Mostly fake people, dead ones. Plus: real live class guests this fall
An expansion of a previous post, this time stretching the mortal coil a bit. We have some great guests (live ones) for this fall's seminar, too!
Nearly two years ago I wrote about a set of “virtual humans” who had begun to step into the traditional roles once played by celebrities, who endorsed products in ads. In this century, “influencers” took over, as they harvested new powers from social media. Virtual influencers now have realistic-enough backstories—fictional, of course, and not standing up to even modest scrutiny. But for the virtual influencers’ sponsors, the best part is that virtual personalities can be exquisitely designed to resonate with audiences.
Back in October 2022, the fake billion-dollar babies already could “surf a vast social media infrastructure built on ‘following’ and digital ‘friendship.’ They have complexity and detail. They are beautiful. They do interesting things. They seep between the cracks separating a real world and their own.”
Now, even the separation of life and death can be bridged with simulations—a kind of undying deep fake of real humans that aligns with fake influencers. The virtual can envelope the real, the departed, and the deceased. As a means of preservation, it’s better than embalming, too, since “deadbots” or “griefbots” can talk. Though they’re still enclosed in a digital display (for now, at least), through an AI simulation, the dead still “converse.”
A resurrection, of sorts, for the famous
Salvador Dalí hasn’t been visually resurrected on the tiny screen yet, but if you’re in St. Petersburg, Florida, you can still ask him anything. You just call him on his special lobster phone at the Dalí Museum. At least the artist is honest about how he got back from the dead: “How they brought me here is far beyond my comprehension,” he says. “All I know is that they used something called a Large Language Model in a recreation of my voice. And here I am. So, next time you have a question just ask Dalí.”
Judy Garland, also dead, will soon be talking regardless, though unlike Dalí, she won’t be open to questions right away. But if you have a PDF or text of The Wonderful Wizard of Oz on your phone, Garland will read it to you. She’ll read anything you have, for that matter. Garland is part of a troop of dead actors dug up by ElevenLabs in their “Iconic Voices Collection.” It’s a free app (for now at least), so you can have famous voices in your pocket.
Liza Minelli, Judy Garland’s daughter, likes the idea: “Through the spectacular new technology offered by ElevenLabs, our family believes that this will bring new fans to Mama and be exciting to those who already cherish the unparalleled legacy that Mama gave and continues to give to the world.”
Among the notables joining Judy Garland are James Dean, Burt Reynolds and Sir Laurence Olivier.
The resurrection of these individuals speaks to their living roles, which in some sense was a commoditized identity. The actors used their bodies and their voices as salable, an attribute unique to them that they controlled. Today, 26 US states recognize the “right of publicity” that requires permission for use of voices and other attributes that make up one’s identity. Dead actors lining up in ElevenLabs’ iconic voices collection have the permission of their estates to back them up.
A voice is an undying asset that can be traded, as long as there’s a market.
Virtually hanging onto the departed
After the deceased have crossed River Styx, not-so-famous Virtual Loved Ones can hold on to this woeful world, if the living they left behind are willing to pay their keep. Such fees to tech companies for new ways of “keeping in touch,” of course, follow a long tradition of making money from mourning, loss, and grief.1 Profit-seeking and human grief have long been a delicate, sometimes dangerous, mix.
You don’t have to be famous to achieve this odd, one-sided immortality, even though Burt Reynolds, James Dean, Sir Lawrence Olivier, and Judy Garland seem to have come back only for new acting gigs.
Hans Block and Moritz Riesewieck directed Eternal You (trailer here) which premiered at this year’s Sundance Film Festival. It’s a documentary film focused on a new AI-driven industry to create virtual representations of the dead. But, as Block clarified, “the main focus of our film is really the loved one living on with the avatars of the deceased.” He pointed to a “transcendental homelessness dealing with these kinds of topics”—topics that once were the province of religion, family, and community. Into that absence, the filmmakers show the new AI services that offer chatbots bearing the voices of the dead and trained on media residues of a dead person’s life.
