We discuss some seemingly paradoxical yet valid effects of quantum physics in information processing. Firstly, we argue that the act of ``doing nothing'' on part of an entangled quantum system is a highly non-trivial operation and that it is the essential ingredient underlying the computational speedup in the known quantum algorithms. Secondly, we show that the watched pot effect of quantum measurement theory gives the following novel computational possibility: suppose that we have a quantum computer with an on/off switch, programmed ready to solve a decision problem. Then (in certain circumstances) the mere fact that the computer would have given the answer if it were run, is enough for us to learn the answer, even though the computer is in fact not run.
We discuss some seemingly paradoxical yet valid effects of quantum physics in information processing. Firstly, we argue that the act of ``doing nothing'' on part of an entangled quantum system is a highly non-trivial operation and that it is the essential ingredient underlying the computational speedup in the known quantum algorithms. Secondly, we show that the watched pot effect of quantum measurement theory gives the following novel computational possibility: suppose that we have a quantum computer with an on/off switch, programmed ready to solve a decision problem. Then (in certain circumstances) the mere fact that the computer would have given the answer if it were run, is enough for us to learn the answer, even though the computer is in fact not run.
A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it. – Max Planck, 1936 The only way quantum theory allows us to interact with quantum systems is via quantum processes. Thus, the only way we can extract information about the state of a quantum system is to apply some non-deterministic process and observe which of the branches happened. For that reason, many people refer to the act of applying certain quantum processes as measurement and sometimes even as observation . A consequence of this unfortunate terminology is that one of the most touted ‘strange’ features of quantum theory is often misleadingly described as follows: ‘The mere act of observing a quantum state changes it.’ ‘Misleading’, since the concept of measurement described in the previous paragraph is far from the passive concept of observation familiar from our macroscopic world, but is rather a non-trivial process that will almost always drastically affect the quantum state. So, the mysterious aspect of quantum theory is not that ‘observation’ alters the quantum state but, rather, that it is impossible, even in principle, to ‘observe’ a quantum system in the classical sense. Given this fundamental restriction on how we can extract information from a quantum system, what can we learn about that system by means of a quantum process? The answer is, in fact, very little! First, performing a particular measurement could, for the vast majority of quantum states, yield any outcome i with a non-zero probability. In that case, obtaining an outcome i doesn't tell us much about what the state of the system was. Second, we say ‘was’, because the measurement will moreover irreversibly change the state according to the outcome. So rather than revealing the state of a system, a quantum measurement typically erases that state from history! Nonetheless, these quantum measurements are crucial to quantum theory since they constitute the only interface between us, in our classical world, and the quantum world. The great insight of the quantum computing community is that the non-deterministic changes induced by these quantum measurements are not a nuisance, but rather an extremely useful resource.
In the Name of the Pasta, and of the Sauce, and of the Holy Meatballs … – Bobby Henderson, The Gospel of the Flying Spaghetti Monster, 2006 After the conceptual comes the practical. While the study of quantum foundations is as old as quantum theory itself, the field of quantum computing is relatively new. So new in fact that large-scale, practical quantum computing is still not a reality. A typical ‘quantum computer’ takes many months to set up before performing such astounding tasks as factoring 6 into 3 × 2. Nonetheless, if those machines would exist, we know that we would gain amazing speed-ups in solving some hard (classical) computational problems, such as those involved in breaking a huge portion of cryptographic systems in use today. Before we get into ‘quantum computing’, we should say a couple of things about ‘computing’. What is computing? Our answer by now probably won't come as such a shock: it's a process theory! Computation is indeed all about wiring the inputs and outputs of small processes together to make bigger processes. More specifically, a computation consists of a (finite) set of basic processes, which are wired together according to some (also finite) instructions, which we refer to as an algorithm or simply a program . The only essential difference between classical and quantum computation is the contents of the basic processes. For classical computation, these operations consist of things like logical operations (e.g. XOR) or reading/writing locations in memory. For quantum computation, we can extend this with quantum processes and classical-quantum interactions such as measurements. So quantum computing is all about figuring out how to write new kinds of programs that exploit these new building blocks to build faster algorithms or accomplish new kinds of tasks that aren't possible classically. The first quantum algorithms were ‘proofs of concept’, in the sense that they solved some problem much faster than a classical computer, but the kinds of problems they solved were not particularly interesting in their own right. However, this changed drastically with the advent of Grover's quantum search and Shor's factoring algorithms.
Of the many and varied applications of quantum information theory, perhaps the most fascinating is the sub-field of quantum computation. In this sub-field, computational algorithms are designed which utilise the resources available in quantum systems in order to compute solutions to computational problems with, in some cases, exponentially fewer resources than any known classical algorithm. While the fact of quantum computational speedup is almost beyond doubt, the source of quantum speedup is still a matter of debate. In this paper I argue that entanglement is a necessary component for any explanation of quantum speedup and I address some purported counter-examples that some claim show that the contrary is true. In particular, I address Biham et al.'s mixed-state version of the Deutsch-Jozsa algorithm, and Knill \& Laflamme's deterministic quantum computation with one qubit (DQC1) model of quantum computation. I argue that these examples do not demonstrate that entanglement is unnecessary for the explanation of quantum speedup, but that they rather illuminate and clarify the role that entanglement does play.
Suppose that we are given a quantum computer programmed ready to perform a computation if it is switched on. Counterfactual computation is a process by which the result of the computation may be learnt without actually running the computer. Such processes are possible within quantum physics and to achieve this effect a computer embodying the possibility of running the computation must be available, even though the computation is, in fact, not run. We study the possibilities and limitations of protocols for the counterfactual computation of decision problems (where the result r is either 0 or 1). If p⊂r denotes the probability of learning the result r 'for free' in a protocol, then one might hope to design a protocol which simultaneously has large p⊂0 and p⊂1. However, we prove that p⊂0 + p⊂1 ≤ 1 in any protocol of this type and derive further constraints on p⊂0 and p⊂1 in terms of N, the number of times that the computer is not run. In particular, we show that any protocol with p⊂0 + p⊂1 = 1 – ∈ must have N tending to infinity as ∈ tends to 0. We show that ‘interaction–free’ measurements can be regarded as counterfactual computations, and our results then imply that N must be large if the probability of interaction is to be close to zero. Finally, we consider some ways in which our formulation of counterfactual computation can be generalized.
"Information is physical and any processing of information is always performed by physical means," an innocent-sounding statement, but its consequences are profound. When quantum effects become important, for example, at the level of single atoms and photons, the classical theory of computation becomes fundamentally inadequate. Entirely new modes of computation and information processing become possible. In the last few years there has been an explosion of theoretical and experimental research in quantum computation. In this brief review we describe some of these developments.
"Information is physical and any processing of information is always performed by physical means," an innocent-sounding statement, but its consequences are profound. When quantum effects become important, for example, at the level of single atoms and photons, the classical theory of computation becomes fundamentally inadequate. Entirely new modes of computation and information processing become possible. In the last few years there has been an explosion of theoretical and experimental research in quantum computation. In this brief review we describe some of these developments.