Synthetic Deepfake Music Plays The Spirit Of Time

It’s a nostalgic complaint of the Millennium: music is disappearing from our ears. The habit of fast consumption also takes music away from the person. Unforgettable melodies that descend deep into the soul, process into cells, etched into the brain, are no longer produced. They were replaced by artificial commercial music that captured the brain, with hypnotizing rhythms and frequencies that sneaked into the ear.

As music drifted from instruments to computers, the sphere of dominance of electronics expanded. Electronic music, which was influenced only by listening, but then not remembered, came to prominence. The DJ-made techno music was, in fact, like the siren of the “artificial reality” dystopia. No longer will instruments, notes, melodies, sounds, or emotions be as real as they used to be. Like synthetic foods, it is debatable how healthy synthetic music will feed the human soul. Electronic music is increasingly evolving into deepfake music in the age of artificial intelligence (AI). Composer, lyricist, even singer; are not human…

Age of marketing to ready customer

Marketing algorithms continue to amaze many people yet. If the holiday program at home is talked about, you look, after a few minutes, the holiday promotion messages of the tourist resort or travel agencies are on the phone. You’re looking for a mobile phone model on Google. Whatever website you open, here are mobile phone ads. The brand on the Billboard you pass by sends discount messages inviting you to its nearest store there.

It’s like an online eye and an ear, in close pursuit. Marketing has now stopped running big campaigns to impress everyone. It is making a point shot.  The person works targeted. He doesn’t bother to impress all consumers with a stunning advertising film or slogan and turn as large an audience as possible into a customer. He offers what he wants to the customer who is ready to buy.

The only option; HIT music

Here, as technology entered the business, music also ceased to be an art branch. It became more of a marketed commodity. Did Beethoven, Mozart or Bach, Elvis, Sinatra or The Beatles, even Michael Jackson, Sting or Pavorotti, produce music that “sold”? The artist would pour his own personality, style, emotions into melodies and produce his own music. Millions of people who connected with these melodies would follow him. And synthetic music, whatever is a HIT, derives that music. He is accused of theft of artifacts and unfair competition for copying and modifying everything that holds, sells.

Deepfake vocal channel on YouTube

Tech expert and blogger Andy Baio points out videos on vocal synthesis, a new deepfake channel on YouTube that pairs famous voices with unexpected dialogue. Speech syntheses from text produced by AI, imitating human voices, were very successful. Realistic sounds were produced by training the state-of-the-art neural network on a wide range of data. The examples were striking. Bob Dylan was singing Britney Spear’s famous “Hit me baby one more time.”

Famous thinker-writer Ayn Rand and Marxist sociologist Slavoj žižek, known for their philosophy of objectivism, duetted the lyrics of “I goot you baby” by famous rocker couple Sonny and Cher of the 60s and 70s. Former US President Bill Clinton was reading the lyrics to the song “Baby Got Back” by renowned hip hop artist Anthony Ray, known as Sir Mix-a-lot. Many of the videos were later remastered by fans, with music added to create fun and surreal musical mix. For example, American rapper The Notorious BIG’s “Juicy” is from the voice of former President Obama of the United States, and 50 Cent’s “In Da Club” is from George W. It was rearranged from Bush’s voice.

Copyright notice to YouTube

No one person remained indifferent to audio synthesizers posted on the Vocal Synthesis YouTube channel and fun amateur deepfake videos produced through them. As the middle of the year approached, at a time when the Covid-19 pandemic was at its peak, the world was dealing with a struggle for life, while deepfake music was perhaps facing a significant legal challenge for the first time. Because Jay-Z, who had a reputation as the king of Hip Hop music, had taken action on the grounds that his copyright had been infringed. Shawn Corey Carter, known by his stage name Jay-Z, was also the CEO of Roc-A-Fella Records, an American Music Company owned by Universal Music Group. He was also a strong business person in the music industry, co-founder of Roc Nation LLC, The 40/40 Club, and The New Jersey Nets.

