South Korea used AI to bring the voice of a dead superstar back on stage, but ethical concerns abound

It is an even more extraordinary feat because the famous folk singer is dead.

National broadcaster SBS plans to use artificial intelligence (AI) to bring Kim’s voice back to life in a new show, “Competition of the Century: AI vs. Human”, which will air later this week.

It is not the first time that AI has been used to revive a famous singer in South Korea.

In December, the music channel Mnet aired “One More Time”, a show that used AI and holograms by late artists to honor their work.

And on New Year’s Eve, K-Pop stars BTS performed online with an AI version of singer Shin Hae-chul, who died after surgery in 2014.

Recent AI performances have excited fans of its music and technology, but others have raised concerns about the ethics and legality of raising the voices of the dead. The creation of new works – or revived voices – by AI also raises copyright issues. Who is considered the owner? The creator of the AI ​​program or the AI ​​system itself?

Korean legendary singer

Kim was just 31 when he died in 1996 at the height of his career, after a series of successes, including “A Letter From a Private”, “Song of My Life” and “In the Wilderness”.

His death was considered so shocking that some never accepted the official cause of suicide, preferring to believe conspiracy theories that he was murdered.

Decades later, fans still congregate on a street named after him near his childhood home in the city of Daegu, although this year’s tributes were mostly held online due to the coronavirus pandemic.

The artist’s murals contemplate the street, where tourists sit on guitar-shaped benches and listen to musicians playing their greatest hits.

A hologram concert by the late South Korean singer Kim Kwang-seok was held in his hometown, Daegu, on June 10, 2016.
Tourists can visit a street dedicated to Kim Kwang-seok in the city of Daegu.

So when national broadcaster SBS announced that it was using AI to recreate Kim’s voice on a new show that would air this month, it left fans in a frenzy.

A one-minute promotional clip of Kim singing “I Miss You”, a ballad launched by Kim Bum-soo in 2002, has been viewed more than 145,000 times on YouTube since December. And a video showing the process behind the episode was seen over 750,000 times since it was posted on January 6.

“The recovered voice sounds very similar to his, as if Kim had recorded it live,” said Kim Jou-yeon, a Kim fan for 30 years.

This even impressed a young fan Park Hye-hyun, who was just two when Kim died. “One of my wishes was to hear more songs in Kim’s voice … I cried,” she said.

Show inspired by a Go game

SBS producer Nam Sang-moon said the idea of ​​an AI competition against humans came after he saw world champion Lee Se-dol take on South Korea’s AI program HanDol in the former 2019 strategy game Go .

Surprisingly, Lee won one of his three games.

Just a month earlier, Lee had announced his retirement from the professional competition Go, saying that AI was “an entity that cannot be defeated”.

Go requires two players to place black and white stones on a 19-by-19 grid. Each competitor tries to claim most of the territory by surrounding his opponent’s pieces so that they are removed from the board.

Go world champion Lee Se-dol accepted Google’s AlphaGo program in 2016.
Lee’s game against HanDol reminded Nam of Lee’s previous game against AlphaGo, an AI program developed by Google DeepMind, in 2016. At that time, AlphaGo won four out of five games, and then Lee said he had “misjudged “the capabilities of the machine.

“Lee-AlphaGo was a big shock followed by a short but huge interest in AI, which is still being developed somewhere, but we are not yet familiar with it,” said SBS Nam producer, in a telephone interview.

Nam realized that the Korean audience was ready for an update and started putting together a series of six-part AI competitions, including the late Kim Kwang-seok’s AI performance.

Recreating Kim

The audio AI company behind Kim’s reincarnated voice is Supertone, a South Korean startup founded in 2020 that provides AI audio solutions to content creators, according to co-founder and chief operating officer Choi Hee- doo.

“For example, BTS is very busy today and it would be a shame if they couldn’t participate in the content due to lack of time. So, if BTS uses our technology to make games or audiobooks or dub an animation, for example, they wouldn’t necessarily have to record in person, “said Choi.

Supertone’s Singing Voice Synthesis (SVS) technology learns voices by listening to various songs with corresponding notes and lyrics, explained Choi.

The system learned 100 songs from 20 singers before receiving 10 songs from Kim Kwang-seok to learn. Now he knows his voice well enough to imitate the singer’s unique style and pronunciation, said Choi.

