Tag Archive for Music

Is China leading the next music disruption?

Are AI-powered music-generating and streaming platforms from China part of the next disruption? If we consider 1995-2005 as the “resistance era” against streaming (i.e., major labels clinging to physical media profits through very creative methods, in yellow below), perhaps we can consider 2005-2025 to be the “simmering period” of streaming consolidation. We are now witnessing the next paradigm shift led by AI-powered platforms (in pink below).

Of special note are Mureka and Melodio, developed by China’s Kunlun Tech, their calculated disregard for copyright fueling their ascent. Mureka’s advanced model can generate music from “reference tracks” which could be an uploaded track or a YouTube link, effectively crowdsourcing training data.

So, what happens to copyright in this case? I bit the bullet, subscribed to the “Pro” plan and proceeded to upload a track from my album. I did not encounter any content filters or copyright detection tools to stop me from uploading the track. Could I have uploaded a work that I didn’t have the rights to? If I’m not mistaken, at the very basic level, Mureka is asking for the user to become complicit in training the model in real time with copyrighted material.

To be fair, American AI-powered music-generating platform Suno and Udio have also been vague about training data and are currently being sued by multiple labels. 1 But they do not ask the user to upload tracks or use YouTube links as reference material.

Melodio seems to be more like Endel, a lean-back listening experience as the platform creates “an endless stream of perfect ambient music.” 2 This can further push the divergence of functional music from the rest of the recorded music output, allowing low-effort background music to be replaced by AI-powered music streaming platforms, reducing royalty payouts by minimizing human labor. Platforms can retain more revenue as production costs plummet. Again, copyright issues of its training data are currently not discussed.

How good are the Chinese AI-powered music models? Will they be popular? Regardless, at the very least, they are heating up the AI competition at the expense of rights holders.

OpenAI (now of Ghibli fame) has grabbed this moment to highlight the dangers of losing out to China. Regarding the White House’s development of an AI Action Plan3, OpenAI commented on March 13, 20205: “Applying the fair use doctrine to AI is not only a matter of American competitiveness—it’s a matter of national security. The rapid advances seen with the PRC’s DeepSeek, among other recent developments, show that America’s lead on frontier AI is far from guaranteed. Given concerted state support for critical industries and infrastructure projects, there’s little doubt that the PRC’s AI developers will enjoy unfettered access to data—including copyrighted data—that will improve their models. If the PRC’s developers have unfettered access to data and American companies are left without fair use access, the race for AI is effectively over. America loses, as does the success of democratic AI. Ultimately, access to more data from the widest possible range of sources will ensure more access to more powerful innovations that deliver even more knowledge.”4

But is “fair use” the right approach? On March 14th, RIAA commented: “Existing U.S. copyright law is sufficient to address copyright concerns arising from AI. In particular, our fair use doctrine…provides a manageable and nuanced approach to addressing liability for unauthorized uses of copyrighted works to train an AI system. As noted above, one court has already ruled that fair use did not apply to the type of AI training at issue in that case.”5

The rapid pace of development of AI-powered platforms around the world is challenging the music industry to reconcile technological progress with ethical and legal guardrails. If history is any indication, we are in for decades of chaos. In no scenario do I see the valuation of digital music increasing.

That said, there is enormous demand for music as a form of communication which fosters social bonding. From micro-events like themed house concerts to large-scale arena concerts, live music is booming. People want to get together. Fans want merch to show solidarity with their favorite artists as well as with fellow fans. Although it’s not always clear how all that translates to artists actually being able to make a living off of music, it’s still a very positive sign that music truly matters. Streaming is just one part of that eco-system.

4/12/2025: Editing to add that on April 1 2025, Udio also released a feature where the user can upload a reference track.

4/13/2025: Editing to add a link to AI 2027, prediction scenarios of AI impact written by Daniel Kokotajlo, Scott Alexander, Thomas Larsen, Eli Lifland, Romeo Dean. AI-powered music platforms fit right in.

