George Johnson
Fire in the Mind
Viking 1996

George Johnson
Fire in the Mind
Viking 1996

pg 110


Keywords: mass, energy, space, time - INFORMATION - Most of us are used to thinking of information as secondary, not fundamental, something that is made from matter and energy. Information seems like an artefact, a human invention. We impose pattern on matter and energy and use it to signal our fellow humans. Though information is used to describe the universe, it is not commonly thought of as being part of the universe itself -
Information can be thought of as a measurement of distinctions, the simplest being 1 or 0, the presence or absence of a certain quality. By this measure, there is more information in something that is orderly than in a homogeneous, undifferentiated mess. - quantum theory - probability wave - measurement problem - Superpositions and Quantum Decoherence - many worlds











COMPLEXITY, ENTROPY, AND THE PHYSICS OF INFORMATION

Wojciech H. Zurek

In building a tower of abstraction, one must start with a foundation, those things that are taken as given: mass, energy, space, time. Everything else can then be defined in terms of these fundamentals. But gradually over the last half-century some scientists had come to believe that another basic ingredient was necessary: information.

"The spectre of information is haunting the sciences", Zurek's manifesto began. There is a "border territory", he believed, where information, physics, complexity, quantum theory and computation meet. What Zurek and his colleagues had in mind was a return to basics, a rethinking of reality's pillars as thorough as any undertaken by Thales, who thought all is made of water, or Heraclitus, who thought all is made of fire.

Most of us are used to thinking of information as secondary, not fundamental, something that is made from matter and energy. Information seems like an artefact, a human invention. We impose pattern on matter and energy and use it to signal our fellow humans. Though information is used to describe the universe, it is not commonly thought of as being part of the universe itself.

But to many scientists the world just didn't make sense unless information is admitted into the pantheon, on an equal footing with mass and energy. A few went so far as to argue that information may be the most fundamental of all, that mass and energy could somehow be derived from information.

There was, first of all, the mysterious connection that seemed to exist between information, energy, and entropy, the amount of disorder in a system. We learn in school that, left on its own, any closed system becomes more and more disorderly, its entropy increases. It is because of this fact, embodied in the second law of thermodynamics, that neat geological strata become gnarled into formless Precambrian rock. The planar geometry of an adobe village melts until it is barely distinguishable from the surrounding hills. Along the way, pattern is washed away; information is lost.

Information can be thought of as a measurement of distinctions, the simplest being 1 or 0, the presence or absence of a certain quality. By this measure, there is more information in something that is orderly than in a homogeneous, undifferentiated mess.

On the other hand, by gathering and processing information, we can create order -we can take the matter and energy of our world and arrange it into songs, civilisations, fragile eddies in the entropic tide. Using our powers as information processors, we can find unlikely structures that already exist - water trapped in a mountain lake, carbon molecules strung in a volatile chain, protons and neurons stacked into a precarious nuclear sphere. And then we simply let them follow the path of least resistance. As they topple and move down the hill from order to disorder, we can extract work by harnessing the entropic flow. The nucleus disintegrates, the bonds of the carbon atoms break, or water flows from its pool to the formless sea. Entropy increases, information is lost, but the energy released in the process can be tapped to build new structures, to create information, though all our creations must eventually succumb to the second law.

No wonder the mind craves patterns. It is the ability to find order in the world that allows us to make use of its resources. For many scientists this would be reason enough to believe that information is fundamental.

But, going beyond the laws of thermodynamics, some believe information plays an even deeper role. According to some interpretations of quantum theory put together by Zurek and his circle, without information there would be no resources to exploit and no one to exploit them; there would be nothing that resembled what we call the real world.

The mathematics used to describe the subatomic realm tells us that, left to its own devices, an electron lacks the very attributes that we, on our microscopic plateau, consider the very hallmark of existence -a definite position in time and space. It exists, we are told, as a probability wave, a superposition of all the possible trajectories that takes on substance only when it is measured, when, as it is often put, an observer collapses the probability wave.

How this transformation occurs is one of the deepest mysteries of physics, the so-called measurement problem: How does the rock solid classical world, in which things occupy definite positions in space and time, crystallise from the quantum haze?

In the past, quantum theory has often been embraced by those who would elevate subjectivity over objectivity, championing a mystical world view in which consciousness brings the universe into being.