On the SBS show, AI Kim will not be competing against a human singer – he will be singing a duet with one. “We revived Kim Kwang-seok to show his ability,” said Nam, the show’s producer.

Instead, Ock Joo-hyun, the former lead singer of the female band Fin.KL, will take over the AI ​​machine. Just as the AI ​​system has learned Kim’s voice, he will be trained to imitate her as well.

While some may consider a competition between an AI singer and a human to be harmless fun, others warn that the technology poses threats that need to be addressed by stricter guidelines and regulations.

Ock Joo-hyun gained fame in the 1990s as the lead singer of Fin.KL

Potential hazards of AI

South Korea, tech savvy, is considered to be at the forefront of AI technology, but there are demands for more regulation.

Within days of launching a petition last week, more than 376,000 people signed calls for harsher punishments for fake video makers, especially those who use the faces of female celebrities for pornography.
Deeptrace researchers from Amsterdam counted more than 14,000 deepfake videos online in 2019 and found that almost all – 96% – were non-consensual deepfake porn using images of female celebrities, many of them from South Korea.
Replicating someone’s voice also poses serious risks, especially if fake audio is used in misinformation and fraud campaigns. In 2019, scammers convinced an executive to transfer hundreds of thousands of dollars in cash to an account, falsifying his boss’s voice using audio software.
“Each year, we see about $ 470 million in losses from fraud, including wire transfers and phone scams. It’s a huge scale,” Vijay Balasubramaniyan, CEO and co-founder of Pindrop, told CNN last year.

All over the world, nations and organizations are calling for legal reforms, ethical guidelines and social agreements to prevent the misuse of AI.

UNESCO is working with its 193 member states to establish the ethical foundations for regulating technology. Last September, he released a draft of his “Recommendation on the Ethics of Artificial Intelligence”. The final report is due to be presented to the UNESCO General Conference later this year.
In December, South Korea’s Ministry of Science and Information and Communication Technology released the “National AI Ethical Guideline”, a document that sets the basic standards for people involved in the development and application of AI. He says that AI must be “developed and used according to its purpose and intention as a tool for human life, and that the process must also be ethical”.
In the same month, the ministry released what it calls a roadmap for AI legislation, which proposes to expand South Korea’s intellectual property rights. as well as works created by humans for the investors and inventors behind AI creations.

Supertone’s chief technology officer, Heo Hoon, said he supports the laws that regulate the industry. He said that to mitigate current risks the company does not deal directly with the public, preferring to work with businesses that share its ethics.

“We are clearly aware of the possibility of our technology being misused when in the public hands,” said Heo. “I think reviving the voice of the deceased would have a negative reaction, which would hopefully start a social debate and lead conclusively to legislation.”

The company has created a level of protection in its recordings, marking the audio as being produced by its AI, although casual listeners are unlikely to know.

“We have watermark technology, which is information planted in audio, which cannot be heard, but tracks where it was made and how it is being distributed,” said Heo.

Ownership issues

The use of AI to create works traditionally produced by human creativity raises the question of ownership. If AI creates the works, does the program own the copyright or does it belong to the programmer?

Lawyer Ko Hwan-kyoung, an expert in AI and data protection, said the issue needs to be addressed as AI becomes more advanced.

“What is interesting about AI is that it is evolving to a level where it can write and compose by studying data,” said Ko.

Last year, singer Hayeon released “Eyes on You”, a single composed by an AI program. Human producers later refined the song, according to their label, Enterarts.

Elsewhere, AI has been used to create new stories, books and even art. Usually, the creator is considered the owner of the work, but in the case of AI it is less clear.

“Are we going to recognize AI as a legal entity with legal personality like humans and grant copyright?” Ko, the lawyer, asked. “We need the necessary regulation that guarantees human security, not excessive regulation that gets in the way of the development of AI technology.”

In the case of Kim Kwang-seok, producer Nam said that SBS obtained the consent of Kim’s family to reproduce his voice before proceeding with the show.

SBS paid a one-time fee for his family for presenting his voice on the show, as they did with other cast members, Nam said. Parts of the show will be published on YouTube after it airs, but neither SBS nor Supertone plans to release Kim’s song as a single.

“We agreed not to officially release the songs featured on the show,” said Choi of Supertone.

So for now, Kim’s fans will have to settle for playing the old songs recorded by the real singer on a microphone for over 20 years.

Seoul’s Son Somi also contributed to this story.

.Source