  1. Tencer, D., & Tencer, D. (2024, June 24). Major record companies sue AI music generators Suno, Udio for ‘mass infringement’ of copyright. Music Business Worldwide. https://www.musicbusinessworldwide.com/major-record-companies-sue-ai-music-generators-suno-udio-for-mass-infringement-of-copyright/ ↩︎
  2. Melodio AI |Vibe Your Moment – Official website. (n.d.). https://www.melodio.ai/ ↩︎
  3. The White House. (2025, March 24). Public comment invited on Artificial intelligence action plan. https://www.whitehouse.gov/briefings-statements/2025/02/public-comment-invited-on-artificial-intelligence-action-plan/ ↩︎
  4. OpenAI’s proposals for the U.S. AI Action Plan. (n.d.). OpenAI. https://openai.com/global-affairs/openai-proposals-for-the-us-ai-action-plan/ ↩︎
  5. Leibfreid, M. (2025, March 20). RIAA leads music community in AI filing – US creative leadership paves the way for innovation – RIAA. RIAA. https://www.riaa.com/icymi-riaa-leads-music-community-in-ai-filing-us-creative-leadership-paves-the-way-for-innovation/ ↩︎

Guiding Musical Conversation through Structured Improvisation

(For my students in 2025 Composing for Improvisers class at Bennington College)

March 14, 2025 Kyoko Kitamura

In this class, we approach Structured improvisation as one of many ways to create an environment where participants can engage in a musical conversation using a shared set of building blocks. (Many words can be used in place of “building blocks”, e.g., parameters, conditions, rules, etc.). The building blocks can help shape the improvisation, much like the framework of a language or the topic of a discussion. While there are infinite possibilities, some of the most common building blocks include:

  • Melodic idea
  • Chord changes
  • Time signature
  • Rhythmic feel (e.g., swing)
  • Tempo
  • Rhythmic figures
  • Pitch
  • Timbre (e.g., “guttural”)
  • Time duration (e.g., stopwatch music, durational music)
  • Other sound sources (e.g., rocks, books)
  • Volume
  • Speed changes
  • Text (e.g., setting a poem to music)
  • Key words
  • Agreed upon conduction signals

They provide both support and contrast, ensuring that improvisation has direction while leaving room for spontaneity. If participants have played together before, they may instinctively understand certain shared musical ideas, but establishing these building blocks will pretty much ensure that the musical conversation will develop meaningfully, even with participants who are new to the ensemble.

It will also often provide an ending. A well-shaped ending can elevate an entire piece, yet it often receives less attention than the beginning or the middle during the composing process. By understanding the importance and the role of the future ending, we can sometimes be more in the present.

By intentionally shaping these parameters, we harness the strengths of the group. While there may be a main composer guiding the structure, every participant is also an improviser-composer, contributing in real time. We introduce elements of unpredictability through individual decision-making (those dopamine hits that keep us engaged!) and the work gains the potential to become much greater than the sum of its parts.

Through this process, the ensemble builds its own vocabulary, expanding its range of shared building blocks and deepening its musical connection. For some, this will be an entirely oral and aural experience; for others, written notation will play a role. Many working musicians today will engage with a combination of both.

Ideally, we want to surround ourselves with musicians who are interested in exploring this music together—and then just keep playing. Practice, play, repeat. You can experiment with the range of building blocks, from none (free improv) to fully-notated material. More than the final product, it is the process itself—the act of creating, listening, and responding in real time—that defines this music.

Writing can play a crucial role in this process, offering a way to revisit, refine, and solidify musical ideas, allowing us to track our progress and communicate ideas with others. Writing can travel over time and geographical distance; however, this can also introduce challenges, as written material tends to take on a life of its own and can be misinterpreted. Therefore, an oral component is essential. The strongest musical bond, in this type of music, is the connection formed by those playing in the same room.

This is communal music with its own deep histories. Even when solo music emerges from this practice, it remains deeply rooted in the communal experience. My own trajectory is via Black music, having been mentored by many great Black musicians, along with my background in European classical music and my Japanese heritage. It will be different for every one of us, allowing us to develop our own individual sound. Think of structured improvisation as the sweet spot between preparation and spontaneity, built on a foundation of personal and communal history, allowing for creative freedom within meaningful interaction with other great players. Above all, strive to make each other sound great and enjoy the process.