By making information fundamental, Zurek and some his colleagues hoped to demystify quantum theory. For what is an observation but the gathering of information? And if information is fundamental, it exists as surely as does matter and energy, without the need of conscious beings. The quantum wave might collapse not because it was beheld by mind but simply because information flowed from one place to another in the subatomic realm. Of course, it is easy to be fooled by our own metaphors, becoming so dazzled by the concepts we invent that we can see the world only through their glare.

In the 19th century, entropy and the laws of thermodynamics were invented to deepen our understanding of the steam engine and make it as efficient as nature would allow. Any closed system, sealed off from its environment, would inevitably march from order to disorder. Soon scientists and philosophers were applying these new mental tools to the universe itself, declaring that, as the most closed of close systems - what could possibly be outside of it? - the universe was marching inevitably toward thermodynamic death, a state of equilibrium, lifeless, unstructured, random.

In the 20th century, information theory was invented to help engineers make electronic communications channels as efficient as possible. And before we knew it, people were speaking of information as real, a few going so far as to imagine that we live in a universe of computation, created from the shuffling of bits.

One of the challenges implicit in Zurek's manifesto was to find new ways to think about whether computation - and therefore information - is natural or artificial. The computers we have built over the years have been crafted from microscopic parts: first gears, then vacuum tubes, then transistors, and now chips inscribed with thousands of transistors that get smaller and more densely packed every year. We stamp our designs on nature's designs; circuitry onto a silicon lattices. But the finer of the blueprints of our artifices, the more they begin to clash with the physics underneath.

Quantum randomness scrambles out to neat choreographies of 1s and 0s. But perhaps as engineers reach tinier and tinier scales they can somehow exploit the natural behaviour of atoms to make their machines more efficient, bridging the divide between the circuitry we design and the circuitry of nature.

And atoms with an electron that could be in one or two states might naturally be thought of as a register containing a 1 or a 0. How thin can be make this gap between the laws of computation and the laws of physics? Where will the shrinking bottom out?

If computation can take place only down to a certain scale, requiring components made up of many, many molecules, then perhaps information is simply an arifice, secondary to laws of physics, a pattern imposed by people as they struggle to describe the world. But if single molecules or even atoms can be said to somehow process information, then maybe computation is as fundamental as what we think of as the laws of physics. Like mass and energy, information would be irreducible, at the roots of creation.

pg 161

Superpositions and Quantum Decoherence

In trying to make sense of quantum theory, some have been led to believe that consciousness, or at least the act of measurement, is necessary to bring about what we consider the real world. But many scientists are suspicious of what sometimes seems like a self-centred attempt to elevate humanity and the classical world we experience to a special, almost God-given role.

If we follow the approach of some of the people at Santa Fe and Los Alamos and admit information as another fundamental, along with mass and energy, then quantum theory can be viewed in a subtly different light. All that is required to break the symmetry of the wave function is information processing.

Not only are conscious observer superfluous - the theory does not even require artificial observers like photographic emulsions or photoelectric cells. The universe itself might process information just as it processes matter and energy.

Seen in this light, our role as informational spiders, stringing and restringing our conceptual webs, is as natural as anything in the cosmos. We try to set ourselves apart from the universe and pretend to see it whole. But we are inevitably a part of what we are observing, and our observations may be but a single circuit in a great web of flowing bits.

At Los Alamos, Zurek and some of his colleagues have been examining the difference between classical and quantum measurements in an attempt to better understand how we come to know the world.

The trick, they say, is to follow the information. Where does it go when we make a quantum measurement? In addition to its static attributes - mass, spin, that which makes an electron, a photon, or whatever - at quantum particle carries this huge complex of dynamic information: the wave function describing every possible state, and every possible combination of states that it might assume. When it is measured and takes on one of these states, to the exclusion of all others, what happens to the extra information? Does it dissipate into the environment in an irreversible act of erasure?..... Where does the information go? It is dissipated into the environment.

Apparently we don't need a measurer or even an inanimate detector to cause decoherence. It seems that the environment itself can absorb the excess information and cause the possible outcomes to come unglued.

And there lies the beauty of this interpretation: there is no reason to give special status to an observer or to sanctify the measurement act. Anything that can absorb information can be thought of as making a measurement. The collapse of the wave function can be shifted from the observer and placed on the environment itself.

"The essence is that the environment knows - it has a record", Zurek said. ..."It is as if the watchful eye of the environment monitoring the state of the quantum system forced it to behave ineffectively classical manner".

