Quantum Physics Meets the Qubit
by Mark K. Anderson
2:00 a.m. Jan. 9, 2001 PST
AMSTERDAM -- Information, say the politicians and pundits, is the natural resource of the New Economy.
But hold that thought. To what kind of information do we refer? Is it the abstract concept of "0" and "1" -- entities, like god himself, that never bow to the physical world but just are?
Scientists are convening in Amsterdam this week to continue a quest begun two decades ago by Nobel laureate Richard Feynman, a quest to dispel this Platonic fantasy. The Fourth Workshop on Quantum Information Processing will be drawing scores of information researchers from around the globe.
These scientists and mathematicians investigate the manifold ways everything -- from computing to cryptography to the nature of information itself -- changes when quantum mechanics, that 20th century camel, sticks its nose into the tent.
Information, or at least the only kind of information computers can work with, is innately physical. It must be represented in some physical system -- say, the millions of on-off switches representing "1" and "0" in a computer chip. And it must be manipulated and processed within that physical system -- say, through the logic gates of a computer's microprocessor.
No sand and wires, no info.
The distinction may seem inconsequential, but then again so did the niggling that spawned Einstein's Theory of Relativity (minor discrepancies in the measured speed of light).
Perhaps because of the extreme pace of miniaturization and sheer number-crunching might over the past several decades, an invincible mindset has emerged that sees information as exempt from the physical universe around it.
"These [computers] -- the most complex things produced by the human mind -- can be made indefinitely small because of a crucial distinction," writes George Johnson in a Dec. 31 New York Times editorial. "While ordinary machines work by manipulating stuff, computers manipulate information, symbols which are essentially weightless.
"A bit of information, a one or a zero, can be indicated by a pencil mark in a checkbox, by a microscopic spot on a magnetic disk or by the briefest pulse of electricity or scintilla of light.
"The special nature of information confers another advantage. The power of computation can be leveraged and leveraged again. Design a computer and then use it to help you design a better computer, ad infinitum."
Well, sort of.
Charles Bennett of IBM's Thomas J. Watson Research Center studies the fundamental, physical nature of information and information processing and says that the time is long overdue for a re-think not only of the computer but also of all the 0s and 1s that flow through its veins.
And we can start with those two little numbers.
"When this abstraction was successfully made -- in the 1930s to the '50s -- the people who did it had left out the principles of quantum mechanics, which have to do with distinguishability," Bennett said.
The early computer innovators set up a framework that continues to this day. They thought of the bit as theoretically identical to zeros and ones written on long strips of paper. A zero can forever be zero; it will only change to one if the computer or computer operator makes it so; it cannot be both zero and one at the same time; everyone who wants to can see it; and observing it does nothing to alter its "zeroness."
An obviously productive idea, although it's also one that has little relevance to the elemental laws of physics that our ever-smaller and faster computer chips now brush up against.
"(John) von Neumann and (Claude) Shannon, all those people thought of information as classical, even though they knew better," Bennett said. "They were aware of quantum mechanics. And indeed, von Neumann had contributed fundamental things to quantum mechanics. But they thought it was somehow irrelevant to information processing, like it would just make the information less reliable. It would be a nuisance."
So "quantum information processing" is, in a sense, where all computers will be headed when the IT size and speed scales improve by several more orders of magnitude. It's time we get used to it: as far as the laws of nature are concerned, zero and one are quaint abstractions.
What comes closer to mirroring the physical universe is not the bit --robust, unambiguous and infinitely manipulatable -- but rather the quantumbit or "qubit."
Start computing with individual atoms, molecules, photons or nuclei, and it quickly becomes evident how fragile, ephemeral and intricately entangled the fundamental units of information can be.
To be discussed at QIP 2001 includes the kinds of programs that can be run on a computer composed of qubits; the kinds of ways qubits can be manipulated to do things no conventional computer can; the kinds of quantum leaps in security and communication that qubits bring; and the systems that
need to be devised to make it all happen.
No small task, and one that will undoubtedly open more hidden doors and passageways than the ones that are already known.
But at the root of it all is the qubit itself.
"What I say to people who don't like to think about many-dimensional spaces is that classical information is like the information in a book and quantum information is more like the information in a dream," Bennett said.
"If you have a dream
and somebody asks you about it, there's a certain privacy to it. Pretty
soon, you're remembering your explanations rather than what the dream originally
was. So in the course of making it public and making many copies of it,
the original content of it is altered in an
"Of course, dreams are terribly inexact. But there's an exact mathematics for handling this dreamlike information -- and it was known even in the '20s. But ... it wasn't thought of as something that had to do with the central notion of information. And that's what our insight over the last several years is. That it should have been included in our central notion of information, and that we are just beginning to discover all that can flow from that."