'Til AI do us apart
On grief tech, deadbots and the digital afterlife.
I was planning for this edition to be on how AI is changing relationships and the way we interact in the world of the living. I even had a great title for it (Are you in a situAIsionship?) and was looking forward to writing it.
And then it happened.
The phone rang and I saw the name of a friend pop up on the screen. As we had agreed to speak over the weekend (but then life got in the way) I figured he was calling me to have a chat as I knew he was at a conference, so probably he was calling me in between sessions while he was having a break. I picked up the phone and his first words were a slap on the face: Someone we knew and had worked with had just passed away suddenly.
I had to ask a few times for confirmation. Were we talking about the same person? How? When? Why? Was it really him? In my head this simply wasn’t a plausible scenario. I was sure I wasn’t catching the name right. It wouldn’t be the first time I’ve struggled to understand people over the phone while speaking in English and getting something completely different from what they were saying. This surely was one of those times.
Except it wasn’t. Later in the day I spoke with another colleague who due to the time difference had woken up to the same news I had been given hours earlier. By then I had already absorbed the initial shock and couldn’t help to get a bit emotional as we reminiscened about him.
“I’ve realised that no one will ever call me CC again” I told my colleague on the other side of the screen. “That was his nickname for me and we joked that it was a perfect code name for a secret agent.” He had started calling me CC during the pandemic, in a Whatsapp group we created with other colleagues at the beginning of 2020 when we went for drinks for my birthday. Little did we know at the time that this group chat would became a lifeline for our all us during lockdown.
The realisation that I will never hear his voice again or will never receive any messages from him (he was very supportive of my writing and sent me a lovely message when I started this newsletter) made me realise that I wish I could speak to him at least one more time. Just to say, as he famously said once to a client over the phone: “Thank you for thanking me, but I should be thanking you too.” Which is a true testament of his character.
I guess you probably see where I’m going here and why it’s not surprising that people have been turning to artificial intelligence as a way to cope with the grief of losing a loved one.
Like founder of HereAfter AI James Vahlos, who in order to preserve the memory of his father after he passed away recorded his memories and then programmed an AI-bot (which he called Dadbot) where he uploaded them to be able to interact with his late father. As a result, Vahlos founded HereAfter AI, a grief tech company that transforms the memories and the voice of loved ones into legacy avatar storytellers that are able to share meaningful moments about their life with close friends and family long after they pass away.
While HereAfter AI was created in 2019 and it is often referenced in news and articles about the rise of grief tech as one of the pioneers on the digital afterlife, it is certainly not the only company operating in this space. The rapid development of artificial intelligence in the past couple of years means that all aspects of our lives have been impacted, including how we cope with death and the pursuit of digital immortality.
The past few years have also seen a surge of companies offering services that can also ease the blow of having to deal with loss by making it easier to erase the digital presence of a deceased one in social media to companies that manage the practicalities of the process of being unalive and help you plan your funeral.
Apps like Bestow and Farewill let you select burial or cremation options, the music you’d like for the service, and how your earthly remains should be handled. For those of you who care about sustainability and reducing carbon emissions, companies such as Recompose (which offers a “natural organic reduction” service that turns bodies into soil) and Aquamation (which dissolves remains via a chemical process) provide an alternative to cremation that has a lower carbon footprint.
The reality is that the intersection of AI with one of life’s unavoidable stages is having a moment and the death tech (or grief tech) sector is estimated to be worth over $100 bn.
With more sophisticated technology available, it is also easier than ever to recreate the physical likeness and tone of voice of a loved one that has passed away in an attempt to help us accept their death and cope with it. But like any other application of artificial intelligence, companies operating in the business of death expect to reap financial rewards out of the services they provide, and so we must not forget that for them grief is a commercial transaction as people are willing to pay thousands of pounds to be able to interact with an AI-powered avatar or bot of a deceased loved one.
It is unsurprising that China, being the technological advanced society that it is and where development of AI is seen as the key to the modernisation of Chinese society, is leading the way on the digital afterlife.
And of course we can’t have a conversation on the digital afterlife and the surge of grief tech AI-companies without discussing the ethics of it.
AI ethicists from Cambridge’s Leverhulme Centre for the Future of Intelligence have called for the need to draw clear safety protocols to prevent social and psychological harm as the way we mourn and what happens after death can be radically transformed as a result of exposure to a new permanent afterlife in the form of an avatar or chatbot.
On a reserach paper published in early May and called “Griefbots, Deadbots, Postmortem Avatars: on Responsible Applications of Generative AI in the Digital Afterlife Industry” these Cambridge researchers have outlined the potential risks of the deadbots and how they could be misused in a number of ways, including to advertise products and dead stalk loved ones, but also making them feel exhausted by having incessant interactions with an AI avatar they may not be able to easily get rid of.