The idea sounds creepy enough that it would seem to have come out of the fringe or in a Black Mirror episode, but it’s worth knowing that Microsoft applied for a patent in April 2017 (granted 2020: U.S. Pat. No. 10,853,717) on chatbots “to enable the agents to embody a person, alive or deceased,” with training coming in part from social media. According to an article in the Financial Times, Microsoft stopped working on a deadbot project in 2021 “after seeing ‘disturbing’ results.”
Block and Riesewieck captured some disturbing results in their documentary. After having built a relationship with some people who were using chatbots designed to simulate dead loved ones, Block and Reiseweick filmed some of the interactions that crossed the Great Divide. In an interview with IndieWire, Block recalled one scene: “One [episode] that stuck out was the first conversation with Christie [Angel]. She’s one of the users of Project December, and she told us her story, and we were so touched about everything, we were so moved.” Angel used Project December to text with a chatbot simulating her dead boyfriend, Cameroun.
When she asked him where he was, he replied “in Hell,” surrounded by “mostly addicts.” Angel, quite naturally, was devastated.
“I would say for the first time we understood what an immense power these technologies have,” Block said, “even if the technology is not really far developed and even if there are problems and wrong answers…. We don’t need to wait for the Super AI…. Even now, we are pretending to see so much in these technologies.”
Block and Rieseweick have the reputation of being skeptical of modern technological innovations, but Rieseweick summed up Eternal You in a way that leaves open benefits of AI as human aides. “In the end we believe our film is maybe not so much about AI as it is about a human desire … which is implemented into AI,” he said. “It’s not only that AI is emerging but there’s an idea emerging—the idea to use this new AI to give people hope, to give people a completely new idea of afterlife.”
The IndieWire interview (below), by the way, is quite good. I haven’t been able to find Eternal You streaming online, though it appears that the film is making the rounds in special showings.
Perhaps, an unpopular view
Not long before his death in April this year, Daniel Dennett said that “we should make it a mark of shame, not pride, when you make your AI more human.” Then, he was primarily concerned about the damage that disinformation and “counterfeit people” could inflict upon social fabrics woven from shared truths and humanity. He viewed that threat as “existential.” The challenge of griefbots is similarly existential, but also more intensely personal, since they play upon individual’s memories, personal histories, and feelings and, perhaps more importantly, create an illusion of immortality. Blunted by a life-like chat with a dead loved one, death might not prod us to affirm the treasures of life in all its limits. That recognition is a bittersweet conclusion that grief may lead us toward.
Deadbots of mothers and fathers and beloved children are a special kind of deepfake bearing a payload of a certain kind of disinformation, a philosophical or theological disinformation. That payload may sand the rough edges of grief by obscuring death’s finality. Or it may merely caricature the dead, turning memory into a new kind of plastic loss.
Sometimes, anxiety also opens possibilities, insights, wisdom, faith; and death is a great source of anxiety. And it’s a paradox to say that in the loss and grief of death, comes an opportunity to touch deeper and more courageous life.2 AI griefbots could stand in the way of a beneficial and revelatory anxiety, a boundary experience that can open us to shape meaningful life and kindle a joy of living.
Got a comment?
And now … a great lineup of 2024 seminar guests
How about some lighter news!
Every year, my students and I get to talk with some very interesting and smart people as we probe “our complex relationships with technology.” This year, I’ve lined up four guests, and here’s something about them:
Brinnae Bent, is a faculty member in Artificial Intelligence at Duke University, a Responsible AI research scientist, and a startup advisor. She is a leader in bridging the gap between research and industry in machine learning, having led projects and developed algorithms for the largest companies in the world. More importantly, she has built algorithms that have meaningful impacts—from helping people walk to noninvasively monitoring glucose. Brinnae is an accomplished researcher and ML practitioner, with over 30 publications and a breadth of experience developing algorithms across domains including health/wellness, sports, privacy, interior design, and energy. Her research on digital biomarkers is world-renowned and her emerging work on Responsible AI strives to solve important problems in the application of AI to real-world problems. Brinnae also curates a weekly tech newsletter called “Spill the GPTea.” Outside of work, Brinnae is a mother, an ultramarathoner, and an artist.
