Ghost in the Machine

The phrase, “ghost in the machine,” was coined by 20th century British philosopher, Gilbert Ryle as a means of characterizing 17th century mathematician and philosopher, René Descartes theory of mind. Descartes theory was an offshoot of the age-old dualism theories that dated back to Plato and Aristotle. His theory—referred to as Cartesian Dualism—distilled dualism into today’s mind-body theories, wherein “consciousness and self-awareness” were mind, and brain was “intelligence.” Descartes further went on to propose that the seat of the soul was the pineal gland.

The term soul, of course, was often used interchangeably with the term, consciousness.

By the late 19th century, the seat of the soul was upseated, by self-proclaimed mystic, Madame Blavatsky, who put forth the theory that the pineal gland was an atrophied “organ of spiritual vision“—what we refer to today as the “sixth sense” or more specifically, extrasensory perception. Of course, some individuals, especially skeptics and those prone toward the hard sciences will argue that the aforementioned were given to flights of fancy. And while the fantasy may provide relief in times of loss (i.e., the soul goes to a better place), we are nothing more than a sum of our biological parts, that work in unison—one with another—wherein behavioral and physical characteristics can be explained by a combination of genetics and our interatction within the greater ecosphere of society and nature.

This was about to change.

In the past decade, another school of thought has been emerging. One that acknowledges “consciousness” yet defines it within the physical constructs of the hard sciences, and begins with the machine.

In 2002, Ray Kurzweil and Mitch Kapor struck a wager. Kurzweil believes that by 2029, machine will pass the Turing test, as described by Alan Turing in his 1950 paper, “Computing Machinery and Intelligence.” Kapor isn’t so sure. Enough so, to wage $20K dollars. Arguably, the debate has less to do with what technology may (or not) do, and more to do with definitions.

For example, if one were to consider Alan Turing’s challenge in its most simplistic form, one might argue that artificial intelligence was achieved with the first and most well-known simulation, a Rogerian Psychotherapist, Eliza, written by Joseph Weizenbaum in 1966. Although we can assume that Kurzweil does not consider Eliza to have even remotely passed the Turing test, his vision still appears to be more toward a novelist approach—a machine that tells a good yarn. Kapor’s vision, on the other hand, appears to ask the question: can machine gain some degree of sentience—a soul, if you will—a ghost? Nonetheless, both appear to agree that in order for machine to think, it must contain vast knowledge.

Now, while this is certainly an interesting thought exercise, Kurzweil’s and Kapor’s wager is nothing new. In 1990, Hugh Loebner offered $100,000 to the first person who could write a conversational A.I. that would truly be indistinguishable from human. Thus began the first annual Loebner Prize contest. By the following decade, Dr. Richard Wallace had entered the contest and picked up three first prizes, with his Alicebot. Wallace developed a tagging language that he coined, AIML (A.I. markup language) as well as open sourcing his design. Because of this, Wallace’s AIML has become an accepted industry standard where conversational bots are concerned. In fact, the chatbot interface used by most second life residents to power their greeters and helpers are AIML-based Pandorabots. The bots however, still lack cognition.

Enter Mindpixel, the brainchild of Chris McKinstry, and Open Mind Common Sense, the brainchild of Marvin Minsky, Push Singh, and Catherine Havasi. Tragically, both McKinstry and Singh committed suicide in 2006. Still, their projects live on.

Importantly, both projects were based upon the premise of collecting and cataloguing all known factoids such that a sort of mega-knowledgebase could be achieved. The result would then become a source whence machine could find its intelligence. McKinstry’s mindpixel eventually found its home in Cornell University, and the paper, “Action dynamics reveal parallel competition in decision making,” by Rick Dale and Michael Spivey, with posthumous credit to McKinstry, was published in January of 2008. The OMCS project proceeds as well, and has amassed over fifteen thousand contributions. Yet, even if the contributions were increased by an order of magnitude, will that be enough to achieve mind? And what about body?

Enter Project Natal, mocap, and brain-computer interfaces.

To address this, three startup companies were eyeing electroencephalographs for controlling avatar movement, and the mocap industry is blooming as a mechanism for recording human movement. And just this last spring, Microsoft debuted Project Natal by introducing us to Milo. Milo is a comprehensive system that responds to voice, facial changes, and body movements. Other companies have moved toward touch screens. Imagine, if you will, a future where you can reach out and literally touch an avatar, and the resulting touch triggers the other player’s sensation gear? No. We’re not going “there” today! Seriously, the goal of immersion is to teach you to “bond” with your computer.

But back in the real world!

Regardless of the technology strides, we have yet to find soul. Enter quantum physics.

In 2002, the same year Kurzweil and Kapor were striking up their wager, two researchers, Johnjoe McFadden, a professor of biochemistry at the University of Surrey, and Susan Pockett, professor of physics at the University of Auckland, had put forth their own theories of consciousness. Dr. McFadden proposed an electromagnetic field theory of consciousness. More specifically, that consciousness lay in the brain’s naturally occurring electromagnetic field. He has since gone on to proffer that consciousness is a result of quantum evolution. Dr. Pockett challenged his proposal, while also noting in her discourse, “Difficulties with the Electromagnetic Field Theory Of Consciousness” (2002, Pockett):

My version of the electromagnetic field theory of consciousness is presently quite restricted in scope. It is that qualia are identical with certain yet-undefined spatio-temporal patterns in the electromagnetic field. At present, the only objects in the universe that generate such patterns are biological brains, but in principle there is no reason why the patterns should not be generated by artificial means.

I contacted both McFadden and Pockett, as I was toying around with my own theories and opined that their work might prove useful within the area of forensic psychology, specifically, utilizing mind printing, if you will, to detect factitious disorder. While Dr. McFadden simply pointed me to his website, the resulting discussion with Dr. Pockett was, needless to say, both informative and provocative. In a missive, Dr. Pockett noted, among other things:

With respect to your proposal, in principle you’re right, but in practice it’s simply not possible at this stage to measure what you want to measure, for technical reasons. I am presently trying to measure the em patterns that covary with the simplest possible kinds of sensory experience – taste, touch, that sort of thing – in normal subjects and even this is extraordinarily difficult. One major impediment is the inverse problem. This amounts to the fact that the em patterns are badly smeared by passage through the brain, skull and scalp before they get to the recording electrodes. You can work back mathematically to what the patterns would have been at their site of generation, but the problem is invariably “ill-posed” ie there is never a unique solution.

Fast forward to today.

Kurzweil has gone on to found Singularity University in preparation for uploading mind into computer. Wallace and various bot developers continue to attend Loebner’s prize competitions, vying for the grand prize. The startups, OCZ Technology and Emotive debuted their products in 2008 and this year, respectively. And Microsoft’s Project Natal is slated to debut in November of 2010. Dr. Pockett has recently completed research regarding conscious and unconscious states using intracranial EEGs. She is also exploring the question of whether the conscious mind is required in the decision making process. Dr. McFadden has since moved to the department molecular genetics and continues to flesh out his theory of quantum mind. And, of course, as mega dictionaries full of factiods (isn’t that what the google machine is?) continue to grow, and computer and cognitive scientist alike, are attempting to grok the results from a statistical analysis point of view.

All things considered, we can count ourselves blessed or damned, or perhaps, both. For as we stand at the door of tomorrow, witnessing the fluid fiction of technology, we are beginning to see a notable shift. One that redefines family, community, business and education. One that will also involve a change in the way we interact, not only with “computer,” but with each other as well.

Even so, will any of these “technological wonders” truly address the question of the ghost in the machine?