StoryFile is an a.i.-powered video company that sets up conversational interactions with dead people. No, not seance style, but on whichever screen you use to watch videos. I saw an ad of theirs featuring Captain Kirk as spokesman. We’ve still got a ways to go before we can order our devices to beam us up.
This company and others film a person’s responses to a series of questions. The more questions customers want answered, the greater the inventory of responses that need to be collected and processed. StoryFile’s site has dozens of stories telling how museums around the word are using their technology
Museums really like having videos and models of people telling stories to visitors
The Walmart Museum has its very own Hologram Sam in Bentonville, Arkansas. This digital image was created using technology that merges audio and video with the ability to answer questions in real time. The museum describes the image as “authentic.” This stretches mighty thin because authentic is defined by most sources as “not a copy.” Plus, Sam was flesh and blood. Hologram Sam is pixels.
Advanced conversational video technology
ADCONVIDTECH (?) provides what is described as “an immersive educational experience.” Visitors get to ask Hollow Sam, the Hologram questions about Walmart’s history, business philosophy, and Walton’s personal insights, receiving authentic responses that reflect his legacy.” I’d be surprised if Sam is programmed to reveal much personal history outside those involving his stores.
It’s not hard to figure that Walmart spent a lot of money, time, and talent on putting this advanced conversational video display together
How lifelike Sam appears or sounds or reacts to questions depends on those who programmed him. The tech companies involved actually tout their products as “cutting edge,” of course, a term that has been around so long it has whiskers.
One of the big tech sector growth areas is Video Grief
The life expectancy in the U.S. is 78.4 years. There are around 70 million people in the U.S. who are between 60 and 80 years old. The math is easy.
At the low end, companies sell so-so images that answer a small number of pre-programmed questions
More money gets you 3-D images that answer your questions in real time. All a.i./Large Language Models require a serious investment in collecting, using, and constantly monitoring the information that feeds them.
A major limitation of many current offerings in grief tech
They’re not only generic, but are also ignorant of who they’re “talking” to. They don’t know who you are. To them you’re nothing but a keyboard prompt. They’ll give the same replies to whoever asks the questions. Think Chatty Cathy, but without the string.
Susan Dominus told a story in the New York Times about an 83-year-old man who found out he will die within a year
Cancer is the bad news. Knowing he’s got a year left gives him and his family time to do and think about a lot of things. The good news is that the 83-year-old’s son hired StoryFile.
The son said, “It feels like watching a grandfather tell grandkids great stories from when he was younger”
This is great, unless you’d rather remember him from the years when he was younger and healthier and more active and you did things together. Not everyone has that chance, though. It’s also unlikely that these structured videos will feel real to everyone. After all, how often did dad/granddad tell stories while staring straight ahead at a camera in a room with enough lighting to melt the candles?
As a researcher who has interviewed thousands, I’m aware of how easily external factors affect people’s behavior
In this case, grand-dad sat in a chair and answered questions for five straight hours, according to the New York Times. It reminded me of the old-time photographs that took so long the subjects’ heads had to be kept immobile by a large clamp.
The man stared straight into the camera lens the entire time
Why is he so far from the camera? Won’t those socks get more annoying each time you view it? And all those bright lights that make people look sickly unless they use stage makeup. What other things do you think have an effect? Having everything you say recorded by videographers while light and sound crews watch isn’t what I’d call natural, either.
Some viewers claim they are making eye contact, but it ain’t so
From what I’ve seen, the dear departeds are making eye contact with a camera, not with a person they can’t see on the other side of the lens. I’d rather have Old Grand-Dad tell all his best stories so we can listen to him tell them over and over.
For me, dead-gone-remember beats dead-gone-pretend, but I’m only hypothesizing
As long as I’m producing my own show, I want each story to have the departed dressed differently, sitting in a different chair in a different room with a different background. I’d want to get them on video in casual, relaxed, impromptu situations with no studio makeup. And as I don’t need them to stare at me while they tell their favorite stories, I’ll add some variety by using several different camera angles.
