memory-card
Social Networks

Eternime plans to create an immortal library of mankind

When we lose someone most of us are only left with a few photos, videos, and our personal memories of our loved one. The death is final. But startup Eterni.me believe it does not have to be and is working on an avatar that will learn everything about someone throughout their lifetime in order to generate an AI avatar of them, so in essence they “live on” and can still communicate with loved ones even after they are gone. Some have called this “creepy” and have questioned whether this is morally sound. We catch-up with Marius Ursache, Co-Founder and CEO of Eterni.me to find out more.

 What do you mean by “become virtually immortal”?

The idea behind Eternime is to preserve your “essence”—which is your mind—for eternity, and make it accessible to people in the future. We collect as much information about you as possible, during your lifetime, through analysing your digital footprint and having frequent conversations with an avatar. We then find meaning in this data and build it into the avatar’s memory. This avatar will offer other people an immersive way to interact with your memories (during, and after your lifetime). In this way, at least part of you will live forever.

Why did you come up with this product?

Right now, when one of us passes away, all that’s left are some photos, a few videos, a Facebook wall full of cat pictures or shared links, maybe a diary or (if we’re famous), an autobiography. Billions of thoughts, stories, memories, ideas, creations or discoveries are lost forever. Some of them are important to our family, our friends, or other people in our lives. Some of them are important at a macro scale, for the whole of humanity. But right now, most of them die with us, because they are only stored inside our mind, and a few memories of people around us. But those memories fade away, and we’re forgotten. Nothing important is tangible, preserved for the future. From a historical point of view, it’s like we never mattered.

We want to stop this. We want to save your thoughts and memories—for your children and grandchildren, but also for people in the distant future. And we want them to be able to interact with them by having an experience as close as talking to a real person as possible, instead of browsing a timeline or reading a biography.

There’s another reason, too. Our “big, hairy, audacious goal” is to create a huge library of mankind, with billions of people and their memories accessible for eternity. This could be one of the most important things ever created.

Could you elaborate on what you mean by “huge library of mankind”?

Anyone who has an Eternime profile will be “findable” by anyone in the future. Of course, when someone builds their avatar, they can choose four levels of privacy for their data: public, only friends, only family, private (for the user’s eyes only, helps build personality traits and other things, but does not give away the facts). All the information will be stored in Eternime and you could access it by chatting to the avatar. You won’t be able to download someone’s full profile.

How will the personality of the departed be digitally reconstructed?

Our initial plan and proof of concept was to mine data from email and social networks, but unfortunately that information is limited, and does not touch at all the essence of who we are.

The first beta version we are preparing to launch soon is an avatar who behaves more like a virtual friend who chats with you frequently to learn more about you. By replying to hundreds of handpicked questions (we have a psychologist on board), and recording those answers as videos or text stories, you help us collect and preserve your most important memories.

Think of it currently as someone interviewing you about your life, in order to create an interactive biography. And through these frequent conversations (at least weekly, maybe, for the next 30 years of your life), the avatar will learn more and more, and will become smarter. We’ll also get a lot of information and context from your digital footprint (Facebook, other social networks, email, wearables, smartphone), to complete your picture.

People constantly change and evolve throughout their lives. Will this not pose a significant challenge when trying to reconstruct a digital profile?

Of course this is a challenge, and we’re structuring the conversations with the avatar so that a lot of them are periodic. This way, we’ll see how a person evolves and keep older stories and memories, but always adjust the avatar traits based on the latest conversations and input.

How will loved ones be able to communicate with the departed?

Once you sign-up, you start talking to your avatar and teaching him facts about you. You can also allow him to access various digital streams—your Facebook, Twitter, emails, photos, geolocation history, maybe even Google Glass or Fitbit data. The data will be collected, filtered and analysed in order to “make sense” of it.

Once we have enough data, we’ll generate an AI avatar that emulates you (initially in text chatting, then as personality, then also as looks). It will take a long time for these to happen, but a lot of people don’t have the luxury to wait for the technology to be there. This is why at this stage our focus is on collecting and curating the contents of your mind (thoughts, stories, memories, ideas, creations, theories or discoveries), not on creating realistic avatars who know nothing.

