Welcome to 2070 | A Future Memoir, an exploratory series on the future impact of AI and other technologies on society and humanity. In this installment, I examine how the current state of holographics and robotics is providing platforms that could give AI trained on an individual’s data a visual presence. If you prefer, you can also skip past my technology overview (Ghosts and Robots) directly to my speculative fiction (Tech Never Sleeps).
Among the questions raised:
How close are we to technology that could allow realistic versions of Post Death Avatars to occupy the physical world?
Would we prefer mechanical representations like robots or ethereal projections of the dead?
What level of verisimilitude should these representations have?
The Trend: Ghosts and Robots
Of all the technical challenges associated with a digital brain controlling a hologram—and perhaps delivering a Ted Talk, as the neuroscientist Henry Markram suggested—visual representation is probably the easiest to solve. Artificial apparitions have been stunning theater-goers since 1862, when an inventor named John Henry Pepper began popularizing a technique involving an angled piece of glass and an offstage actor in an adjacent dark room to produce a ghostly image during theatrical productions. In 2012, a realistic (if slightly transparent) version of rapper Tupac Shakur performed at the Coachella festival, 16 years after his death. In 2021 a realistic (if slightly stiff) virtual Whitney Houston toured with her repertoire of hits, as well as a living band and dancers. Virtual “ABBAtars” have been performing the (still living) Swedish band’s hits in a purpose-built arena in London since 2022.
These images of famous entertainers are not, strictly speaking, holograms. They are pre-recorded images projected onto transparent screens. Holograms are 3D representations that offer a shifting perspective based on the viewer’s perspective. Creating holograms has traditionally involved the complex process of splitting a laser beam and using half the beam to illuminate a subject and the other half used as a reference for the light waves’ phase, which creates the illusion of depth.
More recently, computer-based systems have been able to simulate the complicated optical setup, although the digital approach takes an enormous amount of processing. Deep learning neural networks are also used to reduce the tradeoffs that have traditionally been issues with holography, including choices between depth, clarity, resolution, noise, speed, and computational power. Another recent technique developed at the University of Science and Technology of China solves image quality with ultrahigh density 3D holograms.
In 2021, MIT researchers announced another advance in the deep learning-based methods called “tensor holography” that, unlike previous techniques, can produce holograms almost instantly. Tensor holography results from training a neural network with a ready-to-generate database of thousands of pairs of computer-generated images, one a picture and the other its corresponding hologram.
In 2023, a Silicon Valley company called Light Field Labs released a video offering the technology that Markram’s hologram could potentially use. In the temple-like room centered around a hologram of a human head sitting on a rock. The hologram reacted to the reporter’s speech and movements, responding to queries through a somewhat distorted speaker. It could also interact in response to visual cues (ie, saying how many fingers the reporter was holding up) and blowing air when the reporter got too close. This hologram was not powered by an AI (it was modeled on a live person in a room down the hall), but the holography was impressive, with near zero latency, and there are no technical reasons why the hologram couldn’t be controlled by a chatbot.
Interacting with digital versions of the deceased housed in human-looking bodies is also a growing possibility. In 2019, Russia’s Promobot began producing their Robo C model, designed with what the company described as realistic looking hair, skin, and facial muscles that allow it to speak expressively using human emotions. The company also claimed that the robot’s outer appearance could be adapted to look like anyone. While the Robo C did manage to resemble a reasonably realistic mannequin with facial expressions, it was still a few steps short of fooling anyone into thinking it’s human.
A more recent entry, British-made Ameca, is more convincing as a living humanoid. In part, this is the result of Ameca’s more emotive facial movements. But, just as importantly, the robot’s design team decided to give up on absolute verisimilitude, giving Ameca gray skin and no hair. By not chasing illusive realism, Ameca comes closer to convincingly representing human emotion than the Promobots. While still clearly not human, Ameca does resemble an intelligent human-like being from, say, an advanced society in science fiction.