MIT Technology Review News Editor Charlotte Jee’s article in the October 18, 2022 issue asks us if we’re ready to “speak” to our dead relatives
Jee tells us that using technology to talk to people from the past has been a science fiction staple for decades and is “an idea that has been peddled by charlatans and spiritualists for centuries.”
Call them what you want: digital clones, replicas, ghosts, or virtual relatives, they are not beings with feelings for you or anyone else. How many buyers/users do you think will not be able to understand they are not talking to a real person? A few? A lot? Certainly some.
For the sake of a simple descriptive reference, we might call them virtual beings
As with all Large Language Models, the emphasis is on large. The more information you feed the machine, the more raw material the computer has to learn from. For that training to happen properly, humans must provide constant ongoing feedback to correct errors as they appear.
In 2020, I wrote What Can Mickey Mouse Teach Us About A.I.?
In the Sorcerer’s Apprentice, Mickey turned brooms into robots, gave them instructions, set them in motion, and left them unsupervised. They caused a disaster so enormous that the sorcerer was the only one able to put things back in order. I don’t think the a.i. industry has a sorcerer, but they do have robots that follow the instructions they’re given and turn decision-making over to black boxes using secret math formulas.
Automated decisions, answers, and replies are based on flawed data and biased algorithms
They are not objective and cannot be trusted. It’s like giving the car keys to a seven-year-old and sending the child to do your shopping. What could go wrong?
Well, Large Language Models need ongoing feedback and re-priming to “learn”
Without good training, they never do. Most don’t know that the people “tuning” these Large Language Models are paid less than fast food workers.
What is called lifelike now will be snobbed on in the future
Some of you may remember when our computer screens were no bigger than one hand and what we typed showed up on the screen in neon-green or weird orange. These days, so many people are so accustomed to high resolution that they can’t imagine such prehistoric times. 3-D and holograms will get so much better that today’s tech will appear primitive to successive generations, which is pretty much how things work.
One day, will avatars no longer be stuck in one spot?
They’ll be getting up and walking around with us, won’t they? One day, won’t avatars be more than just images? Once the developers come up with sense of touch, millions will be lining up to hug grandma.
As long as you’re thinking about death and dying
If you go shopping for Video Grief services, make sure you are getting what you want. Some readers have found my December 4, 2023 article, Who Is Elizabeth Kübler-Ross? and my March 18, 2024 article, Why Do We Have Funerals? helpful in addressing the subject of death and dying.
One of the best applications of these technologies is in recording and presenting oral histories
I’d like to hear Ben Franklin telling me about the kite, Dwight Eisenhower telling me stories about D-Day, and Paul Revere describing his ride through the night. Yes, those stories would be enhanced if they all appeared and acted like their real selves in museum exhibits. Maybe they’d start talking when you triggered a sensor, like the lights in the freezer cases at the grocers that are turned on and off by people walking by.
Bonus
Terry Birdwhistell and I were undergrads together, living in the same house. He was one of the distinguished group that created the University of Kentucky’s Louie B. Nunn Center for Oral History program. The Center has more than 14,000 oral history interviews emphasizing 20th century history. The interviews cover more than 600 subjects, including horse farms, quilters, tobacco farmers, teachers, Appalachia, and Peace Corps Volunteers.
Birdwhistell interviewed and documented the lives of the famous and the unknown for 50 years until his death in 2023. He also interviewed thousands of students, teachers, and administrators. None of his interviews was with a tele-clone.
I also interviewed thousands of unknowns, recording them on video as I listened to people tell their stories their way, naturally, with as little interference as possible. Bird’s interviews are archived at the University of Kentucky. Mine, if they are still around, are proprietary, and seen by no one but the client who sponsored the study.
Extra credit if you recognized the name of a former Kentucky governor.
