The British fantasy drama film Truly, Madly, Deeply tells the story of a grieving woman named Nina (played by Juliet Stevenson), overwhelmed with sadness at the recent death of her boyfriend Jamie (Alan Rickman). Sensing the depth of her despair, Jamie reappears as a ghost, with the intention of helping Nina to come to terms with her loss.
Ultimately, Jamie succeeds in this, getting Nina to question her memories of him, and later, accept that he wasn’t always as perfect as she had imagined. Through this process, she realizes it’s possible to live a life without him.
When the movie was made in 1990, the premise was regarded as a complete fantasy, but fast-forward to 2024, and the concept is not nearly as far-fetched. With the rapid rise of powerful AI algorithms, we’re much closer to achieving immortality, at least in the digital realm.
Science fiction has provided numerous examples of how human minds can be uploaded onto computers, and many are deeply unsettling. But the possibility exists that such technology can be far more benevolent, too.
An everlasting AI legacy
One interesting scenario was the subject of a Korean documentary show in 2020, where AI experts teamed up with virtual reality producers to enable a grieving mother to reunite with her seven-year-old daughter. The experiment was controversial, with many considering the idea to be somewhat ghoulish. But the mother herself said that having the chance to talk to her daughter after she had died helped her to process what had happened and learn to live with it.
In that case, the encounters between the mother and daughter were scripted, and the interactions therefore somewhat limited, but with the subsequent rise of generative AI, we may soon have the opportunity to interact with our departed loved ones in a much more realistic way.
Another South Korean company, Deep Brain AI, is working too hard to make this happen. At the CES Show in January 2024, it unveiled its Re: Memory technology which uses a combination of powerful AI algorithms, video footage, and voice recordings to create realistic avatars that look and sound just like the person they’re based on. They can converse in real-time, responding to questions and phrases in a lifelike way.
Similarly, Twin Protocol is working on an AI twin technology that people can use to create a digital replica of themselves and leave behind a kind of personal legacy that will stick around to converse with others long after they have passed away. The company showcased its vision at the SingularityNET Internal Ecosystem Meeting in Thailand in November 2024, explaining how it combines AI with blockchain to address challenges around data security and potential misuse.
The idea is that anyone can use Twin Protocol to create a digital version of themselves. They’ll have to train it to think as they do by engaging with it in regular conversations. The more one converses with it, the more realistic their AI replicant will become, as it will slowly accumulate their knowledge and experiences and understand their personality.
Twin Protocol CEO Stacey Engle was one of the first to create a digital replica of themselves and has posted a video on YouTube that shows how it can act as a kind of digital memoir, preserving her unique character traits for future generations. The video talks about the innovative processes that go into the creation of AI replicas and the ethical considerations it raises and provides us a glimpse of a future where everyone can create a personal legacy that transcends time.
Retaining our knowledge, in the future and now
The concept of AI avatars helping people to overcome their grief at the loss of a loved one is certainly divisive, but it’s not the only reason to consider creating an ‘immortal’ AI legacy. Potential benefits include being able to preserve the valuable knowledge that can only come from a lifetime of experiences and learning.
It’s a thought-provoking idea. What if Albert Einstein was still around, would he be able to help solve some of the complex challenges the world’s brightest physicists continue to wrestle with, with the aid of modern technology? Would Nelson Mandela be able to broker a peace deal in conflict zones such as Ukraine, where modern leaders have failed? And what kinds of movies might William Shakespeare be able to produce, if given the chance to work with a modern Hollywood studio?
There’s no way we’ll ever know because their thought processes expired for all eternity the moment the neurons in their brains stopped firing. But for future generations, we might be able to capture enough data throughout someone’s lifetime to retain their basic thought processes forever more.
Of course, it’s not only the dead that stand to benefit from such technology. Eric Yuan, the CEO of the video communications platform Zoom, believes it will soon be possible for anyone to create a “digital twin” that would be able to act on their behalf, for example by replacing us in video calls, taking notes and discussing ideas, before reporting back to the real-life individual later.
It paves the way for some intriguing possibilities. Such technology could potentially enable us to be in two places at once, giving us more than 24 hours a day to get things done. Or else a busy CEO, who recognizes they’re not spending nearly as much time with their kids as they should, could simply let their digital twin run the company on a day-to-day basis, while their physical self goes on vacation with the family for a week.
Some might be skeptical that an AI replica could ever be as knowledgeable as a real-life person, but others are more confident. In a 2009 paper, Microsoft researchers Jim Grey and Gordon Bell show that we’d only need around one terabyte of storage space to record every conversation someone has throughout their lifetime. We already have far bigger datasets that are used to train our most advanced generative AI chatbots.
By preserving someone’s memories and personality as a digital AI twin, we would be able to benefit from that person’s wisdom long after they pass away. Even more intriguing is the possibility that our immortal versions would be able to continue learning far into the future, as generative AI chatbots like ChatGPT already do.
But should we do this?
It’s interesting to note that shortly after the invention of photography, some philosophers presented the technology as a way for people to achieve a kind of immortality, leading to a lot of debate regarding the ethics of trying to ‘cheat death’, as it were.
Some argue that death is just a natural part of life and that we need to accept that it is a certainty for everyone. If we were to eliminate death through AI clones, it’s not clear what kind of impact this might have on our society. There are concerns that AI could alter the way we think about death. For instance, if we can still talk to our loved ones after they’ve departed from the physical world, will we be able to mourn them properly?
An even more worrying possibility is that AI immortality might result in some of the living being ignored. If an individual is suffering from Alzheimer’s disease, might that living person be neglected in favor of their ‘healthier’ digital twin that’s easier to converse with?
Still, it’s also in our nature to try and resist death and to live for as long as possible. For millennia, there have been stories told of people trying to discover the secret to eternal life. The search for the Holy Grail has been the subject of countless movies, and it’s an idea that’s appealing to many.
The enduring appeal of eternity, together with the evolution of medical technologies, AI, and other concepts like immersive virtual realities, means it’s pretty much guaranteed that people will continue to try to find ways to extend their lives, and they might now be within reach.
Perhaps, just perhaps, someone who’s reading this article today might get to exist much farther into the future than they ever thought possible. We’ve reached a point where someone living now might plausibly be able to see the day that humans finally get to colonize Mars or journey to the stars, or perhaps even witness the Earth being swallowed up by the sun billions of years from now.