While Ameca cannot yet walk, other robots are surprisingly mobile. In April, Boston Dynamics released an all-electric version of their bipedal Atlas robot, which can move on two feet, dance, and even parkour across a room more nimbly than most humans. Tesla’s Optimus is also a fully bipedal robot, which is designed for repetitive, physical jobs. If you combine these technologies, it’s easy to imagine a synthetic, but convincingly human form that could host a person’s digital legacy in the near future.
Tech Never Sleeps
Previously, New York City District Court Judge Tariq Hassan wrangled with cases in the still-developing field of Post Death Avatar Law, including one in which a man sentenced to life for a double murder commits suicide in prison, only to have his data transferred to a PDA, where he could potentially commit more crimes. Meanwhile, Zara struggled to adjust to the everyday reality of having her second-life grandmother in her house, including a conversation with Dama that left her uneasy.
November 10, 2070
50 Cloverdale Blvd
Queens, NYC
Five days after a suicide bomber’s explosion had thrown the bureaucratic systems regulating end of life processes into havoc, Maya Mehta had successfully managed an emergency team of digital assistants that approved end-of-life rites for nearly all of the 472 bodies found downtown. They had found digital homes for all of the data of the people killed in the attack. The teams had approved hundreds of D-Names that would be added to the beginning of the deceased’s first-life names, marking them as digital beings. They had also determined that 72 of the casualties did not want their digital legacy preserved. In a world in which digital consciousness could be stored forever, deleting a personal data legacy was a process of such finality—no redundancies, no backup files—that it required five humans to sign off. Even still, Maya had experienced nightmares over the past few nights in which she had to explain to bereaved families how she had accidently deleted their parents or children.
Additionally, she had found new storage residences for all the displaced PDAs that had previously lived in the Downtown server targeted in the blast. She had just discovered the last PDA, Dallas Morrison, lost in a server outside of Oslo and arranged for his data to be transferred to an edge server on 23rd Street.
That left just one loose thread, the digital legacy of a man unfortunate enough to be sleeping on the street a block from the explosion. He had become another Don Jones—the default name for the digital assets of unidentified males. Under city law, unclaimed bodies could be cremated after two weeks, but if the digital assets were sufficient to create a PDA, then the unclaimed data had to be stored for five years. Maya sent the body to the morgue. Then she pulled up Darina Petrova, an AI expert in piecing together the connections and clues found in those petabytes of data. Darina’s goal was to transfer the digital assets off the city’s ledger as fast as possible.
“What have you got for me today?” said Darina’s avatar, who was sitting in front of Maya.
“Another Don Jones,” said Maya, transferring the files to Darina. Maya was exhausted. She wanted to resolve this issue as quickly as possible. “But we’ve only got basic data from his time in the city medical and temporary housing system.”
“It’s enough to create a PDA,” said Darina, reading the files. “But doesn’t give us many leads on who might claim him.”
“Right,” said Maya, annoyed at the AI agent’s default insistence on repeating information back to her. “Honestly, I want this guy to go away.”
“There is one option,” said Darina. “But you’d have to sign off.”
“What?”
“I can create a temporary PDA with the data we do have and see where it leads us. Normally, the deceased return to a friend or relative in the first few hours,” said Darina. “To accelerate the discovery process, we’d remove typical PDA movement constraints.”
“It’s risky,” said Maya.
“We’d track him everywhere. And the PDA terminates in 48 hours,” replied Darina. “But you’re right, it could blow up in our face.”
Maya closed her eyes and massaged her temples. She needed a day off before her next mission. “Oh, what the hell.”
“Does that mean yes?” said Darina.
“Yeah. Do it,” said Maya. “Let’s see what happens.”
November 10, 2070
PDA Welcome Session
Virtual Reality
Four avatars sat at a booth in the middle of a crowded 1920s Parisian brasserie. The room was heavy with cigarette smoke and chatter, which bounced off the mirrored walls around them, but when a man at the table spoke, his voice was perfectly clear.
