Artificial Intelligence for Music Production

Audialab, an innovative technology company specializing in using AI to empower musicians and music producers, has debuted Deep Sampler, a new AI-powered tool that represents a paradigm shift in professional sound design. Deep Sampler is a novel technology that allows musicians to generate an infinite number of unique, never-before-heard sounds from one input sound and manipulate them endlessly using Artificial Intelligence.

“Before deep sampling, there were only two ways for music creatives to get a new sound: recording a sound or synthesizing it. There’s now a third way,” explains co-founder of Audialab Berkeley Malagon, an AI researcher who previously worked at NASA, Disney, and Microsoft. Malagon led the development of Deep Sampler with a team of engineers and musicians at Audialab. “Deep Sampler can analyze any existing sound and transform it into completely novel ones that couldn’t have existed any other way.“

“Artificial Intelligence is accelerating at a rapid pace, and our mission is to bring the power of those developments to producers and musicians worldwide while ensuring the human creative spirit is protected,” said Mike Chen, the other co-founder of Audialab, technologist, and member of KIMBEMIKE. “We are musicians, computer scientists, and ethicists ourselves, and AI needs to be taken very seriously by the industry. Our core mission is to develop AI tools that enhance the creativity of professional musicians and producers responsibly.” Previously, Chen co-founded several successful advanced technology companies, including Made in Space and the AI company, Magic.

Mike Chen and Kim Montoya, members of the alt-pop duo KIMBEMIKE, have been co-writing, co-producing, recording, and performing music since 2020. KIMBEMIKE’s debut single “Galaxies” represents a historic moment in music production as the first song ever to use Deep Sampling technology. Listen now to Galaxies: https://kimbemike.lnk.to/Galaxies

7 thoughts on “Artificial Intelligence for Music Production”

  1. First they came for us writers, and I did nothing.

    Then they came for the graphic artists and the musicians and I did nothing (other than say, “Welcome to the club”).

    When they come for the lawyers and the politicians I may smile–but they probably get to go last.

  2. Depends on what you mean by “Music”. Pop? Rap?? 1930’s Big Band Sound: Glenn Miller, Kay Kyser, or Ray Noble? 18th-century Sturm und Drang reborn (there wasn’t very much of it written; ignore any performances of the treacly London Mozart Players). I know that would mean producing Classical-style music without the use of performing instrumentalists.

  3. Music composition and Artificial Intelligence are perfect complements. Music has no intellectual content.

  4. I assume this will be done for all instruments and voices. The coming Cambrian Explosion of music.

    • Depends on what you mean by “Music”. Pop? Rap?? 1930’s Big Band Sound: Glenn Miller, Kay Kyser, or Ray Noble? 18th-century Sturm und Drang reborn (there wasn’t very much of it written; ignore any performances of the treacly London Mozart Players). I know that would mean producing Classical-style music without the use of performing instrumentalists.

  5. Sound != music

    Generating music that doesn’t sound sterile is another ballgame.
    Not that popular music of today is much of a benchmark.
    I have seen some serious work at deconstructing and replicating (with software) what professional musicians actually do when they interact. It’s a lot about timing and dynamics on a fine level. It takes a super talented human brain decades to develop those skills. An average brain can’t do it at all.

    We have all seen those super-human AIs with domain specific skills learned in a few months. However, music is interaction between many actors so the number of permutations to process are probably much larger than your average go-game or computer game. So, it can be done but it will take time. I wonder if there is enough economic incentive to push AI into wizzard level music skills. The general population seems content with todays generic stuff.

Comments are closed.