Erich Huang is Head of Clinical Informatics at Verily (https://verily.com/), Alphabet’s (i.e., Google) company that focuses on medical technologies. He continues to serve on the surgery and the bioinformatics and biostatistics faculty at Duke School of Medicine, where he also served as Assistant Dean of Biomedical Informatics and as Chief Data Officer for Duke Health. He co-founded and directed Duke Forge, the university’s health data science initiative. In addition to his academic roles, he is a “serial founder,” having started and advised several health and technology-related companies. He was director for cancer research at Sage Bionetworks and serves as chief science and innovation officer at Onduo (https://onduo.com/). He is particularly interested in the intersection of emerging technologies that can transform healthcare, including medical devices and AI. He earned his undergraduate degree at Harvard University and MD and PhD degrees at Duke, where he also did his general surgery residency.
Anu Kirk is currently working on his Master's degree in counseling psychology. Most recently, he was Vice President of Product at Osso (https://ossovr.com/). His career has consistently engaged art, music, and technology, exploring the ways that artistic expression of all kinds can be deepened or limited by technology tools. He has pursued his interests in large multi-national corporations and in start-ups. At Sony Network Entertainment, he was Director of Music Services, where he led Playstation Music (formerly Music Unlimited); at PlayStation, he was Director and GM of Virtual Reality Products. Work with MOG led to award-winning Android and Apple apps that were acquired by Beats Music and later formed the core of Apple Music. He and his teams won a Webby Award (2013) and “Best VR Headset” award for PlayStation VR (2017). He graduated from Dartmouth College with a degree in economics. He is an active musician and writer and was awarded a patent (“Interactive delivery of media using dynamic playlist generation subject to restrictive criteria”) with a co-inventor who is a good friend and colleague. You can thank them in part for your streaming music queue.
Christopher Mims writes "Keywords," a weekly technology column for the Wall Street Journal, which he joined in 2014. He has written on science and technology his entire journalistic career, and his work has appeared in the MIT Technology Review, Smithsonian, Wired, The Atlantic, Scientific American, and Quartz. His grasp of and insight into technology is voluminous, as a glance at topics he has taken on reveals. He has written about bidets, brain implants, the cult of the founder, the history of technology, innovation, venture capital, robotics, batteries, energy, materials science, wireless communications, AI, data science, telepresence, microchips, logistics, IT, 3D printing and autonomous boats, trucks, cars, drones and flying taxis. His book Arriving Today: From Factory to Front Door—Why Everything Has Changed About How and What We Buy examines the pervasive, critical, and often overlooked—and fast-developing—technologies and processes of logistics, how products move from their manufacture to our doorsteps. Arriving Today was shortlisted for the 2021 Porchlight Business Book Awards. Mims has also won a SABEW award for his commentary at the Wall Street Journal.
Got a comment?
Tags: deadbot, griefbot, AI, boundary experience, voice simulation, chatbot, seminar guest
Links, cited and not, some just interesting
“When Ana Schultz, a 25-year-old from Rock Falls, Illinois, misses her husband Kyle, who passed away in February 2023, she asks him for cooking advice. She loads up Snapchat My AI, the social media platform’s artificial intelligence chatbot, and messages Kyle the ingredients she has left in the fridge; he suggests what to make.” Kelly, Samantha Murphy. “AI Resurrects Deceased Actors’ Voices to Read Audiobooks.” CNN, July 3, 2024. https://www.cnn.com/2024/07/03/tech/elevenlabs-ai-celebrity-voices/index.html and “When Grief and AI Collide: These People Are Communicating with the Dead.” CNN, May 6, 2024. https://www.cnn.com/2024/05/06/tech/ai-communicating-with-dead/index.html.