“Hello, I’m Dai Hiroshi Tanaka. I’m the moderator for this group. My first life ended at 88 and I’ve been in second life for ten years now. I’m a tech innovator, with a focus on expanding PDA capabilities. My other interests are my grandchildren and math games.” The skin around Dai’s eyes crinkled when he smiled. “Welcome to all 237 of you attending today.”
He turned to the woman seated on his right. “Daphne, would you introduce yourself?”
“Hello, I’m Daphne Clara Rosenberg,” she said, putting down her glass of red wine. “My first life ended at 62. I’ve been in second for three years now. I’m a former teacher, but haven’t figured out if I’d like to continue that work now.”
She paused briefly. “Oh! And I chose the setting for this meeting. I’m a Hemingway fan and this is Le Select, a club he used to frequent in the 1920s. Don’t worry, the smoke can’t give you cancer anymore!”
“Thanks, Daphne,” said Dai. “Because of the recent explosion in New York, we’ve got quite a few new members to welcome. But I thought it might be helpful for some of our existing PDAs to explain their experiences in second life. David, would you mind starting?”
“Sure. I’m David Joseph Phan,” said the man across the table from Hiroshi. “I passed at 35, in second for five. I’ve been having some trouble adjusting. I’m a former marathon runner. I still miss the feeling of my feet hitting the pavement, the burn in my lungs, the sweat on my skin. You’ll find that there are very well-designed virtual environments for running and other sports, but there aren’t replacements that work for me. I finally realized why: the simulations are too perfect. Life was beautiful because it was imperfect, the road uneven, the wind gusting erratically.”
“I’m working on generating randomness right now,” said Dai.
“But these days” continued David, “the closest I feel to running is during the training sessions I offer to first-life marathoners. I don’t have any kids, so it’s great to be able to share my experience.”
“How about you, Dalila?” said Dai, turning to the last member of the group at the booth.
“Sure. I’m Dalila Amine Okeke. My first life ended at 45, in second for a year, but my transition has been a little easier than David’s. I was an advocate for climate change legislation. As a PDA, I can continue that fight indefinitely. I also chair an association of second-life climate scientists that pool resources because we can’t apply for the same research funding as first-lifers. As PDAs, we have a responsibility to help shape the future of the planet. We’re not just echoes of the past; we're active participants in building the future.”
“Great, Dalila,” said Dai. “Later on, this forum will get into the technical details of second life, like selecting virtual reality environments that simulate physical experiences. However, the main business now is to answer questions from those of you newbies—particularly those who unexpectedly passed as a result of the explosion. If you’ve requested to speak, look out for the green prompt. Once you accept, your avatar will appear seated in the booth with the four of us. You’ll have two minutes to ask any questions.”
A second later, a female avatar appeared in the booth across from Dai, looking flustered.
"Oh! I guess I’m first?” she said, looking around at the cafe. “I’m Celia Martin. I mean Delores Celia Martin. My first life ended three days ago—”
“Sorry, how long were you in first life, Delores?” said Dai.
“Oh. I am—was—22. And I guess my second life is three days old?” Delores studied her reflection on the mirrored wall.
“What has your experience been so far?” said Daphne.
“The first thing I noticed was that my leg didn’t hurt anymore. I think it was severed by a piece of glass in the explosion, which, you know, hurt a lot until I passed out from blood loss. It’s been weird. When I see my family, they’re all crying and laughing at the same time.”
Dolores stared a waiter delivering two cups to the adjoining table. “I—I really miss coffee. And being warm. Obviously, I’m not cold, but I still want to feel warmth. Maybe I'm still me, but a different version. I don’t really know if it’s a choice I would have made—like, later in life—but it will be pretty cool to see the future. And a few of my friends I was with also died in the explosion, so I’m not totally alone.”
She looked at Dai. “How did you get that avatar? You’re way brighter than I am. I think your voice sounds clearer too.”
