Regular computers at their most fundamental level use transistors to process information. Each transistor represents a bit, which is a 1 or a 0. When the transistor is on (1), this allows an electrical current through which is essentially allowing a flow of information. When it is off (0), the transistor prevents this flow of information. This simple process is the basis for nearly all computation today. When combined, transistors can carry out logical and mathematical processes and entire programs can be created using these basic operations. With enough, transistors, we can compute nearly anything. In order to increase computational power by adding transistors, but not creating large clunky devices, we have made transistors smaller and smaller. This is initially simple enough (provided we have the adequate technology to do so) as a transistor only has to be able to allow a flow of electrons and also be able to stop it. However, this is starting to become an issue as we are getting to the point where transistors are small enough that electrons can get across via quantum tunnelling. This is an issue because these transistors cannot work as switches if they leak current. As a result of this, scientists have been looking for revolutionary new methods of computation and a popular solution has been quantum computing. Quantum computers are interesting because instead of regular bits, they use qubits. Instead of transistors, we can represent qubits using any particle that exhibits quantum behaviour such as an electron or an atom.

Qubits are a combination of 0 and 1 at once and this is known as superposition. However, when we try to observe these qubits, they must collapse into one state – 0 or 1. What is so amazing about superposition is that one bit can effectively store both possibilities (1 or 0) at the same time. If we have several qubits, this starts to become very practical as if we have n bits, we can store 2n different values simultaneously. In addition to superposition, entanglement is also used to perform computation. Entanglement means that one particle has an effect on other particles such that the quantum state of a particle in a group in not independent from the others and cannot be described without considering the other particles. A very useful aspect of entanglement is that it can occur no matter the distance between two quantum particles. If two entangled particles are a thousand miles apart and we measure a property of one particle, the same property will be observed for the second particle when measured. Entanglement is necessary when performing computations, however, when a quantum system is observed or measured, entanglement is broken. This is because the quantum environment is no longer being preserved and therefore, just like the superposition of a particle must collapse into a 1 or a 0, entanglement is broken. This is useful for making measurements and getting results after quantum computation. It is the use of entanglement and superposition that allows us to store and manipulate data in quantum computing.

The result we get from a quantum computation is in the form of a probability. For example, when guessing a password, we get a probability that the input is the password. Therefore, we need to repeat the computation to approach an actual answer. The algorithm that is used to then find the correct solution is called Grover’s Algorithm. In essence, what is does is takes the wave amplitudes of the particles that represent the qubits (each qubit representing a possible answer) and amplifies the correct answer. This correct answer can then be detected and so we would get a result. How this works is that initially, each qubit is put into an equal superposition of all possible states. Then we use an Oracle Function which will go through each qubit and if a qubit is correct, it will flip its amplitude (turn it from positive to negative). The next step is that this qubit is flipped again, but this time the amplitude is increased, thus increasing the probability of this answer. Now we can use this set of qubits as an input and perform Grover’s algorithm again to increase our chance of finding the correct answer. This algorithm is repeated until we approach the correct answer. So, in essence, this algorithm ensures that the qubits are likely to be in the correct state which then gives us the right answer.

The first and most obvious application of quantum computers is using the quantum properties of qubits themselves to model quantum mechanics. As electrons and other particles with quantum properties are themselves in superposition, normal computers can only create approximations which are often lacking in accuracy. Quantum computers could revolutionise our ability to model the universe. Some more practical areas this ability to model quantum systems can aid us in include nanotechnology and medicine. By utilising properties like superposition and entanglement, it is possible that we could create completely new nanodevices. Similarly, if we can understand proteins better on a quantum level, using these new insights, we can make breakthroughs in medicine.

