The love for a loved one never flatlines, even if the loved one is lost. If you had a chance to speak with an AI-generated “ghost” of a deceased loved one, would you take the chance? The idea might be cathartic, but to 16 volunteers, these conversations were often comforting, even when the chatbot got the facts wrong.
A Quantum of Solace
Information scientist Jack Manning and colleagues report that the volunteers open to using technology like AI ghosts, also referred to as griefbots or deadbots, don’t need to be a spitting image of their lost loved one. In fact, the volunteers were more than happy to have their emotional and social expectations met. Manning, of the University of Colorado Boulder, was surprised by how much the participants enjoyed speaking to the AI-generated ghosts.
He also harbors fond memories of his 13-year-old sister who passed away when he was 11. Although Manning himself is averse to the idea of interacting with her digital ghost, his study volunteers came with a goal in mind. While he wants to hold on to his unwavering version of his sister in his mind and fears being confronted with the fact that he’s probably wrong in how he remembers her, his volunteers were looking to share unanswered questions and details from their lives with the AI ghosts.
Across the board, volunteers reported having a positive experience. “It felt like I was actually talking with my grandpa,” one participant told researchers. Another said, “It just feels like I’m getting the closure I needed so bad.”
How did the Chatbot Work?
A brief survey that the participant completed about their deceased loved one fed necessary information to the chatbot to impersonate. Sometimes, the bot filled in missing details with fabrications, such as mentioning a job someone never held. The participants, however, went forward with their conversation. They were more bothered by mistakes in tone or style.
The bot never used emojis, leaving a void for those who communicated with their loved ones through text. In another instance, a bot called a participant “champ”, a name his loved one never would have used. “If we want to understand how to design these systems ethically, we need more evidence of this kind,” says AI ethicist Tomasz Hollanek of the University of Cambridge, who was not involved in the research.
The Real World is Messier
The study had human mediation to scrutinize the chatbot’s replies before passing them on to participants. That helped prevent deceptive or harmful interactions. The problem occurs in the real world, where things get messy, and mediation is often unavailable. Anyone can prompt a chatbot to impersonate someone who has passed away, and companies now sell services that can “re-incarnate” loved ones through AI-generated text, voices, or faces. While the volunteers enjoyed the experience, they saw the risks. One worried about becoming too reliant on conversations with the bot, while another wondered if they would like the person they become if they kept using this technology.
However, Manning is positive that if designed carefully, the technology might help some people to work through grief. His own aversion to AI-generated ghosts has dampened during the course of the study.
Copyright @smorescience. All rights reserved. Do not copy, cite, publish, or distribute this content without permission.
