January 29, 2009

Quantum Mechanics: Modern Alchemy?

Quantum mechanics (QM) has a serious jargon addiction. On the theory that if thousands of university students can get degrees in QM, then I should have no trouble understanding the basics of what they study, I've been trying to decipher some of the jargon from this infamously confusing field. Here's what I've come up with so far:

When a quantum physicist says a particle is 70% one thing and 30% some other thing, he means that if you were to measure it, there would be a 70% chance of the measurement coming out one way and a 30% chance of it coming out the other. As far as I can tell, this is identical to saying "It has a 70% probability of being in state X and a 30% probability of being in state Y, but I don't know which state it's actually in." However, standard QM jargon has you say "the particle is both 70% X and 30% Y at the same time", and treats this situation as a state in its own right. This makes some sense at the mathematical level, where any particle state can be replaced by a function (called a wavefunction) which gives the probabilities of the particle being in each state.

This leads directly to the jargon about measurements causing "a collapse of the wavefunction", and the idea that an observer is necessary to put anything in a definite state. If you think of a particle as being in one state or the other, then a measurement tells you which state it's actually in, so the wavefunction becomes unnecessary. QM jargon is the only thing leading people to say (and believe) that measurements cause changes in the states of particles rather than changes in the state of our knowledge about the state of particles.

I want to pause here and say that I don't think quantum physicists are wrong. I think their predictions about what will happen when measurements are taken are perfectly accurate. What I'm concerned about is the failure to communicate how such predictions are made to interested people not trained in QM. I'm especially concerned about common mistranslations of QM jargon which lead to claims that QM has discovered faster-than-light communication, matter teleportation, and ways to compute answers to non-polynomial-time problems in polynomial time. Claims like these often lead to more funding for research into QM, which is one reason they continue to appear over and over again. However, although QM is worthy of funding, these particular justifications are no more valid than claiming that a good reason to study alchemy was to convert lead into gold. There were good reasons to study alchemy (which evolved into chemistry) but that wasn't one of them.

The claim that QM will lead to a fundamentally faster form of computation comes from the idea that a wavefunction encodes multiple states at the same time. Superficially this sounds plausible, since if you perform an operation on a wavefunction, the operation is performed on all the states the wavefunction encodes simultaneously. Unfortunately, there is no way to extract more than one result of an operation from a wavefunction. Once you take one measurement, the wavefunction collapses, and taking any further measurements will only reveal the same result. Again, this makes perfect sense when you remember that a wavefunction isn't a state, but a way of encoding our imperfect knowledge of the state. To make a prediction about what a measurement will reveal, we have to compute the result of an operation on all the possible states, each of which will lead to a different result. But when you measure an actual particle, it will only have one state. The "multiple states" are only being computed by the people making predictions.

QM actually does promise vast improvements in computing power, but only because using individual particles to store information would be so damned efficient.

Faster-than-light communication is another perennial favourite claim. Typically "entangled particles" will be invoked, with the claim that altering one particle will instantaneously alter the entangled particle, no matter how distant it is. In practice, when two particles are "entangled", it means they interacted sometime in the past in such a way that their states now depend on each other. For example, if two photons of opposite polarity are emitted in random directions (as from a beam splitter), then the two photons are entangled. The wavefunctions of the photons show that they could have one polarity or the other, but we don't know which one is which. The "entanglement" means that when we measure the polarity of one photon, we will instantly be able to deduce the polarity of the other. In other words, when we take one measurement, the wavefunctions of both photons collapse. It is the collapse of the wavefunction of the photon distant from the measurement which is often interpreted as "altering one particle [collapsing its wavefunction] instantaneously alters the other." Clearly, however, no information is transmitted. Entanglement merely means that if we can learn the state of one particle, then we can deduce the state of another. There is no way to actually change the state of a distant particle.

Yet another piece of QM jargon refers to the collapse of entanglement as "quantum teleportation". In popular usage, "teleportation" means moving matter from point A to point B without passing through the intervening locations, usually instantaneously. "Quantum teleportation" on the other hand, is used to mean that information is transmitted from point A to point B instantaneously, which is quite different (and just as incorrect). Promises that QM will lead to Star-Trek-style transporters become ludicrous once you realize that not only does quantum teleportation not involve changes in the locations of particles, it doesn't involve changes in the location of information either (see the paragraph on faster-than-light communication above).

On the other hand, QM does herald an era where we can manipulate fundamental particles with relative ease, to the point that we could, in theory, build any complex object out of basic materials. In particular, if it were possible to measure the state of an object in sufficient detail (which may not, in fact, be possible), we could transmit the information to another location (no faster than light-speed), where a replica could be built. This is about as close as we could come to teleportation without invoking wormholes.

A special message to quantum physicists: I know most of you are perfectly aware of what is and isn't possible in the realm of quantum mechanics. I've certainly read enough clear disclaimers in the conclusions sections of your papers. However, when you try to explain to journalists (or your university's PR people) the exciting things you're doing, using your jargon to get the journalist exited about the wrong thing is bad. As scientists, whose goal is to seek truth, you should be ashamed when you or your colleagues allow the public to be so completely mislead about the significance of what you do.