An unidentified programmer produced two different videos in Vocal synthesis with Jay-Z’s deepfake voice, one singing William Shakespeare’s famous Hamlet tirade that began with “To be or not to be…” and the other as a cover for Billy Joel’s “We didn’t Start the Fire.” His agency, Roc Nation LL, which represents Jay-Z, posted a message to YouTube saying his artist’s voice had been imitated through AI and that the videos should be taken off the air. But YouTube, over objections, soon found Jay-Z’s copyright infringement application incomplete, allowing the videos to be temporarily rereleased until additional information came in.

Copyright infringement controversial

The removal of videos produced by Jay-Z’s synthetic voice from his YouTube channel has sparked legal controversy. “This content illegally uses an AI to imitate our customer’s voice, “the Roc Nation’s written notice on Twitter said, highlighting the use of AI in particular. However, the videos “Book of Genesis” and “Navy Seal”, which used the Jay-Z deepfake sound, continued to be published on the Vocal Synthesis YouTube channel. Moreover, opinions have also been raised suggesting that it is pointless to mention royalty for a classic work such as Hamlet.

Like many others, these voice Synthesis videos were created by feeding Google’s open-source “Tacotron 2” Text-To-Speech model with Jay-Z songs and lyrics. Synthetic audio was allowed to read pre-written text. It was also argued that if synthetic sound produces a different original content using the likeness of a celebrity, in this case there could be no copyright infringement. “It seems reasonable to consider a model and sound produced from copyrighted sound recordings as ‘derivative works,’” tech writer Andy Baio wrote in assessing the controversy. But is there copyright infringement? Like almost everything in the world of copyright, it depends on how and for what purpose it is used. According to Baio, an example of a record producer using Jay-Z’s voice on a new single without his permission could have been seen as a more overt legal violation in terms of copyright.

Protecting personal rights is a priority over copyrights

The algorithmic redesign of artists’ styles raises complex legal questions. Amanda Levendowski, a Georgetown Law professor and Founding Director of the Department of intellectual property and Information Policy, looks at the issue from a different angle. Levendowski points out that the Jay-Z example, with respect to the data used to create the sound, is that while the issue may seem like a copyright issue, in fact the main important issue is different. Levendowski, an intellectual property expert, thinks that because of the impersonation of the person, the incident is a matter closer to lawsuits related to the rights of the person. In the US, the right to protect a person’s name, image and personality from commercial abuse precedes copyright.

Levendowski recalls a 1988 Ford Motor Company commercial with another singer imitating the voice of singer and actress Bette Midler. In the case filed by Midler, she emphasizes that the Court of Appeal ruled that “sound imitation exploits one’s identity,” not for the protection of the voice. In another example in 1991, Samsung again dresses a robot like TV and movie star Vanna White for an ad. White is winning a lawsuit she filed against the company on the same grounds in California. Levendowski emphasizes that in the United States, copyright is federally governed, while personal rights are a national area of law. ‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎‎

Deepfake Music app at full speed; Jukebox

On YouTube, there was a deepfake struggle between Jay-Z and the Vocal Synthesis channel, while OpenAI, the nonprofit AI research company that Elon Musk co-founded, launched the deepfake Music app Jukebox.  The app has an algorithmic system that can produce music in the style of famous musicians such as Elvis Presley and Frank Sinatra, complete with lyrics.  These algorithms analyze an artist’s song collections and reveal the distinctive features of their musical style from their audio data. It then uses these patterns to create new sound.

Some of the songs created with Jukebox are incredibly good.  OpenAI’s website features many examples “in the style” of famous artists, including Sinatra and Elvis, Katy Perry, Stevie Wonder, Natalie Cole, Ella Fitzgerald. Jukebox accepts genre, artist and lyrics as input and creates new samples from scratch. Users can fine-tune the data up to the year the songs are released. After making these settings, if they want, they can create music similar to Katy Perry, for example, thanks to the app.

Does the spirit of time disappear along with music?

Technology leads to development and increases the quality of life of society along with the quality of products or services. However, deepfake technology is stopping the development of music, almost nailing it to the digital age. New sounds, new styles and melodies, how will they emerge now? How will new artists compete with deepfake music, which has given everyone their most liked music to date? How will it lead people to new options? How will future generations continue their emotional development, now listening to different versions of the same music? When the spirit of time loses its music, will it be matter?