THROW a ball against a wall. Watch it bounce. Catch it. So far, so good. The ball didn鈥檛 suddenly disappear through a wall, or spontaneously turn into something else. This perfectly normal, predictable behaviour you just witnessed is classical physics in action. Up until 1900, it was just called physics. Whether you were discussing particles or planets, the rules that governed the bouncing of that ball could be used to describe anything in the universe. The world made sense.
Then quantum mechanics happened. Over the space of four breakneck decades, our world became a wildly unfamiliar place. Objects acted as though they could be in two places at once, particles led double lives as waves and information appeared to travel faster than the speed of light. The weirdness of the quantum world has become legendary, but the origins of that weirdness remain deeply mysterious. Theorists like me continue to struggle with a question that seems almost unbearably basic: what gives the quantum world its distinctively counter-intuitive quantum flavour?
Advertisement
The truth is that we still don鈥檛 know. And that ignorance has profound consequences for our understanding of reality, as well as our ability to use quantum phenomena in real-world technology. While different proposals have emerged over the years, it seems that we may now be getting closer to the real answer.
Nowhere could this be more transformative than in the quest to build a quantum computer. Billed as the superior successor to ordinary PCs, their fabled edge has been attributed to a whole range of quantum phenomena. If we have identified the true root of quantumness, a new computing revolution will soon be here.
It all starts with Albert Einstein. At the beginning of the 20th century, he was at the forefront of a wave of young physicists making radical discoveries about the world around us. In 1905, he convincingly argued that light 鈥� long believed by the physics establishment to be a wave 鈥� must act in a very particle-like manner. The physics world was confounded further when experiments with subatomic particles called electrons showed them spreading out like waves.
鈥淭he truth is, for all its revolutionary potential, the source of quantum computing鈥檚 power remains shrouded in mystery鈥�
Explaining these new phenomena involved rethinking the structure of light and matter. Instead of being solid balls, microscopic particles and atoms were found to have altogether slipperier identities. They were best described by mathematical entities known as wave functions, which calculated the probabilities of their being in different places. Until you spotted exactly where they were, all those different possibilities .
Replacing hard certainties with wave functions and probabilities had some startling consequences. It meant that when faced with a hard barrier, a particle had a non-zero probability of being located on the other side, allowing it to perform the seemingly impossible feat of tunnelling straight through it.
Many physicists didn鈥檛 like this picture. Chief among them was Einstein, who was appalled at the consequences of the quantum world view he had helped usher in. In an attempt to highlight the absurdity of this radical new physics, he collaborated on a paper with two like-minded colleagues at Princeton University, Boris Podolsky and Nathan Rosen. Known as the EPR paper, after its authors, it laid out a troubling consequence of a universe governed by probabilities. Under the right conditions, EPR explained, two particles could have their wave functions tied together, or entangled, so intimately that any action you performed on one seemed to instantaneously influence the other 鈥� no matter how far apart they were.
This was heresy. In the language of classical physics, signals can only travel at the speed of light. This means that objects need more time to communicate with things further away from them than they do with ones in their immediate vicinity 鈥� a principle known as locality. Following that logic, classical physics says two entangled particles placed a light year apart would need a full 12 months to react to any change in one another. According to EPR, however, such reactions seemed to be instantaneous. Little wonder Einstein called the process 鈥渟pooky action at a distance鈥�.
Overruling Einstein
But not everybody got such a fright. Erwin Schr枚dinger, another pioneer of quantum theory, embraced entanglement as the phenomenon that definitively separated the quantum and classical worlds, calling it the characteristic trait of quantum mechanics. To physicists such as Schr枚dinger, its very spookiness made it an ideal place to find the key ingredient that gave the quantum world its quantumness.
All the while, sceptics like Einstein were desperate to explain this bizarre twinning by purely classical means. One suggestion was that the two particles had predetermined properties that were uncovered by means of observation. If you took a pair of gloves to opposite ends of the universe, for example, it would hardly be surprising to know that one glove was left-handed as soon as you identified the right-handedness of the other. The handedness of one glove didn鈥檛 somehow emerge when the other was examined 鈥� it had been an integral part of its identity all along.
