For years now the physics community has been taking the leap into computer science through the pursuit of the quantum computer. As weird as the concepts underpinning the idea of such a device are, even weirder is the threat that this machine of the future could pose to business and government today.
There are many excellent primers on quantum computing but in summary physicists hope to be able to use the concept of superposition to allow one quantum computer bit (called a “qubit”) to carry the value of both zero and one at the same time and also to interact with other qubits which also have two simultaneous values.
A quantum computer would be hoped to come up answers to useful questions with far fewer processing steps than a conventional computer as many different combinations would be evaluated at the same time. Algorithms that use this approach are generally in the category of solution finding (best paths, factors and other similar complex problems).
As exciting as the concept of a quantum computer sounds, one of the applications of this approach would be a direct threat to many aspects of modern society. Shor’s algorithm provides an approach to integer factorisation using a quantum computer which is like a passkey to the encryption used across our digital world.
The cryptography techniques that dominate the internet are based on the principle that it is computationally infeasible to find the factors of a large number. However, Shor’s algorithm provides an approach that would crack the code if a quantum computer could actually be built.
Does it matter today?
We’re familiar with businesses of today being disrupted by new technology tomorrow. But just as weird, as the concept of quantum superposition is the possibility that the computing of tomorrow could disrupt the business of today!
We are passing vast quantities of data across the internet. Much of it is confidential and encrypted. Messages that we are confident will remain between the sender and receiver. These include payments, conversations and, through the use of virtual private networks, much of the internal content of both companies and government.
It is possible that parties hoping to crack this content in the future are taking the opportunity to store it today. Due to the architecture of the internet, there is little to stop anyone from intercepting much of this data and storing it without anyone having any hint of its capture.
In the event that a quantum computer, capable of running Shor’s algorithm, is built the first thought will need to be to ask what content could have been intercepted and what secrets might be open to being exposed. The extent of the exposure could be so much greater than might appear at first glance.
How likely is a quantum computer to be built?
There is one commercially available device marketed as a quantum computer, called the D-Wave (from D-Wave Systems). Sceptics, however, have published doubts that it is really operating based on the principles of Quantum Computing. Even more importantly, there is no suggestion that it is capable of running Shor’s algorithm or that it is a universal quantum computer.
There is a great deal of evidence that the principles of quantum computing are consistent with the laws of physics as they have been uncovered over the past century. At the same time as physics is branching into computing, the information theory branch of computing is expanding into physics. Many recent developments in physics are borrowing directly from the information discipline.
It is possible, though, that information theory as applied to information management problems could provide confidence that a universal quantum computer is not going to be built.
Information entropy was initially constructed by Claude Shannon to provide a tool for quantifying information. While the principles were deliberately analogous to thermal entropy, it has subsequently become clear that the information associated with particles is as important as the particles themselves. Chapter 6 of my book, Information-Driven Business, explains these principles in detail.
It turns out that systems can be modelled on information or thermal entropy interchangeably. As a result, a quantum computer that needs to obey the rules of information theory also needs to obey the laws of thermal entropy.
The first law of thermodynamics was first written by Rudolf Clausius in 1850 as: “In all cases in which work is produced by the agency of heat, a quantity of heat is consumed which is proportional to the work done; and conversely, by the expenditure of an equal quantity of work an equal quantity of heat is produced”.
Rewording over time has added sophistication but fundamentally, the law is a restatement of the conservation of energy. Any given system cannot increase the quantity of energy or, as a consequent of the connection between thermal and information entropy, the information that it contains.
Any computing device, regardless of whether it is classical or quantum in nature, consumes energy based on the amount of information that is being derived as determined by the information entropy of the device. While it is entirely possible that massive quantities of information could be processed in parallel, there is no escaping the requirement to adhere to this requirement with a quantum computer truly delivering this level of computing requiring the same order of energy as the thousands or even millions of classical computers required to deliver the same result.
I anticipate that developers of quantum computers will either find that the quantity of energy required to process is prohibitive or that their qubits will constantly frustrate their every effort to maintain coherence for long enough to complete useful algorithms.
Could I be wrong?
Definitely! In a future post I propose to create a scorecard tracking the predictions I’ve made over the years.
However, anyone who claims to really understand quantum mechanics is lying. Faced with the unbelievably complex wave functions required for quantum mechanics which seem to defy any real world understanding, physicist David Mermin famously advised his colleagues to just “Shut up and calculate!”.
Because of the impact of a future quantum computer on today’s business, the question is far from academic and deserves almost as much investment as the exploration of these quantum phenomena do in their own right.
At the same time, the investments in quantum computing are far from wasted. Even if no universal quantum computer is possible, the specialised devices that are likely to follow the D-Wave machine are going to prove extremely useful in their own right.
Ultimately, the convergence of physics and computer science can only benefit both fields as well as the business and government organisations that depend on both.