Rolf Landauer never thought his principle would solve the mysteries of quantum mechanics. He did expect, though, that information would play a part in making sense of quantum weirdness.
And sure, nobody thinks that all the mysteries surrounding quantum mechanics are solved now — and many wonder whether they ever will be, for that matter. But a new approach to one deep quantum mystery suggests that viewing the world in terms of information, and applying Landauer’s principle to it, does answer one question that many people believed to be unanswerable.
That question, posed in many forms, boils down to whether quantum math describes something inherent and real about the physical world. Some experts say yes; others believe quantum math is just about what people can find out about the word. Another way of posing the question is to ask whether the quantum description of nature is “ontic” or “epistemic” — about reality, or about knowledge of reality. Most attempts to articulate an interpretation of what quantum math really means (and there are lots of such interpretations) tend to favor either an ontic or epistemic point of view. But even some epistemic interpretations maintain that outcomes of a measurement are determined by some intrinsic property of the system being measured. Those are sometimes lumped with the ontic group as “Type I” interpretations. Some other interpretations (classified as Type II) believe quantum measurements deal with an observer’s knowledge or belief about an underlying reality, not some inherently fixed property.
Arguments about this issue have raged for decades. And you’d think they would continue to rage, as there would seem to be no possible way to determine which view is right. As long as all experiments come out the same way no matter which interpretation you prefer, it seems like the question is meaningless, or at least moot. But now an internationally diverse group of physicists alleges that there is in fact a way to ascertain which view is correct. If you’re a friend of reality — or otherwise in the Type I camp — you’re not going to like it.
There’s no way to decide the debate within the confines of quantum mechanics itself, Adán Cabello and collaborators write in a new paper, online at arXiv.org. But if you throw in thermodynamics — the physics of heat — then a bit of logical deduction and a simple thought experiment can clinch the case for Type II.
That experiment involves the manipulation of a quantum state, which is described by a mathematical expression called a wave function. A wave function can be used to compute the outcome of measurements on a particle, say a photon or electron. At the root of many quantum mysteries is the slight hitch that the wave function can only tell you the odds of getting different measurement results, not what the result of any specific measurement will be.
To dispense with some unnecessary technicalities, let’s just say you can prepare a particle in a quantum state corresponding to its spin pointing up. You can then measure the spin using a detector that can be oriented in either the up-down direction or left-right direction. Any measurement resets a quantum state; sometimes to a new state, but sometimes resetting it to the same state it was originally. So the net effect of each measurement is either to change the quantum state or leave it the same.
If you set this all up properly, the quantum state will change half the time — on average — if you repeat your measurement many times (randomly choosing which orientation to measure). It would be like flipping a coin and getting a random list of heads and tails. So if you kept a record of that chain of quantum measurements, you would write down a long list of 1s and 0s in random order, corresponding to whether the state changes or not.
If the quantum state is Type I — corresponding to an intrinsic reality that you’re trying to find out about — it must already contain the information that you record before you make your measurement. But suppose you keep on making measurements, ad infinitum. Unless this quantum system has an infinitely large memory, it can’t know from the outset the ultimate order of all those 0s and 1s.
“The system cannot have stored the values of the intrinsic properties for all possible sequences of measurements that the observer can perform,” write Cabello, of the University of Seville in Spain, and colleagues from China, Germany, Sweden and England. “This implies that the system has to generate new values and store them in its memory. For that reason, the system needs to erase part of the previously existing information.”
And erasing is where Landauer’s principle enters the picture. Landauer, during a long career at IBM, was a pioneer in exploring the physics of computing. He was particularly interested in understanding the ultimate physical limits of computational efficiency, much in the way that 19th century physicists had investigated the principles regulating the efficiency of steam engines. Any computational process, Landauer showed, could be conducted without using up energy if performed carefully and slowly enough. (Or at least there was no lower limit to how much energy you needed.) But erasing a bit of information, Landauer demonstrated in a 1961 paper, always required some minimum amount of energy, thereby dissipating waste heat into the environment.
A Type I quantum state, Cabello and colleagues argue, needs to erase old information to make room for the new, and therefore a long run of measurements should generate a lot of heat. The longer the list, the more heat is generated, leading to an infinite release of heat for an infinitely long list, the researchers calculated. It’s pretty hard to imagine how a finite quantum system could generate an infinite amount of heat.
On the other hand, if your measurements are creating the list on the fly, then the quantum state is merely about your knowledge — and there’s no heat problem.
If the quantum state is Type II, it “does not correspond to any intrinsic property of the observed system,” Cabello and coauthors note. “Here, the quantum state corresponds to the knowledge or expectations an external observer has. Therefore, the measurement does not cause heat emission from the observed system.”
Fans of Type I interpretations could argue that somehow the quantum system knows in advance what measurement you will perform — in other words, you really can’t orient your detector randomly. That would imply that your behavior and the quantum system are both governed by some larger system observing superdeterministic laws that nobody knows anything about. Bizarre as that sounds, it would still probably be a better defense than attacking Landauer’s principle.
“Landauer’s principle has been verified in actual experiments and is considered valid in the quantum domain,” Cabello and coauthors point out. “Therefore, whenever the temperature is not zero … the system should dissipate, at least, an amount of heat proportional to the information erased.”
If you would rather not take their word for it, you should check out the September issue of Physics Today, in which Eric Lutz and Sergio Ciliberto explain the intimate links between Landauer’s principle, information and the second law of thermodynamics.
“Having only recently become an experimental science,” Lutz and Ciliberto write, “the thermodynamics of information has potential to deliver new insights in physics, chemistry and biology.” The new paper by Cabello and colleagues appears to be an example of just such an insight.
Nobody should expect this paper to end the quantum interpretation debate, of course. But it surely provides a new point of view for discussing it.
“Ultimately, our work indicates that the long-standing question, Do the outcomes of experiments on quantum systems correspond to intrinsic properties? is not purely metaphysical,” Cabello and colleagues write. “Its answer in the affirmative has considerable physical consequences, testable through experimental observation. Its falsification will be equally exciting as it will force us to embrace radically new lines of thought.”
Follow me on Twitter: @tom_siegfried