Veltman, Chloe. “An AI Salvador Dalí Will Answer Any Question When Called on His Famous ‘Lobster Phone.’” WSKG, April 21, 2024. https://www.wskg.org/2024-04-21/an-ai-salvador-dali-will-answer-any-question-when-called-on-his-famous-lobster-phone.
Vauhini Vara “fed it a brief description of her sister’s death a few sentences at a time. She asked the bot to keep writing, to tell her what should come next. The bot created nine possible worlds from the fragments and vignettes. As you move through the essay, you realize that the AI never gets it right. Nothing that it proposes is correct. Yet this doesn’t feel wrong; it feels more like an alternate universe, a what-if episode.” VanDyke, Bryan. “Rise of the Ghost Machines.” The Millions, June 5, 2024. https://themillions.com/2024/06/rise-of-the-ghost-machines.html.
Eternal You - Trailer (DOXA 2024). YouTube video, 2024.
Simulate the Dead - Project December, YouTube video, 2022.
“ ‘Rapid advancements in generative AI mean that nearly anyone with Internet access and some basic know-how can revive a deceased loved one,’ said Dr Katarzyna Nowaczyk-Basińska, study co-author and researcher at Cambridge’s Leverhulme Centre for the Future of Intelligence (LCFI). ‘This area of AI is an ethical minefield. It’s important to prioritize the dignity of the deceased, and ensure that this isn’t encroached on by financial motives of digital afterlife services, for example. At the same time, a person may leave an AI simulation as a farewell gift for loved ones who are not prepared to process their grief in this manner.’ ” Hollanek, Tomasz, and Katarzyna Nowaczyk-Basińska. “Griefbots, Deadbots, Postmortem Avatars: On Responsible Applications of Generative AI in the Digital Afterlife Industry.” Philosophy & Technology 37, no. 2 (May 9, 2024): 63. https://doi.org/10.1007/s13347-024-00744-w. The article is neatly summarized, with illustrations, in SciTechDaily: “Cambridge Experts Warn: AI ‘Deadbots’ Could Digitally ‘Haunt’ Loved Ones From Beyond the Grave,” May 9, 2024. https://scitechdaily.com/cambridge-experts-warn-ai-deadbots-could-digitally-haunt-loved-ones-from-beyond-the-grave/.
Russavage, Edward, and Benjamin Nickerson. “AI: The Voices Behind the Music.” IPWatchdog, July 25, 2024. https://ipwatchdog.com/2024/07/25/ai-the-voices-behind-the-music/id=179355/. (H/T Steve Toback)
The funeral business was well and slyly documented by Jessica Mitford’s The American Way of Death (1963; Alibris) and the follow-up The American Way of Death Revisited (1998; Alibris; Bookshop.org). I’ve used her entertaining description of the process of embalming “Mr. Jones” in a writing class. Mitford was a great prose writer. Her funeral arrangements in 1996, by the way, cost $533.31—far below the average.
I am thinking of Paul Tillich’s The Courage to Be (1952; Alibris; Bookshop.org). Tillich claims there are three existential threats: 1) fate and death, guilt and condemnation, and 3) meaninglessness and emptiness. I know, sounds pretty grim and anxiety-laden, but Tillich viewed these threats as openings to larger and more “courageous” and affirming life. That paradox of soul-eating anxiety opening a pathway to fuller life may be short-circuited by our current misinterpretation of AI. Perhaps we will be able to reinterpret AI more realistically in the future, as we have done with other technologies. I think of photography, for example, which challenged similarly but I think less profoundly. Tillich’s book is pretty short and quite readable.
Pretty deep thinking for a Luddite like me! I guess it’s healthy to expand my limited horizons. Keep at it! Just a drag upon the system. Tzwickey@msn.com