Disclosure—working with Bronze was a powerful amplification of their existing toolset and their considerable production skills. “Bronze is like Logic—it’s a DAW, or digital audio workstation, but it has many capabilities that Logic does not. The integration of working with Bronze and Logic together is very powerful,” Guy Lawrence told nft now in an interview.

While Bronze did the remixing, Disclosure’s experience, skill and taste played a pivotal role in creating the variations. We invited Lawrence to walk us through how the process worked. “For this project, I put the stems (each part of the music, drums, bass, vocals etc.) into Bronze. From there, you can set the parameters of probability that something may or may not occur in the playback, whether that be a tempo change, restructure of how the song plays back, whether certain notes come in—or do not come in—at certain points, whether the drums are half-time or full-time. Anything really,” he said.

Howard and Guy Lawrence, the brothers behind Disclosure

After setting the parameters in Bronze, the heart of the process was repetition: running the AI over and over again, while tweaking the inputs. “The key is that it’s based on probability. So, each time it plays back, you’re getting a different chance of it playing back differently—and with so many different parameters set, it’s very unlikely it would ever play back the same [version] twice,” said Lawrence. “This was obviously super helpful for this project, as we needed 1000 different variations of the same song. It would be a lot of work for a human to have to do that by hand,” he said.

That time-saving factor may be among the most revolutionary aspects of the Bronze AI tool, making the discovery of new musical combinations easy—and engaging. “Bronze made this [process] incredibly speedy, interesting and definitely took the song to places I would not have thought, too—especially if I had to try 1000 times,” said Lawrence.

Matching Music And Visuals

After the creation of the music, Grotesk Group stepped up to create the visuals. “We took those [1000] songs and used the contained data as something like a seed for our tools and techniques,” they told nft now.

As soon as they set the AI running, the art collective began to see results in the visuals. “The actual process was actually quite straightforward…We were happily surprised [at] how much information we were able to pull from the data, how well we were able to instrumentalize it, and how beautiful everything looked as soon as we got our workflows connected. There is—of course—human decision making involved, but in the end that data is based on the essence of Disclosure’s musical style,” they said.

The innate contradiction in the visuals—delicate, almost floral forms, wrought in bold materials like bronze and glass—capture the essence of the song, which is about love and passion. “We were most surprised by how fragile the generated artworks are. It’s easy to see the emotional connection the fragile artworks have to their human-made, audio-only counterparts,” said Grotesk Group.

“I feel like we are currently in a golden age of AI/music collaboration. Hopefully, it stays that way for as long as possible—where AI is just helping us rather than taking us over!”

Guy Lawrence, Disclosure

In creating the collection, Beatport—whose music marketplace, built in partnership with Define Creative, is built on Aventus, a parachain in the Polkadot ecosystem—played a finishing, uniting role. “The collaboration was a symphony of technology and creativity. Disclosure’s input guided Bronze AI in creating the audio variations, while Grotesk Group skillfully translated these auditory elements into visual artworks. Our role was to oversee and harmonize these contributions, ensuring a cohesive final product. Curation was a significant aspect of this project. We ensured that each pairing of audio and visual art was more than just a combination; it was a unified representation of the track’s essence,” they told us.

Reflecting on the intricate, collaborative process of creating Disclosure: Bronze Editions, Lawrence is cautiously hopeful about the future of AI in music. “I feel like we are currently in a golden age of AI/music collaboration. Hopefully, it stays that way for as long as possible—where AI is just helping us rather than taking us over! I like the idea that AI for musicians, at least in the digital world, can be akin to what that paperclip thing used to be like on Microsoft Word for writers—purely, just helping take the human-made work further. I’ve yet to hear an original piece of AI-generated music that moved me. It’s always a variation of a copy of a human-made piece of work. I’m sure that will change though, but for now, I’m loving the dynamic of collaboration between AI and our brains,” he said.

Dive Deep

Features & Guides