metamorworks -

Dissecting the musicians' open letter to AI vendors

The letter attempts to play on public sympathy and move AI technology vendors to take some action toward fairness. It also brings up the question of data ownership.

Another open letter has called out generative AI system creators for allegedly infringing on intellectual property.

More than 200 musicians, including well-known artists such as Billie Eilish, Nicki Minaj and Camila Cabello, have signed an open letter asking AI developers, tech firms and digital platform providers to stop using AI technology to obtain artists' voices and likenesses.

The letter, released Tuesday, accuses powerful AI companies of using the musicians' work to train AI models without permission. The artists also contend that companies are trying to craft AI-created sounds and images to contaminate the royalty pools artists receive.

"This assault on human creativity must be stopped," the letter reads.

Innovating despite problems

The letter comes as AI vendors including OpenAI, Microsoft and Google face lawsuits accusing them of using copyrighted works to train their AI systems.

The New York Times is in litigation against OpenAI and Microsoft for infringing its copyrighted work and training its AI system on the publication's subscription-only articles.

Amid the various lawsuits, AI vendors continue to innovate.

Most recently, OpenAI revealed it developed Voice Engine, a model that uses text input and 15 seconds of audio to generate speech.

While Voice Engine is currently in preview and testing with education and health IT users, the AI vendor said it will not release the technology to the general audience until it determines it is trustworthy for wide use.

Meanwhile, some observers see the musicians' letter, which follows similar moves by authors and artists, as a natural reaction to the rapid spread of generative AI technology. And some see positive uses for generative AI technology for artists.

"Artificial intelligence is both a threat and an opportunity to musicians and these artists," Futurum Group analyst Cory Johnson said. He added that the technology is coming to the edge, with new semiconductors from AMD and Qualcomm that will bring native generative AI capabilities to millions of smartphones and PCs.

"It's clear that we are moving into a phase of AI development in which virtually any type of human voice-generated sound will be replicable," said Michael Bennett, AI policy advisor at Northeastern University. "It certainly behooves professional artists and organizations and alliances to which they belong to do everything they can to counteract these developments and safeguard the opportunities that they have to continue making art in the future in ways that allow them to have a decent living."

The question of whether these tools will be used for good or bad is yet to be answered, Johnson said.

The music industry has been accused of misusing new technology before. For example, the LinnDrum machine upended the music industry and supplanted some human drummers' livelihoods in the late 1900s, though since then electronic music has flourished while human-produced music has also remained healthy.

"AI could be worse," Johnson added.

Public sympathy

Thus, while the musicians' open letter may seem like just another complaint against AI developers and creators of generative AI systems, it serves some purposes, Bennett said.

For one, absent a significant court ruling on copyright infringement and fair use, the open letter is a campaign that attempts to play on public sympathy, he said.

In trying to convince the public, the musicians might pressure the tech vendors and AI developers to make a move based on perceived fairness.

"If artists can get those companies to do something before a decision from the court ... then that behooves artists," Bennett continued.

A public sympathy campaign also targets legislators and could lead to laws at the state level to protect musicians' livelihoods.

While it's unclear whether this open letter from the musicians will be effective, there is evidence that public opinion is enough to move big technology companies, said Juliette Powell, an author and adjunct NYU professor who has written about AI.

For example, after Amazon released its facial recognition tool, people were falsely accused of crimes they didn't commit. Human rights groups and engineers released multiple open letters calling for a change. The letters led Amazon shareholders to vote on whether they should discontinue the technology.

The shareholders decided to continue selling the facial recognition tool until the public outcry over police brutality videos that emerged in 2020. After that, Amazon said it would stop providing the technology to police for a year. While Amazon did not provide a specific reason for the momentary ban, the vendor decision came about a month after a video was released showing a Minneapolis police officer kneeling on George Floyd's neck during an arrest.

"When there's enough public pressure that is on the board, where all of a sudden it's going to be their bottom line and their reputations, there are examples where it does make a difference," Powell said.

Since the initial moratorium, Amazon has resumed providing the tool to law enforcement. The Department of Justice revealed earlier this year that the FBI is in the initial phase of using the software.

Data ownership and augmentation

Furthermore, the musicians' open letter steers the conversation toward the question of data ownership, she said.

People should have control over our own data.
Juliette PowellAuthor and professor, NYU

"People should have control over our own data," Powell continued. "That is a fundamental human right."

While the musicians' open letter is another example of prominent people challenging big AI companies, a piece of the conversation is missing, said Davi Ottenheimer, vice president of trust and digital ethics at private data storage firm Inrupt.

"People don't understand that technology could solve some of these problems," Ottenheimer said. "The technology is ripe for an information revolution that includes music."

For creators of the AI systems, the technology can make mundane work more efficient and is the next step of innovation. For example, in the case of the Times lawsuit, Microsoft argued that the VCR did not destroy the Hollywood movie industry and, thus, it is unlikely LLMs will force the newspaper out of business.

AI is thus expected to be a tool for creativity in every industry, Johnson said.

"It can take a creator from ideation to creation almost instantly," he said. He added that much in the same way jazz legend Charlie Parker imitated or used parts of the work of another jazz great, Lester Young; or hip hop artists sampled funk innovator James Brown's "Funky Drummer," cutting-edge creative artists will be standing on the shoulders of AI for the music of the future.

The next step for musicians and those who feel victimized by AI technology, then, is to figure out how to augment their work with the technology.

For example, musicians can try to use AI capabilities to distribute their music, therefore using the technology to augment their craft.

"If musicians can use the AI and the technology in ways to augment themselves, and [the technology vendors can] use AI with their permission to augment their music as a derivative, I think everyone would be happy," Ottenheimer said.

However, if musicians are instead asking for royalties in exchange for technology vendors taking and using their music with AI technology, then that feels too much like unequal labor, he added.

Moreover, tech vendors are perpetuating the use of the technology in a way that makes it seem too domineering, he continued.

Instead of showing the value of the work and the data they're using, vendors are claiming not to gain value from the technology and just allowing others to derive value from it.

"What we should be asking AI companies to do is to innovate, as opposed to asking us to believe that they're innovating," Ottenheimer said.

Esther Ajao is a TechTarget Editorial news writer and podcast host covering artificial intelligence software and systems.

Dig Deeper on AI technologies

Business Analytics
Data Management