Wednesday, January 10, 2018

A Refutation of Heat Death (On Entropy and Consciousness)

"The distinction between materialism and spirituality is an artificial one; the dichotomy which sets religion against science is a false one. The two worlds are one; the heavens and the earth are entangled and intertwined."
The second law of thermodynamics states that the entropy of a closed system tends to increase. Within this lies the possibility of two subsystems, one of decreasing entropy, and one of increasing entropy. Intelligence functions to distinguish between the two systems, playing the role of Maxwell's Demon--sorting the larger system into two subsystems, one of decreased entropy and one of increased entropy. Biological intelligence has evolved to distinguish between the two systems via the system of perception in order to decrease the entropy of the environment of the individual and its offspring in order to increase the possibility of the propagation of the individual's genes.

However, a gene is only one sort of abstract pattern which can be protected and vied for through evolutionary processes. Any abstract pattern is subject to natural laws of evolution, reproduction, and propagation. Concepts, in and of themselves, are patterns which are subject to these natural laws. Self-replication typically increases the fertility or fecundity of a pattern. If a pattern is found to possess both fertility (the tendency to self-replicate) and durability (a long half-life), it will likely enjoy a high prevalence (the degree to which it is widespread, or common, within the system.)

Noting that entropy is a measurable quantity, it follows that for the entropy of one subsystem (Subsystem A) to decrease while the entropy of the entire system increases, the entropy of the rest of the system (Subsystem B) must increase by a magnitude which is at least as great as the magnitude of the decrease of Subsystem A.

While the universe as a whole must experience an increase in entropy, it is possible for some section of it to experience a decrease in entropy. It is primarily by the organizing capacity inherent in intelligence and consciousness that this can occur. Intelligent life is therefore that pattern which drives the separation of the whole system (universe) into the two final subsystems, which might be characterized as Chaotic (or "Disordered") and Assembled (or "Ordered.")

This seems to provide an answer to the supposed heat-death of the universe: While the entropy of the system taken as a whole must continue to increase, subsystems may be created by intelligent agents within which entropy can continue to decrease indefinitely.

Wednesday, July 12, 2017

Psychiatry for Artificial Minds

Artificial neural networks are miniature manifestations of the mind, and as such, can be expected to display similar phenomena to the biological mind, to the extent to which they are modeled after the biological mind. If, as suggested by the diffusion-based neuromodulation paper, we make our model more sophisticated -- make it resemble the biological mind with greater fidelity -- we can expect the model to produce behaviors which more closely resemble those exhibited by the biological mind.


So perhaps neural networks will need psychiatrists someday.

And perhaps the engineers that work on neural networks will be able to bring insights from experimentation performed in the domain of artificial intelligence, where things are more easily quantified and the space less ethically-fraught, and apply those insights to the field of human psychology.

If we continue to model the artificial mind after the human one, we can expect to witness the presentation of disorders -- though they may at first be thought of as "bugs" -- similar to the disorders that arise in human minds today: Autism, seizures, bipolar, Alzheimer's, etc. Some of these are apparently disorders of architecture (i.e. autism and other life-long disorders which are effectively untreatable.) Others are disorders of neurochemical modulation (i.e. seizures, as evidenced by their presentation upon alcohol withdrawal; bipolar disorder, etc.) Others are disorders of maintenance, of hardware degradation (i.e. Alzheimer's, dementia.)


Architecture refers to the number of neurons present and which are connected to which.

Neurochemical modulation refers to the way that neurochemicals (or their artificial counterparts, in a computer model) affect the function and behavior of neurons and the systems which they compose.

Maintenance and degradation refer to the upkeep of neural systems over time. To the extent that our inorganic components are subject to degradation and pollution (i.e. diffusion of impurities in circuit elements,) we will see maintenance disorders manifest in inorganic neural systems.

Of course, there will always be fundamental differences between organic and inorganic minds, and therefore we can expect to be presented with some entirely new problems, too. One such problem is the one of catastrophic forgetting. Biological brains retain knowledge and skills quite well in the face of new information, compared to artificial neural networks. ANNs have a problem of forgetting old skills quickly. In other words, if an ANN is trained with additional data, it is often then observed to perform poorly on a task at which it previously performed well.

Coming to understand the nature of human mental disorders may be the key to debugging our artificial intelligences, and vice versa. Our first attempts will be clumsy, and we will likely see many iterations of artificial minds which are deficient in one way or another -- one might have a good eidetic memory but be terrible at adapting to new situations; while another might very dynamic, but not very resilient in its "core beliefs."

Indeed, we will likely come to a deeper understanding of how architecture gives rise to consciousness and behavior as we create minds from schematics and observe their behavior. What architectures are capable of creating good music? Eloquent speech? Visual problem-solving? What about mathematical analysis? And motor control?

