Intelligence is not Artificial

Why the Singularity is not Coming any Time Soon And Other Meditations on the Post-Human Condition and the Future of Intelligence

by piero scaruffi
Cognitive Science and Artificial Intelligence | Bibliography and book reviews | My book on consciousness | Contact/feedback/email
(Copyright © 2018 Piero Scaruffi | Terms of use )

(These are excerpts from my book "Intelligence is not Artificial")

Mind Uploading and Digital Immortality

Of all the life extension technologies proposed so far, perhaps none has captured the imagination of the machine-intelligence crowd more than mind uploading. Somehow the connection between the Singularity and digital immortality was made: at some point those super-intelligent machines will be able to perform one great task for us, upload our entire self and "become" us. Couple it with the immortality of the "cloud" (see later), and your "self" becomes immortal. It will be downloaded and uploaded from one release of the Singularity to the next one for the rest of time.

In the most memorable of Isaac Asimov's short stories, "The Last Question" (1956), humankind is preserved in cyberspace after the end of the universe. Some forms of mind uploading already appeared in Arthur Clarke's novella "The City and the Stars" (1953) and in Frederick Pohl's short story "The Tunnel Under the World" (1955). I find the latter more realistic because it envisions mind uploading as a trick devised by the advertising industry. Wesley Barry's film "Creation of the Humanoids" (1962) envisioned the technology to upload memories and emotions of a dead man into an immortal machine in order to save the human race from extinction after a nuclear holocaust.

The technology of uploading a human mind into a computer was first explored by a geneticist, George Martin, in "A Brief Proposal on Immortality" (1971). He foresaw that some day computers would be so powerful that they will be able to do everything that a brain can do. Therefore why not simply port our brains to computers and let the computers do the job. Needless to say, philosophers are still arguing whether that "mind" would still be "me" once uploaded into software instead of gray matter.

Hans Moravec speculated that you are just a pattern, therefore you could "transmigrate" to a different body ("Dualism Through Reductionism", 1986).

That vision became more realistic in the 1990s with the explosion of the World-wide Web. A paleontologist, Gregory Paul, in collaboration with a mathematician, Earl Cox, speculated about cyber-evolution that could create non-human minds in "Beyond Humanity" (1996), including the idea of immortal "brain carriers" to replace our mortal bodies. In the days when television was still influential, William Gibson, the science-fiction writer who a decade earlier had invented the term "cyberspace" ("Burning Chrome", 1982), contributed to popularize the concept by scripting an X-Files episode, "Kill Switch" (1998), in which a man uploads his mind into cyberspace.

Ray Kurzweil wrote the article "Live Forever Uploading The Human Brain...Closer Than You Think" (2000).

Then came the deluge with books such as Richard Doyle's "Wetwares - Experiments in PostVital Living" (2003) exploring all sorts of technologies of immortality. Every year the vision of what Martine Rothblatt calls "mindclones", implemented in "mindware" (the software for consciousness), has to be updated to the latest computer platform.

In 2012 a Russian tycoon, Dmitry Itskov, pretty much summarized the vision of the immortality field: firstly, build brain-machine interfaces so that a human brain can control a robotic body; secondly, surgically transplant the human brain into a robotic body; and, finally, find a method to achieve the same result without the gory surgical operation, i.e. a way to upload a person's mind into the robotic body or, for that matter, into just about anything.

The question, of course, is whether that "you" will that still be you. Or just a machine mimicking you? You can ask someone to impersonate you, but that does not mean that he or she is you. That someone has absorbed the "pattern" of your behavior, but s/he is not you. By the same token, if a machine absorbs some pattern found in your brain, it doesn't mean that the machine has become you. We literally don't know what in the brain makes you "you" (and not, for example, me). This disembodied and reconstituted "mind" might well be immortal, but is it you?

In other words, this program is predicated on the assumption that "i" am entirely in my brain, and that my body is simply a vehicle for my "i" to survive. If so, such body can as well be replaced by some other material substrate. The brain is disposable, according to this view: the brain is merely the organ of the body designated to host the processes that construct the "i", but the "i" truly is only those processes, which, luckily for us, turn out to be information-based processes, which, luckily for us, can be easily transplanted from the mortal (and, let's admit it, quite repulsive) brain into the kind of information-processing machines that we started building in the 1940s and that are getting more and more powerful, rapidly approaching the capacity required to simulate the entirety of those brain processes.

This movement has revived the project of whole brain emulation. Ray Kurzweil and others have estimated that "artificial general intelligence" will be achieved first via whole brain emulation. The basic idea is to construct a complete detailed software model of a human brain so that the hardware connected to that software will behave exactly like the human would (which includes answering the question "is it really you?" with a "yes").

But first one needs to map the brain, which is not trivial. In 1986 John White's and Sydney Brenner's team mapped the brain of the millimeter-long worm Caenorhabditis Elegans (302 neurons and 7000 synapses). As far as i know, that is still the only brain that we have fully mapped. And it took twelve years to complete that relatively simple "connectome". The term was coined in Olaf Sporns' "The Human Connectome, a Structural Description of the Human Brain" (2005). A connectome is the map of all the neural connections in a brain. In 2009, a few years after the success of the Human Genome Project, the USA launched the Human Connectome Project to map the human brain. The task, however, is not on the same scale as mapping a worm's brain. The entire human genome is represented by about a few gigabytes of data. Cellular biologist Jeff Lichtman and Narayanan Kasthuri estimated that a full human connectome would require one trillion gigabytes ("Neurocartography", 2010). Furthermore, we all share (roughly) the same genome, whereas each brain is different. The slightest mistake and oops they may upload the brain of someone else instead of yours.

Once we are able to map brains, we will need to interface those brains with machines. This may actually come sooner. In 1969 the Spanish neurophysiologist Jose Delgado implanted devices in the brain of a monkey and then sent signals in response to the brain's activity, thus creating the first bidirectional brain-machine-brain interface. In 2002 John Chapin debuted his "roborats", rats whose brains were fed electrical signals via a remote computer to guide their movements. His pupil Miguel Nicolelis achieved the feat of making a monkey's brain control a robot's arm. In 2008 the team made the monkey control a remote robot (in fact, located in another continent).

By the time science is capable of uploading your mind to cyberspace most of us will probably be dead, and with us our brains. That disturbing thought predated the very science we are talking about. Robert Ettinger's book "The Prospect of Immortality" (1962) is considered the manifesto of "cryonics", the discipline of preserving brains by freezing them. It was actually cryonics that started the "life extension" movement. In 1964 another founding father, Evan Cooper, launched the Life Extension Society (LES). In 1972 Fred Chamberlain, a space scientist at the Jet Propulsion Laboratory, founded the Alcor Society for Solid State Hypothermia (ALCOR), now called Alcor Life Extension Foundation, to enter that business.

The similarities with the most successful organized religions of the Western world are too obvious to be overlooked. The end of the world is coming in the form of the Singularity, but, not to worry, we will all be resurrected in the form of mind uploads made possible by the super-machines of that very Singularity. The only difference with the ancient Western religions is that people from previous ages are dead for good, forever: we don't have their brains to upload anymore. But then maybe those super-human machines will find a way to resurrect the dead too.

Back to the Table of Contents

Purchase "Intelligence is not Artificial"
Back to Cognitive Science | My book on consciousness | My reviews of books | Contact