Such a theory, which appeared to explain entanglement while preserving locality, became the raft that quantum naysayers clung to with increasing desperation. That all changed in 1964, when physicist John Stewart Bell conceived a series of thought experiments that would be able to distinguish between true quantum entanglement and a classical equivalent that preserved locality (see diagram, below).
In a series of ever more accurate experimental implementations of his ideas from 1972 onwards, entanglement was shown to be a reality. Quantum physics was gleefully non-local. 鈥淭hese experiments sealed the fate on any hope of rescuing a crystal-clear picture of reality,鈥� says Matty Hoban, who works on quantum information theory at Goldsmiths, University of London.
But while non-locality is clearly an integral part of what makes quantum mechanics weird, it wasn鈥檛 the end of the story. For one thing, non-locality only applies to two or more particles. It isn鈥檛 present in any weird quantum effect involving a single particle, such as its ability to tunnel through walls or take multiple identities. Some other law of classical physics was also being broken.
Unreliable witnesses
The answer to this dilemma, once again, can be traced back to Einstein. Another assumption that EPR and its supporters had made was that quantum experiments obeyed similar rules to classical ones. They had assumed that any object had fixed properties, which could be uncovered by asking the right experimental question. A left-handed glove was always going to be left-handed, it was an intrinsic property that didn鈥檛 vary. Whether you tested its handedness by putting it on, or asking a friend to put it on, or putting it on blindfolded while underwater, those different contexts shouldn鈥檛 change its identity.
What Bell鈥檚 tests had shown was that in the quantum world, they do have an effect. The result you got when measuring one of a pair of entangled particles was strongly linked to the measurement being made of its partner. In other words, each particle鈥檚 answer must depend on the context in which it was being questioned. Viewed through that lens, all of the quantum world鈥檚 most counter-intuitive results suddenly made sense. Interrogate a quantum particle under the right conditions, and you could influence the nature of its confession. Now a particle, now a wave; now on one side of a wall, now the other. 鈥淐ontextuality is a hallmark of quantumness,鈥� says Mark Howard, who works on quantum computing at the National University of Ireland in Galway.
For Ana Bel茅n Sainz at the Perimeter Institute for Theoretical Physics in Canada, that makes contextuality the more inherently quantum phenomenon, with non-locality simply one way in which it manifests itself.
To find out if it really is the source of the quantum world鈥檚 inherent weirdness, however, we would need to construct a test analogous to the one Bell devised for non-locality. That would involve setting up the same experiment in two different contexts, and comparing outcomes. This turns out to be a significant challenge, as even minor imperfections in equipment can lead to experimental noise, introducing slight variations in your results. How do you tease apart differences due solely to experimental noise from those due to quantum weirdness?
An answer came in 2016, when two teams conducted independent experimental tests of quantum contextuality. Their approaches were robust enough to experimental noise to convincingly show that objects in the quantum world really are unreliable witnesses: the answers they give depend not only on the questions you ask, but on the context in which you ask them. For Sainz, these experiments represent a milestone in the history of quantum physics.
Despite its more central role in the quantum universe, for decades contextuality has lacked the recognition it deserves. The lack of a snappy Einstein-endorsed catchphrase is only the tip of the iceberg. 鈥淎 lot of it comes down to shock value,鈥� says Hoban. For physicists, non-locality seems more shocking because it touches upon the fundamental constituents of reality: things like space, time, and causality. Whatever the reasons, its dominance may be about to be challenged.
If you believe the hype, 2019 could be the year the world鈥檚 first large-scale quantum computer is built. According to the leading contenders in that race, such a computer would harness the quirks of quantum mechanics to perform tasks beyond even the fastest supercomputers, from discovering new cancer drugs to improving weather forecasting and traffic control. But what makes them so special?