Perhaps we will find that the mind is more self-organizing than we ever suspected, and that our so-called knowledge of the structures of the human brain and their respective functions is all a fallacy, that these functions arise simply as a matter of physical proximity to the inputs and outputs of the respective sense-organs and musculatures.

But, perhaps not. Perhaps we will find that there is a strict genetic code that produces the particular circuits within the brain which make possible certain behaviors and yield certain propensities -- a talent for music, for speech, for mathematics, for spatiotemporal performance (i.e. athletics,) for competition, for theory-of-mind. Perhaps we will find that certain neuromodulations correspond to certain characteristics of personality. Does norepinephrine calibrate curiosity? Does prolactin calibrate extraversion? Does GABA calibrate conscientiousness?

Tuesday, April 4, 2017

Why the Mind?

Hawking. Einstein. Curie. Darwin. Freud. Aristotle. Da Vinci. Christ. Buddha. Confucius. Gandhi. Mother Theresa. Martin Luther King, Jr. What do these historical greats have in common?

They earned their place in the history books without armies or advertising budgets. Their ideas, their philosophies, inspired their followers and set them apart from the pack. Their concepts changed the world forever.

Ideas, philosophies, and concepts -- all products of the mind.

What is the mind? Why do we care about it?

The mind can make history, but more than that, the mind is the essential difference between a man and an ape. Gorillas, monkeys, and orangutans experience roughly the same emotions as we do: Love. Fear. Happiness. Anger. Pleasure. Jealousy. Hope. Despair.

The mind gives rise to language. It gives rise to structure, to organization, to rules and regulations, to empathy and enmity. The mind conceives of gardens, rockets, guns, towers, tanks, music, sculpture, scripture, treaties, movies -- all of human culture, past and present.

But does that make it worth studying? Does that make the mind worth our time and consideration?

Why is the study of the mind important, here and now?

In the past century, we have peered deeper and deeper into the universe's infinitesimal depths, gaining an increasingly intricate understanding of chemistry, physics, biology, and data. Not only have we accumulated new knowledge, we've also developed new ways to organize, access, and process that knowledge.

Modern computing technology is allowing us to create virtual minds that outperform humans at increasingly complex tasks. These artificial intelligences can now outperform our simian ones at image recognition, board games (first Chess, and now Go), and other tasks. Within ten years, we hope to have cars that drive themselves, freeing up billions of man-hours to apply to other tasks.

Meanwhile, genetics and chemistry are giving us unprecedented insight into the workings of our human brains. We know the molecular structures that give psychoactive substances the ability to alter behavior, and we are using this knowledge every year to conceive, manufacture, and sell "new and improved" variations of these structures, in the form of prescription drugs. New molecules. New mixtures, and old mixtures in different proportions. New delivery systems. We regularly identify and mimic the neurotransmitters that interact with our brain cells, and increasingly, we understand what's happening both inside and outside those brain cells when these chemicals enter our bloodstream. Increasingly, we're able to predict the behavioral outcomes of these chemical changes, and the practice of prescribing pills to fix problems is now a societally-ingrained approach.

As a result of the simultaneous, parallel advances in these different fields of science, we stand at a crux. We have the potential now to understand the mind like never before in history.

Some of the greatest minds of the last century have chosen to concern themselves primarily with the study of the mind itself.

Ray Kurzweil, for example, dubbed by Forbes the "ultimate thinking machine," publishes books with titles like: "The Age of Intelligent Machines", "The Age of Spiritual Machines", "The Singularity is Near", and "How to Create a Mind," all centered around the theme of engineering artificial intelligences by understanding and mimicking human neurobiology (and featuring Kurzweil's predictions about the results of such an undertaking.)

Or take Freud, whose opinions, controversial as they have been, have undoubtedly served as a foundation and a catalyst for much of modern psychology and the medicine of the mind.

Apes have tongues and fingers, the dextrous organs that make possible both sophisticated communication and use of tools. (Even if horses had minds like humans', there would be no "horse technology", for they lack the physical organs which make technological innovation possible.) But in contrast to the primitive ape, the human mind has given rise to music, machine guns, and more. Wars are begun and ended by the products of the mind: Hitler's ideology, Einstein's bomb. Without the mind, we are but fleshy shells, vegetative, unfamiliar. And it is only due to the advanced, sophisticated mind which is unique to us among all primates, that we live as we do today, as part of a vast, contrived civilization, far separate from the order of the natural world.

The mind presents many faces for us to study, many rabbit holes to go down. We are fascinated by it because it is both enigmatic and fundamental. The mind gives rise to the entire human experience. How then could we not investigate it to the utmost?