How Do We Find Music?

Ruben Whitaker's chart tracking exposure points to the music they listen to.
Ruben Whitaker's chart tracking exposure points to the music they listen to.
(Above chart by Ruben Whitaker)

How do we find music? Recommendations from friends? Movies? Video games? Algorithm? Music is everywhere. But when 120,000 tracks are being uploaded every day to digital streaming platforms1 and Spotify gives us access to over 100 million tracks,2 the chances of us encountering a piece of music become surprisingly low. 

Yet, we find them.

This semester at Bennington College, my class “Musical Taste and Monetization” took another deep dive into music discovery methods. The image above is a page from a project by Ruben Whitaker, recent graduate of Bennington College, who analyzed years of their playlists and made the following observation: “Friend recommendation + Algorithm seems to be a powerful phenomenon of exposure especially if the artist and sub-genre are completely new…I was surprised at the low number of artists and songs I encountered through algorithms, this highlights that peer-to-peer recommendation is strong…” 

Ruben continues: “I think a lot of people listen to…music they know personally or locally because it’s more of a personal relationship. I think this aligns with the hypothesis of music being developed for social bonding. This is something to think about in terms of how to build strong musical networks and offer something that AI music or mass-produced popular tracks cannot.”

When I first designed this course back in 2021, I anticipated students to rely heavily on the Spotify algorithm. How wrong I was! Although Spotify is the main platform many use to listen to music, the ways in which students discover music were remarkably diverse and creative, with person-to-person communication and live concerts playing huge roles, that I had to completely rethink and recalibrate all I had thought I knew. 

Daniel Ek recently made comments about the cost of creating content, and my class too had previously reached a similar conclusion specifically regarding digital content: anything that can be reproduced digitally ad infinitum with a click of a button skews the traditional supply and demand curve. We discussed what it means to live in a society surrounded by technologies of abundance: “New technologies continue to democratize, decentralize, and disrupt production, offering the possibility that scarcity will be a thing of the past for many industries. We call these technologies of abundance. But our economy and our legal institutions are based on scarcity.” 3  

Music is so much more than digital content. Reexamining the non-digital component of music feels especially important right now. Music has been integral in every culture since the dawn of humanity, with many researchers highlighting the relationship between music and social bonding.4 In that spirit, I continue to make music and explore the opportunities of social bonding through music. The next generation already understands this. These students are creating our future. We have so much to learn from them. 

[All images by Ruben Whitaker, used with permission. Ruben Whitaker is a Chicago-based songwriter and multi-instrumentalist, and a graduate of Bennington College. During college, Ruben ran a radio show called “Folkloric Geography” which showcased music from around the world. Ruben has several releases under the name “Ruben Jai” including the the album Present for You and the singles Washburn and Neighborhood Planets.]

  1. Stassen, M., & Stassen, M. (2023b, May 24). There are now 120,000 new tracks hitting music streaming services each day. Music Business Worldwide. https://www.musicbusinessworldwide.com/there-are-now-120000-new-tracks-hitting-music-streaming-services-each-day/ 
  2. Konstan, E. & Spotify Technology S.A. (2023). FORM 20-F. In UNITED STATES SECURITIES AND EXCHANGE COMMISSION [Report], P.35, https://s29.q4cdn.com/175625835/files/doc_financials/2023/ar/26aaaf29-7cd9-4a5d-ab1f-b06277f5f2a5.pdf 
  3. Desai DR and Lemley MA (2023) Editorial: Scarcity, regulation, and the abundance society.
    Front. Res. Metr. Anal. 7:1104460. doi: 10.3389/frma.2022.1104460 
  4. Mehr, S. A., Krasnow, M. M., Bryant, G. A., & Hagen, E. H. (2021b). Origins of music in credible signaling. Behavioral and Brain Sciences, 44. https://doi.org/10.1017/s0140525x20000345 

(This article originally was written for, and appeared on, New York Jazz Workshop blog)