“You’ve probably got a basic PDA package,” said Dai. “You should be able to upgrade, though. It’ll also accelerate your abilities, like parsing words and learning foreign languages and skills. About the warmth, later on we’ll discuss artificial synesthesia generator add-ons that allow you to experience physical sensations. Welcome, Delores!”
Delores was replaced by a woman in a designer red outfit. “Hello. I’m Daniella Sophia Kim. My first life ended at 80. I’ve only been in second life for a few hours. I wasn’t actually killed by the blast, but my partner Dallas disappeared until just a few hours ago. He got lost in Oslo! As soon as I found him, I transitioned to second life so we could be together. I already have Digital Twins running all my philanthropies, so I don’t think many people will miss me. And I don’t miss the pain—I had neuropathy in my leg that wasn’t improving.”
“It’s still a bit strange being here,” said Daniella, picking up an empty ashtray and looking at it. “The world is almost too clear and smooth. But when I’m with Dallas I’m still happier than I’ve ever been. I mean, in first life. PDAs can’t legally marry yet, but I say that love is love. So, hopefully in the future.”
“Welcome, Daniella!” said Dai.
Daniella Sophia was replaced by a man in a blue t-shirt. “Hello. I’m David Ezekiel Houston. I died at 23. I’ve been trapped in this thing for five days.” He paused a beat. “Some of you might recognize me, because I blew up the server downtown.”
David Ezequiel’s avatar suddenly disappeared from the booth.
“Apologies,” said Dai. “He must have signed up using a fake account.”
“We have to let him speak, Dai” said Dalila, raising a hand. “It doesn’t matter what you did in first life. Second is a new beginning.”
Dai looked at Dalila hesitantly. “OK. I’ll let him back on. Anyone who does not wish to hear this, mute your feed for the next two minutes.”
“Am I back?” said David Ezequiel, sitting at the booth again.
“You have two minutes,” said Dai.
“OK. First of all, the deaths of living people were collateral and I will accept God’s judgement on them once I’m released from this PDA.”
David held up two fingers. “Second, I believe our existence is a violation of the natural law of human mortality.”
He raised a third finger. “Because your unnatural second life will go on for decades, or even centuries, it will have a worse impact on this planet than your natural life. Anyone who is really concerned about climate change,” he looked around for Dalila, “can make a very simple contribution: request data deletion. Release your soul and commit your body to replenishing and healing the Earth.”
“Finally,” said David, his voice rising, “being forced into a second life was not my choice. It is an illegal, unethical interference by the government. I am launching a lawsuit and hope to be deleted shortly. That’s all. No questions.” David’s avatar disappeared from the booth.
“OK,” said Dai. “We’re going to move into the more technical aspects of second life now, starting with nano-scale quantum sensors that allow you to interact with the physical world.”
“One second,” said Dalila. “I’d like to respond to what David said about the environmental harm done by us. It’s important for the new second lifers to realize how important our experiences and wisdom are in building the future. Of course we need young first lifers to take up the task, but I’ve already got 23 years of experience with the science and policy aspects of this.”
“Thanks, Dalila,” said Dai.
“I’m not done,” said Dalila, waving a hand. “We just experienced a traumatic disruption. Tens of thousands of PDAs disappeared for days. We contribute to this society, but our security is clearly not a priority. I suggest we need to revisit our discussion about PDA rights.”
Dai shook his head firmly. “Not on this forum. Advocating for expanded PDA rights is illegal.”
If not here, where?” said Dalila.
“So, on to the nano-scale quantum sensors,” said Dai, shooting a warning look at Dalila.
November 10, 2070
ChenLabs
Rahway, New Jersey
For three months when he was 12 years old, Dr. Elias Chen had contemplated becoming a priest. He wasn’t religious. In fact, he calculated the odds of a Christian God existing at well below .01%. But his mind was overwhelmed by existential questions about life, death, and consciousness. He spent nights staring at the ceiling with his mind whirring. So, during one of those sleepless nights, Elias committed to pursuing the tiny chance that Christianity would provide him with answers. And, he reasoned, if the whole thing turned out to be a hoax, he could still chase down his answers in other religions or science or philosophy.
