Digital Echoes: The Rise of AI Afterlives and Their Impact

In the near future, people might interact with AI versions of those who have passed away. These AI entities, called generative ghosts, can create new content and engage in conversations, unlike simple griefbots that only replay recorded messages. Jed Brubaker, a professor at the University of Colorado Boulder, studies these generative ghosts and their potential impact on society. Brubaker notes that people often find new technologies unsettling at first. For example, photos and videos were once considered creepy, but society eventually adapted. Similarly, generative ghosts may seem strange now, but they could become commonplace. Generative ghosts have various applications beyond helping people cope with grief. They can be used in education, such as the virtual Benjamin Franklin created by Khan Academy. Museums, especially Holocaust museums, are interested in using generative ghosts to preserve history and educate the public. However, there are ethical and philosophical questions surrounding generative ghosts. For instance, who should create them? Should they be a reincarnation or a representation of the deceased? And should they evolve over time? There are also risks to consider. People might become overly attached to these AI entities, and there are concerns about privacy, security, and the potential for these AI to tarnish the memory of the deceased. Despite these risks, Brubaker believes that generative ghosts could have significant benefits for society. Brubaker and his colleague Meredith Ringel Morris wrote a paper exploring these benefits and risks. They emphasize that their work is not a prediction but a way to start a conversation about the future of AI and its role in our lives. Brubaker is personally interested in creating his own generative ghost. He sees it as a way to preserve his knowledge and continue to contribute to society even after he's gone.