South Korea has used AI to bring a dead superstar’s voice back on stage, but ethical concerns abound

It is a performance that is all the more extraordinary because the famous folk singer has died.

National broadcaster SBS plans to use artificial intelligence (AI) to revive Kim’s voice on a new program, “Competition of the Century: AI vs Human,” which airs later this week.

This 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 program that uses AI and holograms from late artists to pay tribute to their work.

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

Recent AI performances have excited fans of their music and technology, but others have expressed concern about the ethics and legitimacy 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?

Legendary Korean singer

Kim was just 31 years old when he passed away at the height of his career in 1996 after a series of hits, including ‘A Letter From a Private’, ‘Song of My Life’ and ‘In the Wilderness’.

His death is considered so shocking that some never accepted the official cause of suicide, but rather chose to believe conspiracy theories that he was killed.

Decades later, fans still gather in a street named in his honor, near his orphanage in the city of Daegu, though this year’s tributes were mostly held online due to the coronavirus pandemic.

The artist’s murals look across the street, where tourists sit on benches like guitars and listen to musicians playing his biggest hits.

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

When national broadcaster SBS announced it was using AI to recreate Kim’s voice in a new program airing this month, it sent fans into a frenzy.

There has been more than 145,000 times a one-minute promotional track of Kim singing ‘I Miss You’, a ballad released by Kim Bum-soo in 2002. since YouTube on YouTube. And a video showing the process behind the episode was more than seen 750,000 times since it was posted on January 6th.

“The recovered voice sounds a lot like him, as if Kim is recording it live,” said Kim Jou-yeon, a thirty-year-old Kim fan.

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

Show inspired by a Go contest

SMT producer Nam Sang-moon said the idea for an AI versus human competition came to him after he saw world champion Lee Se-dol in South Korean AI program HanDol in Go’s ancient strategy game in 2019 has.

Surprisingly, Lee won one of their three games.

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

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

World champion Lee Se-dol embarked on Google's AlphaGo program in 2016.
Lee’s game against HanDol reminded Nam of Lee’s earlier game against AlphaGo, an AI program developed by Google DeepMind in 2016. AlphaGo won four out of five games at the time, after which Lee said he “misjudged” the machine’s capabilities.

“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,” SBS producer Nam said in a telephone interview.

Nam felt that the Korean public was ready for an update and began compiling the six-part AI competition series, including the AI ​​execution of the late Kim Kwang-seok.

Kim recreates

According to Audio co-founder and chief operating officer Choi Hee-doo, the audio AI venture behind Kim’s reincarnated voice is Supertone, a South Korean venture founded in 2020 that offers AI audio solutions for content creators.

“For example, BTS is very busy these days, and it would be a shame if they could not participate in content due to lack of time. For example, if BTS uses our technology to make games or audiobooks or copy an animation, for example, they do not necessarily have to personally, ‘Choi said.

Supertone’s SVS (Singing Voice Synthesis) technology teaches voices by listening to multiple songs with matching notes and lyrics, Choi explained.

The system learned 100 songs by 20 singers before getting 10 Kim Kwang-seok songs to learn. It now knows his voice well enough to mimic the singer’s unique style and pronunciation, Choi said.

In the SBS show, the AI ​​Kim will not compete against a human singer – he will sing a duet with one. “We revived Kim Kwang-seok to show its ability,” said Nam, the show’s producer.

Instead, Ock Joo-hyun, the former lead singer of the girl group Fin.KL, will take on the AI ​​machine. Just as the AI ​​system taught Kim’s voice, it will be trained to imitate her as well.

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

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

Potential hazards of AI

South Korea, which is technically proficient, is considered to be at the forefront of AI technology, but there is a call for more regulation.

Within days of the launch of a petition last week, more than 376,000 people had signed the calls for tougher penalties for creators of profound videos, especially those who use the faces of female celebrities for pornography.
Amsterdam researchers, Deeptrace, counted more than 14,000 deep-faked videos online in 2019 and found that almost all – 96% – were non-consensual deep-fake pornography, using images of female celebrities, many from South Korea.
Repeating someone’s voice also carries serious risks, especially if false audio is used in misinformation campaigns and fraud. In 2019, scammers persuaded a driver to transfer hundreds of thousands of dollars in cash to an account using his boss’s voice through audio software.
“We see about $ 470 million a year in fraudulent losses, including bank transfers and telephone fraud. It’s huge,” Vinday Balasubramaniyan, Pindrop’s CEO and co-founder, told CNN last year.

Globally, countries and organizations are calling for legal reform, ethical guidelines and social agreements to prevent the abuse of AI.

UNESCO is working with its 193 member states to lay the ethical foundation for regulating the technology. Last September, he unveiled a draft of his “Recommendation on the Ethics of Artificial Intelligence”. The final report is expected to be submitted to UNESCO’s General Conference later this year.
In December, the Ministry of Science and Information, Communications Technology of South Korea released the ‘National AI Ethical Guideline’, a document setting out basic standards for people involved in the development and application of AI. It is said that AI ‘should be developed and used in accordance with its purpose and intent as a tool for human life, and that the process should also be ethical.’
The same month, the ministry released a roadmap for AI legislation, proposing to expand South Korea’s intellectual property rights. beyond works created by humans, to the investors and inventors behind AI creations.

Supertone chief technology officer Heo Hoon said he supported laws to regulate the industry. He said to mitigate the current risks, the company does not deal directly with the public, and prefers to work with companies that share its ethics.

“We are clearly aware of the possibility that our technology could be abused in the hands of the public,” Heo said. “I think the revival of the voice of the deceased would give a setback, which would hopefully start a social debate and eventually lead to legislation.”

The company has built some protection into its recordings by marking the sound that it is produced by the AI, but it is unlikely that casual listeners will know it.

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

Questions about ownership

The use of AI to create works that are traditionally produced by human creativity raises the issue of ownership. Does the copyright of the program hold if AI creates the works, or does it belong to the programmer?

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

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

Last year, singer Hayeon released ‘Eyes on You’, a single composed by an AI program. Human producers later refined the song according to her record label Enterarts.

Elsewhere, AI is used to create news stories, books, and even art. The creator is usually 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?” Ask Ko, the lawyer. “We need the necessary regulation that guarantees the safety of people, not an excessive one that hinders the development of AI technology.”

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

SBS paid a one-time fee to his family for offering his voice in the program, as with other members of the cast, Nam said. Parts of the show will be released on YouTube after it airs, but no SBS or Supertone plans to release Kim’s song as a single.

“We agreed not to officially release the songs featured in the show,” Supertone’s Choi said.

Kim fans will now only have to be content with playing the old songs that the real singer recorded through a microphone more than 20 years ago.

Sun Somi in Seoul also contributed to this story.

.Source