Will AI Music Have a Napster Moment?

By Edward Balassanian, Founder and CEO of Aimi

The year is 1999. The world is on edge as an impending technological meltdown is just around the corner, and a burgeoning global recession is rearing its ugly head. Amid all of the chaos, a little-known peer-to-peer music-sharing service is on the rise, one that went down as a pinnacle of the early aughts zeitgeist. 

Many people, especially GenX and millennials, remember Napster as a breakthrough in music-sharing services. The music-sharing service allowed anyone with a computer and music files to connect to a network of peers, creating a distributed database of music that was freely available to all users. At its peak, Napster had about 80 million registered users. The company seemed untouchable because users maintained the network itself, and it seemed out of reach from the legal arm of the music industry. 

AI companies are trying to copycat Napster’s illegal loophole

But, the fairytale proved short-lived. From the infamous lawsuit filed by Metallica and Dr. Dre to the RIAA, Napster was essentially sued into bankruptcy. Many stand behind these suits, as the service accommodated, encouraged, and largely glorified the theft of copyrighted material. 

Today, we can see familiar tension toward copyright laws within the AI sphere. AI models require massive data sets to be effectively trained. Since many generative music AI initiatives aim to mimic music industry professionals, the best data that can be used to train these models are often copyrighted. 

Piracy doesn’t pay

Efforts are already underway, under the disguise of a non-profit banner to usher in potential illegal streaming services. Like Napster before, these efforts will invariably raise the ire of the music industry, provoking lawsuits that will ultimately shut these services down for good. The very nature of these efforts will likely result in new legislation making the use of copyrighted material for training neural networks without permission illegal.

The lesson from Napster and various other companies, such as Limewire, is that piracy doesn’t pay, no matter how you try to frame it.

Investors, keep an eye out for real-life connections

While it might be tempting to invest in AI efforts that are seeking to train their models on pirated music, these efforts will likely run afoul of the RIAA and the music industry at large. In that case, artists will likely push back against these services, as we saw with Napster since their creative efforts will be expropriated in a manner that disenfranchises them.

Looking at the numerous generative music platforms and experiments cropping up, such as Harmonai and Playlist AI, we see the relationship between human creativity and AI being shaped before our eyes. Google’s MusicLM, which generates music from text descriptions, has acknowledged that they will keep the model off the market. So while you can’t produce your own music using the tool, Google has released samples to go toward research in the generative AI field.

For investors, pirating data for training a model is not a realistic business model. More importantly, the use of AI models in music to mimic the works of art of creative artists is problematic for many other reasons besides the high likelihood of getting sued. 

Music created by a purely generative AI is not necessarily an attractive business model. One of the core elements of a musician’s success is his or her fans. Without them, a meaningful connection to an artist is lacking, and in the case of AI, this organic creation fosters a relationship that cannot be regenerated. A singer’s broken heart pulls one’s heartstrings, and a starstruck teenager waits in line for tickets to their favorite artist’s show. Those looking to back AI-powered music platforms need to look for users’ ability to connect to the platform.  

Finding a harmonious balance between creators and technology

As AI becomes more prevalent in all industries, especially music, artists should be considered an ally for any music company to have on their side. It’s vital that business models elevate artists rather than marginalize them. If the sole goal of a generative AI solution is to create copyright-free music, the landscape of low-margin companies that already provide this without AI is clear evidence that these models are problematic. Instead, AI companies could target the creator by making creativity more accessible and less tedious.

About the author: 

With more than 25 years of experience in the technology industry, Edward Balassanian has founded and funded numerous startups ranging from operating systems to consumer packaged goods. Edward recently founded Aimi, a groundbreaking music platform that allows fans to enjoy immersive listening experiences while providing artists with a new medium to create and monetize AI-powered music.

Prior to Aimi, Edward founded startups including BeComm Corporation, a pioneer in building operating systems for digital devices, Be Labs, an IP-centric incubator, Vital Juice, a Food Manufacturing company, and Strings Inc, a content publishing platform, among others. A graduate of the University of Washington with Distinction in Computer Science, Edward started his career at Microsoft where he worked from 1990-1995. He is the named inventor of over 80 patents in technologies ranging from gesture interfaces, networking, distributed processing, and digital music synthesis.

The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.