Music industry veteran Mike Caren is joining AI and metaverse tech startup Futureverse as a founding partner of JEN, the company’s text-to-music AI generator.
The move is part of Futureverse’s strategy to develop AI music generators with the input of music industry insiders. According to the company, Caren will lead music industry alignment and expansion at JEN.
“With Caren on board, JEN has both the music industry access and expertise, combined with a leading music AI technology platform designed to collaborate with artists, producers and creators,” the company said in a statement on Monday (October 30).
Caren has written and produced records with numerous artists such as Beyoncé, Kanye West, Bruno Mars and others. A veteran of Elektra Records and Atlantic Records, he served as President, Worldwide A&R and later as Creative Officer at Warner Music Group. But Caren is best known as the Founder and CEO of music recording and publishing companies Artist Publishing Group (APG) and Artist Partner Group.
APG’s long-running joint venture with Warner came to an end in 2020, when Warner’s Atlantic Records bought APG’s existing recordings catalog and artist roster, moving them into Atlantic and leaving Caren’s APG as a fully independent record company.
Although it was known for its successes in pop and hip-hop, APG expanded into the alternative and rock genres with the signing of a publishing deal with writer and producer WZRD BLD (Drew Fulk) in 2021. APG also entered into a partnership with Luke Mitzman’s 100 Management to build a boutique company for music producers and entrepreneurs that “will take on a small, select roster of clients that balance creative and business skills”.
Caren will be joining Futureverse co-founders Shara Senderoff and Aaron McDonald, as well as Futureverse Head of AI Research Dr. Alex Wang, as a founding member of the JEN project.
“We’ve been working with Mike on JEN for nearly a year now, and are honored to formally announce his addition to our founding team,” Senderoff and McDonald said in a joint statement.
“Mike’s ethos and relentless mission to innovate the industry in terms of technology, business model and creator opportunity aligns identically with ours. We strongly believe that including producers and artists into the process of developing AI tools is critical to its success and expansion.”
Caren added: “Shara, Aaron and our team of advanced technologists are deeply aware of the concerns and fear that many music artists and executives have around AI development. With JEN, we are committed to shaping the model with talent at our side as we further develop the technology to the standards of commercial music production. We aim to put these tools directly in the hands of the creators to ultimately empower and enhance a producer or artist’s creative process.”
Futureverse, which was created out of 11 startups in 2022, announced last August the development of JEN-1, which it touts as a “universal high-fidelity model for text-to-music generation.”
The company published a research paper authored by its Altered State Machine innovation team that delved into the complexity of music’s wide frequency range and the need for higher sampling rates to capture its nuances effectively.
“With JEN, we are committed to shaping the model with talent at our side as we further develop the technology to the standards of commercial music production. We aim to put these tools directly in the hands of the creators to ultimately empower and enhance a producer or artist’s creative process.”
Mike Caren, APG/Futureverse
Futureverse claims that its research paper and the method adopted by its researchers surpassed that of industry giants Google and Facebook.
JEN-1 represents an “exceedingly efficient approach that demonstrates higher-quality outputs than the state-of-the-art baselines (such as MusicLM and MusicGen) previously released by Google and Facebook,” the company said in a statement.
Earlier this summer the company raised USD $54 million in a series A funding round from backers including 10T Holdings and Ripple.
Futureverse also announced the release of a second research paper on JEN, which delves into the topic of separating generated audio into controllable stems. The company also released its “JEN Doctrine,” a set of principles designed to train AI models with a “fair and equitable approach.”
Among the principles are transparency about how and on what AI was trained; equitable compensation for rights holders, artists and users; and limitations on generated audio based on copyright.Music Business Worldwide