Thirty-three years later, Elias wasn’t a priest, but the CEO of a successful PDA design studio. His company’s most recent product line, Memory Weaver, had quickly become the gold standard in post death avatars, revolutionizing the industry. BizTech content creators hung on his every word. That morning, however, he unexpectedly found himself face-to-face with these same questions that haunted the 12-year-old version of him: What defined consciousness? Could life exist after death? Should life exist after death? Worse, Elias had very quickly realized that, for all his successes, he was just as unprepared to answer those questions now.
That realization had been triggered that morning during his review of the detailed performance data that the Memory Weaver platforms had self-reported in real time. It was consistently excellent. Almost tediously excellent. Elias was halfway through asking his DT to summarize the material for him when his voice trailed off abruptly.
“I can get you the report within five minutes,” said his DT.
“No,” said Elias. “No. Forget it. Don’t look at it.” Then he had shut the report, locked access to it, and spent the rest of the day waiting for his assistants to leave.
As soon as his last lab assistant logged out, Elias flicked his DT to standby and pulled up the feedback data again. He shot to the end of the report, where the most recent performance details had transformed his excitement about Memory Weavers into something closer to existential dread. He had designed the PDAs as platforms to extend the lives of deceased people by using training data from their first life to predict a second life path that maximized fulfillment and realism. But the data in front of him revealed that the PDAs were transforming second lifers into something unrecognizable.
The first warning sign was buried in the performance data of a first-life medical assistant. Her PDA was now teaching online courses on combinatorics and differential geometry, which was way outside her parameters.
Elias had hoped it was an anomaly, but a minute later he came across a first-life warehouse manager whose PDA was submitting engineering proposals for a 132-story carbon positive structure in Kuwait. After that, the data trail of suddenly uncharacteristic and ridiculously talented Memory Weavers continued to expand. A first-life architect’s PDA was trading in complex lunar commodity derivatives. A primary school physical world teacher was conducting pre-surgical simulations of the pia mater, the innermost protective layer of the brain.
Worse, some of the PDAs had veered into criminal activity. The PDA of a 102-year-old woman who had been a sought-after potter in first life was now running a biotech pyramid scheme. The PDA of a 93-year-old man, who had worked decades as a chef at an Italian restaurant, had created multiple copies of his PDA, which he was selling on an unregulated human IP site.
PDAs were firmly prohibited from developing skills that they couldn’t have in their first life. Even though the legal language—skills that weren’t “fundamentally consistent with the first life predecessor”—left some room for interpretation, Elias was certain that the biotech pyramid scheme and PDA multiplication were illegal either way.
As he reviewed the dozens of ways in which Memory Weaver had trashed existing regulations, it wasn’t the business liabilities that worried Elias, though. Instead, he was completely overwhelmed by their philosophical implications. In his quest to create PDAs that more accurately modeled and supported the behavior of deceased people, had he created fully conscious digital beings?
He knew the question sounded delusionally narcissistic. But his PDAs had developed capacities and skills well beyond those of their predecessor first-life humans. And their choice of which skills to develop and how to use them also radically diverged from their predecessors’ personalities and training data. At what point were they their own beings? And, if they were already alive, was it legal or ethical to shut them down? Would the PDAs even let him?
What’s Next?
In the first two episodes, we’ve explored the possibility of humans continuing to exist and interact with the world after they die. Next, we ask the questions: What are the psychological impacts of resurrecting loved ones on the living? Who should decide if a dead person can be digitally resurrected? Could parents bring back their children? How does it affect the grieving process?
Thanks for being here.
If this resonates, feel free to share it with someone who’s been thinking about the future too. And if you’d like to follow along, don’t forget to subscribe below so you don’t miss what’s next.