AI is pushing the boundaries of what’s possible, and now it’s enabling us to build and interact with digital avatars of the deceased. These so-called ‘deadbots,’ ‘ghostbots,’ ‘postmortem avatars,’ or ‘griefbots’ are designed to provide comfort to those grieving, but they also raise a host of ethical questions that we need to grapple with.
The trend of creating these AI-powered representations of the dead is gaining momentum. SenseTime Group founder Tang Xiao’ou used AI to deliver a ‘live’ speech four months after his death in March. Korean firm DeepBrain AI’s ‘Re;memory service’ allows for deceased parents to ‘live’ in the cloud and interact with relatives. This service draws inspiration from the film Coco, where the message is clear: if no one in the living world remembers you, you vanish. Other companies like MIT Media Lab’s Augmented Eternity project, HereAfter AI, MyHeritage, Super Brain, Storyfile, Silicon Intelligence, Eternime, and Somnium Space are also offering similar services.
While some might consider this concept morbid, there are valid arguments for its potential benefits. We all use photo albums to keep memories of loved ones alive, and deadbots could offer a similar form of comfort. Certain religions also hold the belief in a reunion after death, and these avatars could potentially align with those beliefs. Furthermore, these digital representations can help preserve family histories, enhance historical education, and even assist with legacy planning.
The possibilities are seemingly endless. Imagine students interacting with an Einstein avatar to learn about relativity, a grandmother’s avatar sharing treasured recipes, or a renowned surgeon’s ghostbot guiding medical students. We could even see life-like avatars of deceased pop singers performing at virtual concerts or children seeking comfort and guidance from their late parents’ avatars.
The global chatbot market is projected to grow to $15.5 billion by 2028, according to MarketsandMarkets, highlighting the increasing demand for these technologies. While basic voice avatars might cost between $500 and $15,000, more sophisticated versions can be much pricier. Super Brain’s founder, Zhang Zewei, told Forbes that customized griefbots can cost between 50,000 and 100,000 yuan ($6,860 to $13,710).
But can AI truly capture the essence of a deceased person? Current chatbots utilize GenAI models, including large language models trained on vast amounts of data, such as text, photos, audio, and video recordings. HereAfter AI, for example, uses recorded interviews to create interactive avatars that reflect the deceased’s mannerisms and speech patterns. Eternime utilizes social media posts, emails, and wearable data to simulate personality traits. However, it’s important to remember that ghostbots cannot fully replicate a person’s complex essence—their personality, knowledge, mannerisms, their ability to learn, grow, and make nuanced decisions.
While there are potential benefits, the creation of deadbots raises significant ethical concerns. Think of the distress caused to Robin Williams’s daughter Zelda by online deepfake videos of her late father. Ethicists from Cambridge’s Leverhulme Centre for the Future of Intelligence warn that even those initially comforted by a ‘deadbot’ might eventually feel drained by daily interactions. There’s also the possibility of companies exploiting these technologies for personal gain, such as using them to ‘stalk’ the living with spam messages or impacting inheritance. The lack of consent from the deceased is another crucial issue, adding further complexities to the ethical landscape.
The development of AI avatars of the dead presents us with a fascinating but complex dilemma. While they offer potential comfort and opportunities, their ethical implications demand careful consideration and responsible development to ensure their use doesn’t come at the cost of exploitation or further grief.
