On June 18, AI image-generation company Midjourney released a tool that lets users create short video clips using their own images as a template. Days later, Reddit cofounder Alexis Ohanian posted on X about how he used the tech to animate a photo of his late mother, which shows him as a child wrapped in her embrace.
In the artificial video, she laughs and smiles before rocking him in her arms. “Damn, I wasn’t ready for how this would feel,” he wrote. “This is how she hugged me. I’ve rewatched it 50 times.”
Ohanian’s post, viewed almost 30 million times, has reignited a longstanding debate over how technology mediates grief and memory—and whether it’s magical or dystopian. TIME spoke with experts on grief and memory to understand how this latest advance in “digital resurrection” is changing our relationship with the dead.
False Memory
Human memory has always been fallible: while we typically remember the gist of an event, details are forgotten or distorted. Memory is not “a personal library of all the things that have ever happened to you,” says Julia Shaw, a criminal psychologist specializing in false memories. “It was meant to help you survive.” While Shaw feels positive about using AI to reanimate people, she says the technology poses the risk of contaminating and overwriting our memories. “AI is a perfect false memory machine,” she says.
Of course, people are capable of distorting their memories without technological assistance. “My grandfather used to yell at my grandmother all the time, but after he died, he was the most wonderful man in the world,” recalls Elizabeth Loftus, a professor of psychology and law and pioneer in memory research. And it’s well-established that tools like Photoshop and doctored videos affect what people remember about the past.
But AI changes the ease and extent to which content can be altered. A recent study that Loftus conducted with the MIT Media Lab found that exposure to even a single AI-edited visual affected people’s memory of the original. Participants “reported high levels of confidence in their false memories,” with younger people proving particularly susceptible.
The researchers also found that while this technology could have beneficial uses, such as reframing traumatic memories or enhancing self-esteem, there is a considerable risk of creating false memories in high-stakes contexts like courtrooms, and using the technology to spread misinformation.
Grief, Interrupted
One possible harm: engagement with digital simulacra of the deceased could complicate the grieving process. Mary-Frances O’Connor, a neuroscientist and author of The Grieving Body, explains that grieving is a process by which one learns to reconcile the reality of a person’s death with the sense—encoded at the neurobiological level in one’s brain—that they should still be here. She notes that for many people, the dead continue to live amongst us, insofar as people report experiencing their presence. “Many bereaved people describe how every time they walk into a room, they see a hole that no one else is seeing.”
O’Connor notes that “all cultures, in all periods of history, have used whatever technology they could to connect with their deceased loved ones.” Once cameras were invented, for example, people began keeping photos of the deceased in their homes. In 2020, documentarians in South Korea used virtual-reality to create a structured experience for a mother to reunite with her daughter, who she lost to a rare medical illness. While the experience helped the mother process her daughter’s death, it was met with concern by Western media.
Perhaps the key question, she says, is whether AI helps us connect to our late loved ones, or reinforces the idea that they are everlasting. Given the unprecedented nature of the current moment, it may be too early to tell.
“We’re in a massively novel situation: the dead have never been this talkative before,” says Elaine Kasket, a cyberpsychologist and author of All the Ghosts in the Machine. Between traces left online and the ability to digitize old letters, photos, and other records, we have access to more “digital remains” than ever. Kasket believes she has access to sufficient material from her friend, for example, to have a conversation with a machine that would be “functionally indistinguishable” from one with his human counterpart. As human memory is already hallucinatory and reconstructive, she wonders: “is the fiction from the machine unhealthier than the fiction from within our own heads?” It depends what function it serves.
Wu watches a video created with artificial intelligence showing the face and voice of his son, who died last year at the age of 22 while attending Exeter University in Britain. (Hector Retamal—Getty Images)
Dead Intelligence
With frontier AI companies investing billions of dollars in creating “agents,” AI systems may become increasingly convincing stand-ins for the dead—it is not difficult to imagine, for example, soon being able to videocall a simulacra of a grandparent. “I think that would be a beautiful future,” says Shaw, while emphasizing the need to prevent the AI being weaponized against the person. “It feels like an atheist version of being able to talk to ghosts,” she says.
Alongside the questions of whether this is good or bad, and whether it is truly distinct from what has come before, is the question of who stands to benefit. O’Connor notes that people have long profited from the bereaved, from mediums and seances to intercessionary prayers in the Catholic church, where a priest would only pray for the soul of the deceased for a fee.
There may be real therapeutic and emotional value in being able to reconnect and potentially achieve closure with lost loved ones, in the same way that some people find value in texting or posting to somebody’s social media feed after they’re gone, says Shaw. “If people want to do this in their own private world, because it makes them feel happier, what’s the harm?” says Loftus.
For O’Connor, cause for concern arises when somebody is engaging with the deceased to the exclusion of other important aspects of their life, or when they become secretive about their behaviour. On the whole, though, she emphasizes the remarkable resilience of human beings: “this will be one more thing we learn to adjust to.”
Kasket sees a risk that reliance on digital reincarnates renders us brittle: if all the “difficulty and mess and pain” associated with human relationships can be scrubbed away, we may be left vulnerable to life’s unexpected challenges. At the point where we “pathologize and problematize the natural finitude and impermanence of carbon-based life forms such as ourselves, we really need to take a beat and think about what we’re doing here,” she says.