Like the inevitable fall from order to entropy, this loss of quantum coherence is irreversible. According to the second law of thermodynamics, the entropy arrow points one way because, as the disorder of universe increases, information is dissipated beyond retrieval. If we knew the speed and direction in which each shard of the shattered pot was flying, we could theoretically reverse their many courses and cause them to reembrace. This lost information is out there somewhere: in the vibrations of air molecules disturbed by the moving pieces. But these air molecules disturb other air molecules and information scatters.

Likewise, the quantum-to-classical transition cannot be undone, because the extra information in the wave function is irritrievably scattered throughout the environment.

The air molecules in the atmosphere, photons streaming from the stars - even in deep, dark space, the cosmic background radiation is there, interacting with quantum particles, causing decoherence, spiriting away the excess information at the speed of light.

The environment is monitoring everything all the time, collapsing wave functions, bringing hard edged classicality out of quantum mushiness. Like a city saturated with sounds, the universe is saturated with information from these constant measurements.

We are used to thinking of the difference between quantum and classical as the difference between the very small and a very large. Zurek sees it instead as the difference between a system that is closed and the system that is open.

A completely closed system - one that is interacting with nothing - is represented by the pristine symmetry of the wave function, with every possible state and every possible combination of states in superposition, interfering with one another.

In its unmeasured state, a quantum die would read six and one at the same time - and two and five, and three and four and five and six, and so on. But these arbitrary superpositions are extremely unstable; they can survive only in a closed system. When we open up the system - forcing it to interact with the environment - they dissipate so rapidly that they cannot be retrieved. Only a small fraction of the possible outcomes that coexist in the wave function are stable enough to endure on their own - the ones that behave classically, in which something can be in only one state at the time.

There are still many possible outcomes left for us to contend with , but these are now probabilities that are knowable. So we go from quantum ignorance to classical ignorance, from the inherent randomness to randomness rooted in what we don't yet know. The interaction with the environment acts like sieve: it sifts the wave function, leaving a set of classical probabilities. Then we resolve our classical ignorance by measuring and gaining information.

The environment can be thought of as any system large enough to absorb the excess information and bring about decoherence. In this democracy of measurement, we cannot really say which is the observer and which is the observed. The photon leaves an imprint on our retina; our retina leaves an imprint on the photon. Measurement is simply the correlation of two systems, which go away from the encounter with a record of each other.

The result of all this is what we might call the classical illusion. Once decoherence occurs, all the pieces of the wave - even the weird juxtapositions - are echoing through the environment somewhere. But all we can experience is the classical outcomes, where something is in one place at the time. For it is only these classical states that are stable, long-lasting, and predictable.

We can measure them without fundamentally disturbing them, recording the information in our brains. There is no reason for our senses to perceive the superposed states, which dissipate so rapidly. They carry no meaning for us, they are indivisible to our information processing.

"Our senses did not evolve for the purpose of verifying quantum mechanics", Zurek has written. "Rather, they developed through a process in which survival of the fittest played a central role. And when nothing can be gained from prediction, there is no evolutionary reason for perception".

It is only when we can make stable records - memories - of something that it can be said to exist. Two things must leave imprints on each other to be mutually real - i.e., in the same universe. Decoherence makes possible information exchange, and it is only through information that we can know the world.

All of this goes along way toward painting a picture in which classical probabilities arise from quantum uncertainty, without the need for observers. We are dealt this hand of premeasured, decohered states, but why do we experience just one of them?

In the terminology of Everett and the many worlds interpretation, we make a measurement and the universe splits into branches: one in which the molecule is on side A, one in in which it is on side B; six branches for each role of the die. But nothing in Everett explains why we are stuck on just one of the branches.

Zurek believes decoherence may hint at an answer. We have to remember that in this democracy of information exchange our brains are also being measured by the environment. As we observe the die, neurons are momentarily placed in superposition between all possible juxtapositions. But our brains are not closed systems. Networks of neurons send signals to one another. And so the superposed brains states decohere long before we are aware of them. The nonclassical states are instantly spirited away.

The brain will still be left with a variety of premeasured classical states,Everett branches. But here as the crux of the argument: there will be only one branch in which the die says six and our brain says six. That is the one that we perceive his real.

"Decoherence is preventing our brain from getting into these funny superpositions",Zurek said. "There is only one option in which I see the cup here and it is here. So we are in one state of mind at any one time and that state of mind is correlated with one state of the universe. "It is sort of the reverse of the observer creating reality. The universe is through our senses ajusts the record in our brain".



 

HOME      BOE     SAL     TEXTE