So it’s more curating a digital legacy during your lifetime, which can be useful in many ways—and not a “social network for/with the dead” how it’s been dubbed by tabloid press.

What sorts of challenges will you encounter in emulating the user?

The biggest challenge is finding meaning in the huge amount of data, and to reconstruct someone's knowledge/memory. The artificial intelligence that we need requires a lot of effort and time—it may take years. A lot of the people who expressed interest in Eternime don’t have this time. Many of them have terminal diseases, incipient Alzheimer’s, or are just old. This forces us to use more or less the approach of the cryonics initiatives—preservation first, technology second. It also helps us develop the AI faster, since we need a lot of personal data for this.

The beta version will focus on collecting the information, and allowing the user to intervene in manually curating the data. In the future we’ll be doing this more and more using technology. It is a challenging project, and the technology is still far from being effective, not to mention “perfect”. But things will evolve along the way and by the end of our lives, the technology will be really effective and allow a more realistic and immersive experience.

Our goal is not to create science-fiction tech, but to preserve the most important thing about who we are—what’s in our mind. This has a small scale impact (honouring people close to us) and bigger impact (preserving important knowledge and information about all humanity, not only famous artists or scientists or politicians).

As this is quite a sensitive area, isn’t there a danger of making the grieving process worse?

We are very aware of the emotionality that is attached to the topic of death or to that of robot clones or avatars and the “uncanny valley” effect. A lot of people labelled our idea “creepy”. For us it is really important to emphasise that we do not want to preserve the banalities of the life of a person, but would much more like to create a legacy that allows your great-grand-children to learn more about their great-grand father—and beyond. Humanity is at the centre of everything we do, and we work with psychologists to avoid any negative-side effects in the future. I can’t say for sure there will not be any, but I think that the benefits will hugely outweigh any negative aspects.

How will you be working with social networks like Facebook?

We’ll use any part of your digital footprint to learn things about you—or at least to understand more about your environment and the context of events. Right now, this data is insufficient and just a small part of who you really are.

Think of a funny scenario where aliens discover Earth and the Facebook data centre long time after the human race became extinct—their conclusion after analysing the Facebook timelines of a billion people would be that we were a superficial kind, obsessed with cats and memes.

A lot of people will draw comparisons to the TV series Black Mirror, where a grieving woman communicates with her partner in digital form. What is your response when people ask you about that?

I watched that episode (as well as Transcendence and The Final Cut) after receiving other press questions & cues. While the underlying idea has some similarities, I think the “implementation” is different. It was really useful in reminding me that there are serious emotional and ethical issues attached to the topic. That’s the reason why right now we’re working with a psychologist/therapist to avoid any additional trauma.

And yes, no real-life creepy clones. We’re not trying to replace people, but honour their contribution both to their small world (family and friends) and to the big world.

How will you handle security issues?

Our current model is to charge a monthly subscription to all users, so we won’t need to disclose anything for advertising reasons. All the data will be private and only the user will choose who has access to what. We’re taking this very seriously, and it’s not only for privacy reasons, but also for preservation reasons.

How much interest have you had so far?

We have close to 30,000 beta signups, but we are initially focusing on a very narrow segment (seniors over 70 and people with terminal diseases), to launch the beta and receive feedback. For them, this is more than a “cool new product”, it’s something important and they are willing to dedicate time to build their legacy.

When will Eternime be officially launched?

We are working on the beta release that will come out this summer. It will be very limited, but our goal is to be able to collect the most important facts and memories from older generations, and be released to a limited number of people. Then, until the end of the year, we'll gradually release to all the people who signed up.

 

PREVIOUS ARTICLE

« Crowdsourcing Innovation: Floris Hoff, 3D By Flow

NEXT ARTICLE

An Interview with the New CMO of Emarsys »
author_image
Ayesha Salim

Ayesha Salim is Staff Writer at IDG Connect

  • twt
  • Mail

Recommended for You

How to (really) evaluate a developer's skillset

Adrian Bridgwater’s deconstruction & analysis of enterprise software

Unicorns are running free in the UK but Brexit poses a tough challenge

Trevor Clawson on the outlook for UK Tech startups

Cloudistics aims to trump Nutanix with 'superconvergence' play

Martin Veitch's inside track on today’s tech trends

Poll

Is your organization fully GDPR compliant?