The dilemmas of using artificial intelligence to bring dead people back to life – 7/5/2023 – Tech

An advertising campaign showing the late Elis Regina and her daughter Maria Rita performing a duet provoked antagonistic reactions on social media.

In Volkswagen’s play, the singer who died in the 1980s was brought back to life using artificial intelligence (AI). She appears driving a Kombi and singing “Como Nosso Pais”, by Belchior.

While many fans and netizens praised and were moved by the advertisement, others questioned whether it is ethical to use the image of a person who is no longer alive in a fictional context.

Glauco Arbix, a sociologist and impact coordinator at the Center for Artificial Intelligence at the University of São Paulo (USP), told BBC News Brasil that the subject is indeed controversial, either because it raises debates about the psychological effects of bringing dead people to life using technology or because it touches on issues such as consent, veracity and the finitude of life.

For Arbix, there are many risks in using AI in a non-transparent, informed or conscious way, especially when there is a spatial displacement or attribution of untrue statements to the portrayed person.

“It’s not because you can do it that you should do it”, he says. “It’s one thing to keep in your drawer a film by someone who died to watch it a few times, it’s another thing to recreate it. [a imagem dela] in new condition, as if she were still alive.”

According to the USP professor, our society is not ready to deal with this spatial and circumstantial displacement of deceased figures and doing so can be “disturbing” for some people.

“The finitude of life is sedimented in social history. Even for those who believe in life after death, it is something always more inaccessible and different from what we see now, for which we are not ready as a society.”

BBC Brazil sought Volkswagen for a position on the subject, but did not receive a response until the closing of this report.

‘It can destroy name and reputation’

Volkswagen’s campaign was not the first to use artificial intelligence to stage realities with people already dead.

In the movie “Rogue One: A Star Wars Story”, actress Carrie Fisher was also digitally recreated to appear as the young Princess Leia.

In June, musician Paul McCartney said that artificial intelligence had been used so that the voice of John Lennon – who died in 1980 and with whom he was a partner in the Beatles band – could be used in a new song.

The technology, also known as deepfake, is still frequently used to create fake videos involving celebrities and political figures.

In the case of the automaker’s advertisement, the artificial intelligence was trained specifically for Elis Regina’s facial recognition, unlike what is done in AI projects that use pre-trained technology from generic data.

According to the company, the AI ​​received extensive training with different technologies, combining the stuntwoman’s performance with Elis’s movements and images, to arrive at the result of the singer’s face in the advertisement.

For Arbix, although the Volkswagen video was made with the authorization and participation of Elis Regina’s daughter, this technology can also be used for dangerous purposes, distorting facts, and even in the pornography or pedophilia industry.

“The person can be subject to a recreation that can end up destroying his name and reputation”, he says. “But it also raises questions from the point of view of the integrity of family life.”

According to the sociologist, there is still no consensus among the medical community on the psychological effects of seeing or even talking through AI with loved ones who have passed away.

Several technology companies, including the American HereAfter AI, have developed technologies to create a digital version of someone. In this way, it would be possible to create an artificial dialogue with a deceased person using personal information, voice tools and advanced artificial intelligence.

“From the point of view of psychology, there are those who say that it can help maintain memory and bring comfort to the family. But there are also those who are totally against it”, says Glauco Arbix.

And there are even those who are already trying to protect themselves from this. Actor Robin Williams, who died in 2014, imposed a restriction on the use of his likeness for 25 years after his death in his will.

According to the information released, the American wanted to prevent his figure from being reproduced through holograms or other technologies for commercial purposes.

Image rights and consent

When it comes to image rights or consent, sociologist Glauco Arbix says he believes that Brazilian law already has all conflicts well resolved.

“The legislation and the way in which our society sees it today already account for the dilemma. Families have copyrights”, he says.

“Discussing whether, for example, Elis Regina would authorize the use of her image in this advertisement is naive, because she also did not authorize the publication of photos, but this issue is foreseen in the legislation.”

For Sara Suárez-Gonzalo, professor at the Open University of Catalonia and researcher on the subject, the debate must go further. For her, the consent of family members is not enough in cases like this.

“Even when they die, people are not mere things that others can do with as they please. This is why our societies consider it wrong to desecrate or disrespect the memory of the dead. In other words, we have certain moral obligations to the dead. , insofar as death does not necessarily imply that people cease to exist in a morally relevant way,” he said in an article published on The Conversation website.

According to Suárez-Gonzalo, the debate is even more complex when it involves bots that collect personal data to reproduce conversations with deceased people, since replicating someone’s personality “requires large amounts of personal information, such as data from social networks that reveal highly sensitive characteristics” .

The researcher also states that another ethical issue involved in the use of AI is accountability for the results of the technology, especially in the case of harmful effects.

If a bot, video or image created with technology, for example, causes damage to the mental health of a family member, who is responsible?

“It is essential to open a public debate that can better inform citizens and help us develop policy measures to make AI systems more open, socially fair and compatible with fundamental rights”, he says in the article.

This text was originally published here.