Have you ever wondered what it would be like to talk with a loved one who has died? To say goodbye that you never had the chance to do, or to fix something that you didn’t have time to do? As people, we often carry the weight of unfinished conversations, unresolved conflicts, and unspoken words – and we long for that last opportunity.
Now we live in a world where AI technology evolves to the point that it can grant us this elusive chance: to reconnect with those who have crossed the border. Companies in China, for example, have been exploit the power of AI to resurrect the dead by bringing together large amounts of data, photos, recordings and messages, to form chat programs capable of imitating the speech and behavior of deceased individuals.
And while the potential emotional benefits of this technology may be obvious, it also reveals many concerns, such as the consequences and ethical limitations associated with these AI-based products. Does using AI to resurrect the dead cross the boundaries of what is morally acceptable, or is it simply a natural progression in how this emerging technology is developing?
Ljubisa Bojic is a senior researcher at the Institute of Philosophy and Social Theory at the University of Belgrade and has worked on a number of projects that explore the intersection between AI and ethics. According to him, examples of such AI products above all raise the question of privacy and consent.
“The deceased, for example, cannot give consent to the use of their personal information or aspects of their personality. Furthermore, given such cases and the context of our research at the level Digital Society LaboratoryInterestingly, as AI models like GPT-4 begin to exhibit traits suggestive of self-awareness and subjectivity, the ethical implications of using these AI models are also evolving,” Bojic tells The Recursive.
It is also extremely important to note that humans are fragile beings, susceptible to dependence and manipulation, he adds. “There is no way to prevent this except by banning AI applications that degrade the mental health of their users, which is very clear with the aim of bringing deceased loved ones back to life through AI,” emphasizes Bojic.
“Get lost in your creation”
For those who actually create the technologies behind these products, things might be a little different, experts say. According to Predrag Tadic, assistant professor at the School of Electrical Engineering at the University of Belgrade, engineers generally do not have subjects dealing with ethics.
“When the rise of these AI technologies started, I started reading a lot about philosophy and morality, because I felt like I needed to know more about what I participates with my own work and the modest way of developing these technologies. I thought I should have a better idea of what is developing,” Tadic said, giving an example of how his students experience the industry, during a conference on the subject in Belgrade.
“I remember a brilliant student who decided to create an algorithm that would scan the shelves of a store and see where these quadrants were where items were missing. And he was really excited about it, he was successful, he got a job at a big company that was impressed with his work. And I asked him if he wanted to do anything else with his time since he’s young and smart and educated, and if he wanted to work on problems intended to help someone. And he said, “Are you kidding? Or are you serious? He wasn’t really sure I was pulling his leg. It tells you about the kind of thought process that goes on in people’s heads to develop this, they don’t really think about it, they’re happy when they see it works,” Tadic told The Recursive.
He himself has experienced the same enthusiasm and for the moment, we will not talk much about ethics, to the extent that there are no major incidents linked to the use of such products.
“I know how exciting it is to create something and have it work the way you want it to. When we think about the broader sociological and moral impact, I don’t think anyone – at least the engineers who build it – think about it. I think they should think, and I always tell my students to think about how they spend their time,” adds Tadic.
The Wild West phase of AI regulation
How the industry will respond to the use of AI to resurrect the dead depends to a large extent on the ethical framework and regulations put in place, Bojic adds.
“Currently, the industry is in a bit of a Wild West phase when it comes to AI ethics. Nevertheless, it is becoming increasingly clear that we cannot afford to ignore the ethical implications of AI development. There is an increasing need to understand the full impact of these technologies on society, which includes potential risks and benefits. And this, in my opinion, is where large-scale cooperation could play an important role. AI should be considered a public service, subject to regulation and control,” the Serbian researcher shares with The Recursive.
But are companies ready to make ethics an integral part of their laws and actions? Branka Andjelkovic, co-founder and program director of the Belgrade-based think tank Public Policy Research Center, said this would be possible if regulators took care of it first.
“You see these elements and principles embedded in the AI regulations. And not just AI law, it’s other laws as well – European law and the US response as well. This all sounds nice – but you don’t build it like a regulator, you just propose regulation that requires ethical rules – but it’s the engineers who build the technology. It is therefore essential to take a step back and think about how you can include ethical principles in the products you use. And some of the products are the ones we are discussing today, you don’t really understand how they are developed,” says Andjelkovic.
Even though the field of AI is exciting overall and the industry is coming up with groundbreaking inventions, it is at this crucial stage that it is steered in a direction that respects human rights and democratic values, says Bojic .
“It is with such precautions and collective focus that the industry can effectively respond to these ethical challenges and work toward an AI-driven future that meets our best interests,” he concludes.
