Humans are the only species on the planet to recognize their mortality. Throughout their entire history, humans have been trying to overcome death and find a path to immortality, be it in this or another world, their own or in a new body. With the advent of digital technologies, this dream now seems nearer than ever: if the human body can never be made eternal, can we transfer the soul into a more reliable vessel?
If a machine can in principle think, it can think like a person, and we could transfer a specific personality to it. If our brain is a supercomputer, then it’s highly likely we’ll find a way to make a backup of the system.
Sci-Fi Ideas and Futurists’ Predictions
Lem Stanislav was one of the first to talk about this idea in popular culture. In his work Star Diaries, electrical “backups” permit a person to basically be brought back to life immediately after death in a body no different from their old one and with a complete set of memories, such that the person brought back to life doesn’t even notice what had happened.
In Charles Stross’ landmark futuristic novel Accelerando published in 2005, the idea of digital immortality was taken a step further. In the first novella of the work, the human brain is “moderately” strengthened by an exocortex, i.e., external processors and storage in a form reminiscent of modern augmented reality glasses that are an important part of an individual’s personality. At the time, the KGB was conducting the first experiments into uploading the consciousnesses of squid into the digital space. At the end of the novel, the characters already “lived” as copies uploaded into a powerful processor.
According to a prediction made by famous futurist Ray Kurzweil, the age of technological singularity and digital immortality will come as early as in 2045. However, in 2017, we are still approaching the stage of those squid: the scientists working on creating connectomes, or digital maps of neural connections in a living organism, have only managed to upload the brain of an earthworm into a computer (or a device based thereon, it really does demonstrate a behavior similar to that of a C. elegans nematode). Also, scientists are working on connectomes for mice, owls, and fruit flies.
A project for creating a connectome of the human brain conducted jointly by several scientific organizations seems over-ambitious. However, at one point in time, the same was said about decoding the human genome.
The problem is not just in the transfer of an interaction map for over 80 billion neurons into a computer. The theory that the system will be able to respond to input signals just like the person after which it was modeled has not been fully proven.
We still don’t know how our brain works. Modern technologies operating at microlevels help create copies of brains of only the deceased, and non-invasive methods, such as MRI, don’t operate at the neuronal level. A living brain is a constantly changing structure, and so far no technology has been able to reproduce these changes.
Attempts to create a different media for our consciousness, or at least to connect a living brain to a processor of some sort, are still very different from those described in science-fiction works.
The 2045 Project launched by Russian businessman Dmitri Itskov in 2011 involves saving a living human brain in the body of a robot by 2025. The transition to an artificial brain is being postponed until more advanced technologies are created.
The Imitation Game
Many of the aforementioned technologies are designed to extend the life of a specific individual, but death, just like immortality, has two sides: those who leave (even though, like Woody Allen, they may prefer to live on in their apartments rather than their offices) and those who remain, and their desire to keep a deceased person for themselves to converse with them.
This was the path taken by the founders of a start-up called Luka that created a chat bot imitating the speech of Roman Mazurenko, the former art director of the Strelka Institute, who died in a car crash. The bot became a digital monument of sorts, a shadow person that gave friends an opportunity to feel the person’s presence once again, if only for a moment.
Later the authors launched the Replika Project, in which users created their own bots and chatted with them. Artificial intelligences based on neural networks that learn from user responses and social network materials will in time behave more and more like their creators.
The creators of eterni.me think it’s a good idea to set up an avatar that your loved ones will talk to after you’re gone. Using information from personal social media, a person may store memories and thoughts they think are important in an artificial intelligence. The system learns during conversations and answers questions better the more someone spends time with it.
There are a number of services for creating avatars for deceased famous people and other living stars and politicians that allow them talk about themselves or answer questions or make museum exhibitions more lively.
The goal of artificial intelligence is not to exist, but to appear to exist. It’s clear that the approach offers not only storage but also the design of personalities, which provides a different set of options from simply using available texts to train it. We may select what information to store for our descendants and what thoughts are extraneous. An avatar with artificial intelligence will not be an exact copy of a person as much as it will be a more or less successful attempt to create an improved version of oneself.
In the event that certain people would like to see a specific side of a deceased person’s personality, and a lot of information has been stored in this regard, then the approach will be successful. For example, digital copies of deceased actors are regularly being used in film, rapper Tupac Shakur has appeared at music festivals as a hologram, and fans of the rock group Grazhdanskaya Oborona now have the ability to evaluate a new album written in the style of lead singer Egor Letov.
Yandex employees Aleksei Tikhonov and Ivan Yamshchikov trained a neural network to write poetry, uploaded Letov’s texts, gave it the required rhythmic structure, and then composed the album “Neural Defense.”
The texts produced by the neural network could not always be distinguished from the author’s.
Despite that such an “imitation game” is a much more practical idea than a digital brain, not every question that digital immortality creates can be answered by technology. A person may only learn to exist alongside his or her digital copies. These technologies already allow bots to be created that may stay online in place of oneself and be taught what comments to like and even to write short comments, giving the illusion that one is online. The question is how your friends will feel once they find out nothing you said was real.
The idea of chatting with a recently-deceased grandfather in a messenger totally destroys our ideas of how we should observe a period of mourning; indeed, many would consider such behavior sacrilege. In this regard, at eterni.me more than 35,000 people, it would appear, want others to be able to talk to them after their deaths.
Author: Anna Degteva