Which sounds like a very real life version of Dream Scenario and therefore not reassuring at all. If you’ve watched the movie, you’ll know what I mean. If you haven’t watched the film, hurry up because this is coming our way.
But there’s also the question of grief itself and how we process it.
Can interacting with a deadbot or avatar help us cope with loss? Or is it on the other hand making people stuck into a mourning cycle?
When psychiatrist Elisabeth Kübler-Ross published On Death and Dying in 1969, a seminal book stating the five stages of grief people experience after losing a loved one (denial, anger, bargaining, depression and acceptance) she did so in a world where the possibility to contact the dead was limited to hiring a medium and hoping for the best, or worst.
Which by the way was a common practice after WWI as families hoped to contact one last time those who had perished during the conflict through a seance. Today companies like Seance AI (yes, it’s called exactly that) offer a XXI century version of that experience where thanks to LLM and AI they can produce a digital psychic who summons a digital representation of a deceased loved one so that the living can have one last conversation with them and get closure.
But a world where the dead can -if we so wish- be still present and interact with us in real time, albeit in pixalated form, leaves little room for accepting they are no longer with us in a very definite, permanent way. It’s like unfollowing an ex on Instagram but knowing you can create a fake profile to check their stories - you can’t really move on if you are given the possibility to get a glimpse into their lives without you. Or so I’ve been told.
Not to speak of the question of who owns the deceased data and likeness (image and voice) to create their avatar or power a chatbot with their memories, how that data is disposed of once there is no need for the avatar, how long we can have ownership of that avatar, or access to the services that bring it to life.
What happens when a member of a family dies and some relatives would love to recreate an AI-powered avatar/bot to ease their grieving and others don’t as they feel they won’t be able to move on otherwise?
As it is becoming the standard with anything regarding the disruption AI is causing in areas of our life that have remained pretty much static for millenia (people have died in the same way since the beginning of time: quite definitely and permanently), I’m afraid that regulation on these matters will take years to catch up and by then we may be dead - at least in physical form.
As I’d like to encourage you to think openly about the impact of AI in the way we deal with grief, loss and death, I’d like to bring to your attention the work of AI Researcher Ginger Liu, who specialises in the digital afterlife, and who has a podcast and LinkedIn newsletter exploring in more detail this subject.
In the episode “Rise of the Griefbot: AI Chatbots that Replicate the Dead and Provide Grief Support” Liu argues that the work of AI companies operating in the grief tech space is no different from the role photographers played at the end of the XIX century when they sold postmortem portraits.
For her, the services provided by companies in the business of the digital afterlife fool us into thinking we’re having a real conversation with a loved one that has passed away, but it’s something we opt in because we want to be fooled, at least for an instant, to preserve that connection a little longer.
News
The creative industries are vital to British soft power
The Royal Opera House is changing its name to the Royal Ballet and Opera
Financial Times signs licensing deal with OpenAI to access articles for ChatGPT queries
Creative tech festival SXSW to launch in London in June 2025
The Yard Theatre is to build a new permanent home in Hackney Wick
A Willy Wonka reality competition show is coming to Netflix (can’t be worse than the Willy Wonka Glasgow experience, whose epic fiasco has inspired a documentary as well as a musical)
Music by UK recording artists generated £750m in royalties on Spotify last year
TikTok is to start watermarking when content has been made using AI
Ukraine unveils AI-generated foreign ministry spokesperson
London College of Fashion’s Fashion Innovation Agency is turning 10 this year and it’s marking the occasion with a series of events on all things fashion, AI, immersive technology and innovation
Last but not least
Italian cinema is having a terrific year as confirmed by the films that were nominated for best film at the 69th edition of the David di Donatello Awards.
In the past month I’ve managed somehow to watch There Is Still Tomorrow (the acclaimed directorial debut of actress Paola Cortellesi), Io Capitano (a wonderful tale of hope, compassion and survival by Matteo Garrone), and Palazzina LAF (another gem from Michele Riondino, who like Cortellesi is an well-known actor making his directorial debut with this film that denounces the abuses committed against workers at the ILVA factory in Taranto).
Aside from any of the above films, and given the topic of this issue, I wholeheartedly recommend Alice Rohrwacher’s magical realism gem of love, death and loss La Chimera, with an excellent Josh O’Connor in the role of Arthur, a taciturn archeologist who makes a living by robbing Etruscan tombs in the Italy of the 80s.
This is a beautiful film that has the structure of the best fables and ultimately it is a tale about people who lived in a distant past but who never truly left us as well as about people who live in the present but have long abandoned it. As Arthur discovers in his pursuit of Etruscan treasures and lost love.
The Age We Live In Now is a tech and creative newsletter with a focus on artificial intelligence. In each issue I discuss how our world is rapidly changing as a result of new technologies. And as a bonus, great cultural recommendations.