How Generative AI Enhances Practices of Remembrance and Commemoration
2 min read
Generative AI, including large language models like ChatGPT and image generators like DALL·E 2, is revolutionizing what is known as “digital necromancy” – the ability to summon the deceased through their digital traces. This concept was first popularized in the 2010s with the advent of “deep fake” technology, which brought back stars like Bruce Lee, Michael Jackson, and Tupac Shakur through video projections. Now, with the rise of generative AI, these technologies are becoming more accessible to everyone.
Start-ups like Here After and Replika are using generative AI to reanimate loved ones for the bereaved. This technology, however, has raised concerns about its cultural and ethical implications. Many people feel uncomfortable with the idea of regularly interacting with digital simulations of the dead, viewing AI-assisted necromancy as dark magic.
However, as sociologists working on remembrance and commemoration, we believe that there is no cause for concern. Throughout history, people have formed emotional connections with representations and relics of the deceased, such as portraits and photographs. Generative AI is simply an extension of these existing practices.
The speed at which generative AI’s necromantic capabilities have been embraced indicates its compatibility with our current ways of grieving, remembering, and commemorating. AI start-ups incorporate texts, audio recordings, photographs, and videos of loved ones to create AI models that enable posthumous interactions.
There are concerns that the reanimated may be made to say things they wouldn’t have said while alive, but we should approach these situations on a case--case basis. It’s important to remember that we already engage in imagined conversations with the dead, reflecting on their potential reactions and advice. Images, texts, and artifacts have always served as conduits to their memory.
Critics argue that digital interactions with the deceased may be fraudulent, akin to the charlatans of the Victorian spiritual revival movement. However, we don’t confuse personal messages and recordings with the actual loved ones themselves. These records act as proxies for us to remember and communicate through.
In conclusion, concerns about digital necromancy are exaggerated. By focusing excessively on its strange and sinister aspects, we overlook how these technologies align with our existing human experiences. Generative AI offers new avenues for remembrance and commemoration, building upon our long-standing practices.