Tuesday, September 25, 2007

The constant smirking of dolphins should have tipped us off

The following is an excerpt from an article in Discover magazine.

"Their constant smirking should have tipped us off, but we were so damned in love, dazzled by the glare from those latex-smooth exteriors, charmed by those adorably wobbly tail stands at Sea World. [...] Now, however, if a team of scottish marine biologists are to be believed - and if you've ever dipped even a toe into a body of scottish water you have some sense of those people's rigor - dolphins are not to be trusted. They gossip. We know this because we know their names. Each bottlenose individual identifies itself by a unique pattern of clicks along the lines of woo-woo-wee-wee, or even woo-wee-woo-woo-wee-woo. What was not known until the Scottish research, however, is that a pair of dolphins use the name of a third dolphin when that third dolphin isn't present. In other words, dolphins gossip."

I just had to post this one.

Tuesday, July 10, 2007

Extropy: Pantheism

While enjoying an espresso and a book on cognitive science at the lovely Gutenberg Cafe, I had a provocative idea I would now like to share with you.

Is AI an emergent property of any tech-species? To put it another way, is AI inevitable given a species which can manipulate parts of the physical world, i.e. technology?

I believe that it is. It all stems from the properties and products of recursive processes. This also has to do with autopoiesis. Once the tiniest most simple instance occurs, an explosion of intelligence has to happen, barring any apocalyptic event.

I see it proceeding as follows: life stems from nature and the ingredients which comprise this planet, then human level intelligence stems from life and the ingredients which comprise organic chemistry, then artificial intelligence stems from human intelligence and the ingredients of metal and silicon.

This progression should be obvious, but is it inevitable?

Now I am really taking the concept of autopoiesis and running very far with it but it was originally introduced as a way of describing self-sustaining life and any other dissipative structure. I saw the convenience of this definition simply as life supporting life. Instinctively keeping one's species in existence and all that Darwinian hullabaloo. The original 'motivation' of life has always been to keep itself alive. This is how evolution came about because with a changing environment some separated areas of life learned how to keep themselves alive better than the rest. (I now slightly digress to ask for some pardon regarding my anthropomorphizing the chemical processes of DNA and mutation. I am doing so to help make it apparent that we are experiencing the same force of nature but seems different because we are sentient and natural selection is not.) As I was saying, life became so good at keeping itself alive it became able to manipulate the environment so that it would be easier to stay alive. This of course refers to us.

Now is where it becomes interesting. Humans have created artificial intelligence and are currently working to make better ones. Why? Humans do this because AI tends to be very good at making living easier for the humans. Humans have invariably continued the process of life keeping itself alive. Not only that, humans have made it much faster. Some say exponentially so.

Now I know you are going to ask how AI has made it easier for you to stay alive. I would respond that what it means to live is much more sophisticated to a human than it is to any other form of life on this planet. We have reached a level where if one had access to any large compendium of information as is normal in any developed nation, one would be able to leave the hustle and bustle of the city and "live off of the land" without any problem. Some have even tried this. I make this point to make it obvious that we have reached a ceiling with regards to evolution through natural selection. Natural selection will not be able to make it easier for our species to survive. We have created our own form of evolution. The evolution of ideas, principles, and information. Our collective minds have made it so easy for our bodies to survive that it has turned on itself and raised its own expectations. We seem to have to keep moving, which may be more evidence supporting the embodied cognition school of thought.

I thoroughly enjoy the concept of extropy, as informal as it may be. It is such a beautiful view of life. The opposite of entropy. While entropy increases disorder in the universe,extropy increases order in a particular neighbourhood. While entropy starts out as a rapid change and slows down as it dissipates, extropy starts out as the slow evolution of life and speeds up as it progresses.

The biggest critique of entropy is that it seemingly tries to nullify the second law of thermodynamics. This, however, is not correct. The second law is only interested in very large systems, i.e. the universe. Extropy, or rather the description I have here, deals primarily with increasingly smaller systems. The human civilization may expand up to and even beyond the solar system, but intelligence is not defined by how much space its proponents take. It is defined by ability and amount of organization. A system can increase in extropy without increasing in size. Consider the Big Bang as an example of entropy along with the fact that the universe is still expanding; while considering the pulling together of mass to form planets as extropy.

I share all of this because it has portrayed to me the world and life in general in such a sublimely beautiful way only a pantheist could experience it. The advancement of intelligence and order in general is the quintessential anthropic concept. It preceded us because we are products of this propensity for autopoiesis to develop self-aware systems. This thing called life has not been given to us by some higher power. We are not children in a playground given to us by some omniscient parent. We were born from non-conscious star dust. To quote Carl Sagan we are: "star stuff contemplating star stuff". The secrets of life are not being governed and hidden from human eyes. They are here and we are immersed. Maybe the future is not variable and can be predetermined, but if it is we are the only ones responsible.

Saturday, June 30, 2007

The Emotion Machine

While I am still reading Godel, Escher, Bach: An Eternal Golden Braid I have interluded said reading with the reading of a similarly themed book titled: The Emotion Machine. It was just recently written by one Marvin Minsky.

Minsky probes deeper into the question of natural intelligence. Don't look for simple explanations: he believes "we need to find more complicated ways to explain our most familiar mental events"; we need to break our thought processes down into the most precise steps possible. In fact, in order to truly understand the human mind, Minsky suggests, we'll probably need to reverse-engineer a machine that can replicate those functions so we can study it.

I love this book the more I read it. It may be dangerous to post about it before I have completed the book, but it is not of the type to finish in one sitting.

What I love most about reading this book is that I invariably formulate theories to describe or solve the problems and situations he presents in the book. He structures the reading such that after many of the problems he introduces he simply states that they need to be solved because they haven't been so far. I like this as it motivates me.

Upon writing this I have realized that I seem to be attributing this method of presenting ideas to Minsky himself when in fact they are due to the fact that the time this book was written no such solutions are in place. Still, the book has helped me to recognize that I have just been guilty of self-reflection. That recognition is of course a type of recursive self-reflection, a self-reflection-reflection if you will. Who or what should I attribute to the fact that I just recognized a SRR (Self-Reflection-Reflection)or even that I recognized a SRRR which leads to recognition of a SRRRR ad infinitum. Perhaps there is some truth to the infinite regress of the homunculus, more on that in a later post.

I will leave you with an excerpt from a review which can be found on Amazon's site for the book. I warn you, it could be considered controversial for anyone with a sense of self, i.e. everyone, and any Stoics out there.

Thus, [Minsky] rejects the idea of consciousness as a unitary "Self" in favor of "a decentralized cloud" of more than 20 distinct mental processes. In this view, emotional states like love and shame are not the opposite of rational cogitation; both, Minsky says, are ways of thinking.

Saturday, May 5, 2007

Do You Know What Artifical Intelligence Means?

A very pertinent article written by Eliezer S. Yudkowsky who is a research fellow for the audacious Singularity Institute can be found here.

The article is very articulate. I strongly suggest you read it if you plan on living for atleast thirty more years.