Under the hood, a computer is just a device capable of manipulating information to perform a desired task. In a regular laptop or PC, this is done through the medium of billions of transistors located on chips at the machine鈥檚 heart. Each transistor can be in one of two states: on 鈥� with current running through it, say 鈥� or off. These two states, referred to as 1 and 0, represent a single bit of information. Stick enough bits together, and you can do anything from calculating the trajectories of a space shuttle to building a model universe.
In quantum computing, the classical bit is replaced with a quantum bit, or qubit. Instead of only existing as 0 or 1, the wave function of a qubit allows it to exist in a new combination of 0 and 1, known as a superposition. So far, so simple. But at this point, many attempts to explain the power of quantum computing go badly off the rails. The standard version goes something like this. Classical bits have to choose between one of two possible states, but 鈥� thanks to superposition 鈥� a qubit can represent both 鈥渁t once鈥�. So while a classical computer has to try each possible solution one at a time, quantum superposition allows a quantum computer to try all possible solutions to a problem at once, making it faster and more powerful than a regular computer. Sounds fantastic. Only trouble is, it is pure nonsense. 鈥淚f things were that simple,鈥� says Howard, 鈥渨e would all be rich and quantum algorithm design would be easy.鈥�
The truth is, for all its revolutionary potential, the source of quantum computing鈥檚 power remains shrouded in mystery. 鈥淯nderstanding what quantum computers are 鈥榓ctually doing鈥� is a difficult problem, even for experts,鈥� says Hoban. In some cases, we know quantum algorithms have been able to perform seemingly impossible feats by using phenomena like entanglement and superposition 鈥� but we also know that some of their achievements can be simulated on a regular classical machine. In other words, just using these weird effects isn鈥檛 enough to give quantum computers the edge. 鈥淲orryingly, we do not yet have a robust way of identifying the necessary and sufficient conditions a quantum system has to possess to see an advantage over classical computation,鈥� says Angela Karanjai, a quantum physicist at the University of Sydney.
But some people think they know where that advantage might lie. In 2014, Howard and his collaborators published a landmark paper showing contextuality could be the engine at the heart of quantum computing. They started with simple systems known as stabiliser circuits, which are inherently quantum but easy to simulate on a regular classical computer. In the parlance of quantum computer scientists, these circuits aren鈥檛 鈥渦niversal鈥�; not every possible quantum algorithm can be run on them. It also turns out that they aren鈥檛 sophisticated enough to display contextuality.
Howard and his colleagues proved that as soon as you give stabiliser circuits the ability to create situations that are contextual, you make them universal. 鈥淭his result broke fresh ground and uncovered part of the bedrock of a quantum computer鈥檚 power,鈥� says Hoban.
Quantum superpower
So is contextuality the secret engine responsible for the quantum computing speed-up? Given the variety of models of quantum computation that are available, Howard is cautious about overgeneralising. Nonetheless, knowing that contextuality can, at least some of the time, make computation truly quantum is still a helpful result. Recent work by Karanjai and her collaborators has shown that the amount of contextuality a quantum circuit exhibits puts a limit on the memory a classical computer needs to simulate it. The more contextuality, the larger the memory required. 鈥淭hese results tell someone building a quantum computer to build it using systems that exhibit contextuality,鈥� says Karanjai, because these can offer more computational power.
But contextuality doesn鈥檛 only have great technological promise. After a century of uncertainty, it could finally allow us to chart the hazy boundary between quantum and classical physics.
For Schr枚dinger, the very fact that entanglement was 鈥渟pooky鈥�, that it upended our classical preconceptions, was enough to herald it as lying on the other side of the divide. But the work of Howard, Karanjai and others suggests that a more rigorous definition might be attainable. Rather than being a ragbag theory comprised of those phenomena at odds with our classical expectations, quantum physics could at its heart be a theory about computing. Mind-blowing as it may seem, the true hallmarks of quantum behaviour might be those that provide a clear computational advantage.
If that holds up, then the race to build a quantum computer will do more than revolutionise computing. Its real legacy, in fact, could be to finally pin down what makes quantum mechanics quantum.
