When Lolo (not his real name) lost his mother to cancer, his grief was unfathomable. In her mid-20s, working as a content creator in China, she was troubled by the unfinished nature of their relationship. Their bond has always been complicated, shaped by unspoken resentments and a childhood where caring was often followed by criticism.
After his mother’s death, Rollo finds himself unable to reconcile the turmoil of their past and the silence that follows. She shared her struggles with her followers on the Chinese social media platform Xiaohongshu (meaning “Little Red Book”), hoping to help them on their own healing journey.
you may like
“I wrote about my mother, recorded all the important events in her life, and created a story about her resurrection in the world of AI,” Lolo told me through an interpreter. “We write down the key life events that form the protagonist’s personality and define their behavioral patterns. Once we’ve done that, the AI can generate its own responses. Once we’ve generated the output, we can continue to make adjustments based on our wishes.”
During the course of his training, Rollo began to reinterpret his past with his mother, changing elements of her mother’s story to create a more idealized person: a kinder, more attentive mother. This helped her process the loss and resulted in the creation of Xia, a public chatbot that her followers can also interact with.
After her release, Lolo received a message from a friend saying her mother would be very proud of her. “I broke down in tears,” Lolo said. “It was incredibly healing. That’s why I wanted to create something like this. Not just to heal myself, but to give others something that might say the words they want to hear.”
Sadness of the Deathbot era
As I detail in my new book, Love Machines, Lolo’s story reflects the new possibilities technology has opened up for people to cope with grief through conversational AI. Large-scale language models can be trained using personal materials such as emails, texts, voice notes, and social media posts to mimic the conversational style of a deceased loved one.
These “deathbots” or “griefbots” are one of the most controversial use cases for AI chatbots. Some are text-based, while others depict a person through a video avatar. For example, U.S. “grieftech” company You, Only Virtual creates a chatbot from conversations (both spoken and written) between the deceased and their living friends and relatives, creating a version of what the deceased looked like to that particular person.
watch on
Some deathbots remain static representations of the person at the time of death, while others are given access to the internet and can “evolve” through conversation. Justin Harrison, CEO of You, Only Virtual, argues that unless AI can keep up with the times and respond to new information, it will not be an authentic version of the person who died.
However, this raises many difficult questions about whether it is even possible with current technology to estimate the development of human personality, and what impact interactions with such beings might have on the loved ones of the deceased.
you may like
Xingye, the platform on which Lolo created the chatbot for his late mother, is one of the key catalysts for new proposed regulations by the Cyberspace Administration of China, which regulates and censors national internet content, aimed at mitigating the potential emotional harm of “human-like conversational AI services.”
What does digital resurgence bring to grief?
Deathbots fundamentally change the mourning process because, unlike looking at old letters or photos of the deceased, interacting with generative AI can introduce new and unexpected elements into the grieving process. For Lolo, creating and interacting with an AI version of her mother was incredibly therapeutic, allowing her to put into words emotions she had previously been unable to express, giving her a sense of closure.
But not everyone shares this experience, including London-based journalist Lottie Hayton. She wrote about her experience of losing her parents suddenly in 2022 and recreating them with AI. She said she found the simulation creepy and painful. The technology wasn’t yet widespread, and the clumsy imitation felt like it was cheapening rather than honoring her real memory.
watch on
There are also important ethical issues such as whose consent is needed to create a deathbot, where it will be allowed to be displayed, and what impact it may have on other family members and friends.
Could one relative’s desire to create a symbolic companion to help them understand their loss give them the right to publicly display a deathbot on their social media accounts for others to see, compounding their grief even further? What happens if relatives disagree about whether a parent or partner wanted a digital rebirth in the first place?
The companies creating these deathbots are not neutral grief counselors. These are commercial platforms driven by familiar incentives for growth, engagement, and data collection. This creates a tension between what is psychologically healthy for the user and what is profitable for the company. A deathbot that people compulsively visit or can’t stop talking to may be a successful business, but it’s a psychological trap.
These risks don’t mean we need to ban all experiments with AI-mediated grief or deny the real solace that some people, like Rollo, find in them. But these mean that decisions regarding the “resurrection” of the dead cannot be left solely to startups and venture capitalists.
The industry needs clear rules around consent, limits on how postmortem data can be used, and design standards that prioritize psychological well-being over endless engagement. Ultimately, the question is not just whether AI should be allowed to reanimate the dead, but who can do it, under what conditions, and at what cost.
This article contains a link to bookshop.org. The Conversation UK may receive a commission if you click on a link and make a purchase from bookshop.org.
This edited article is republished from The Conversation under a Creative Commons license. Read the original article.
Source link