What is perhaps more exciting is the ability of quantum computers to solve what are known as NP-Complete problems. NP-Complete problems are defined as problems that are NP problems (a solution for the problem can be verified as the correct solution in polynomial time) and are NP-Hard (where every NP-Hard problem can be simplified in polynomial time to the hardest NP problem). Although we do not have to actually understand how to solve these problems, the ability to brute-force these problems using Grover’s Algorithm would allow us to achieve things we couldn’t do with current-day computers (however, if this does help us solve an NP-Complete problem, then we could solve all other NP-Complete problems as they can be simplified into each other). In addition to all of this, an entirely new field of problems may be defined by quantum computers – BQP problems. BQP problems are defined as all problems with a yes or no answer that can be solved by quantum computers in polynomial time. The broad category of problems that current computers could be able to solve is known as PH problems. These are problems that you get if you take an NP problem and make it more complex using second-order logic (i.e. using statements like ‘there exists’ and ‘for all’). We are currently unsure if there are any BQP problems that don’t exist in PH, but if there are, then even if we overcome quantum tunnelling, there will be problems that classical computing can never solve even after trillions of years. This is a very exciting possibility that demonstrates the potential that quantum computers could have in the future.

This can be useful in analysing weather patterns or training neural networks. This incredible ability of quantum computers will also be very useful as our world shifts to automation and artificial intelligence. For example, if we look at self-driving cars, there are several factors such as routes and road-closures that an automated system will have to take into consideration to create an efficient system for all self-driving cars. Quantum computers can perfectly plan each car journey, choosing the route of each car and timing when cars merge into a road perfectly to ensure the maximum efficiency for all journeys. This could eliminate traffic jams and make the shift to smart cities much more realistic. In addition, quantum computing can help artificial intelligence find the best solution to a problem much quicker, allowing us to strengthen artificial intelligence models much faster.

However, the ability of quantum computers worries many people as it has the ability to break all modern encryption in a matter of hours (while it would take current supercomputers millions of years). Currently, encryption uses a public key to encrypt communications and a private key to decrypt them. A public key is simply a very large integer and the private keys are the prime factors. However, a public key is visible to anyone and so if we can factorise it, we also know the private keys. The reason why this isn’t already exploited is that the integers are so large that current computers would take millions of years to factorise them. However, in 1994, Peter Shor came up with an algorithm (known as Shor’s Algorithm) for a quantum computer to find the prime factors of an integer in polynomial time. This means that with quantum computers, hackers could implement this to break any modern encryption in a matter of hours. Quantum computers could also render currencies like bitcoin useless as the process of bitcoin mining also relies on breaking encryption. In theory, it would be possible to use quantum computers to mine the majority of remaining bitcoin. Finally, blockchain also relies on current cryptography techniques. Many financial companies looking to ‘step into the future’ are starting to use blockchain, but this could also become a problem because of quantum computing. So many of our modern systems rely on this system of encryption and this worries people as quantum computer could compromise all of this.

But do we need to worry just yet? The important thing to note about quantum computers is that the qubits need to be in very specific conditions to exhibit quantum properties such as superposition. The temperature needs to be almost absolute zero and they can’t be exposed to any radiation. As a result, quantum computation is inaccessible to most everyday users and it is very expensive to run. In addition, it cannot carry out computations for a very long time as things like the quantum logic gates within the quantum computer itself can disrupt the quantum environment. As a result, quantum computing is only accessible to large institutions, corporations and governments. This means we still have time to find new encryption methods that cannot easily be broken by quantum computing in the future. However, while our personal data is safe for now, governments might try and use this ability of quantum computers to hack other governments. China and Russia are already known for carrying out large hacking to steal designs for new military technology or influence elections. In addition, national security could be compromised by quantum computing. But again, since quantum computing isn’t capable of this yet, we might not have to worry as we can exploit potential alternatives such as biometrics.

To conclude, while (as with anything), quantum computing comes with its risks, we need a revolution in computational capacities as transistors are approaching their limits. As our world is rapidly modernising, we are looking harder than ever for breakthroughs in nanotechnology, medicine, weather patterns and artificial intelligence to help us solve large issues that face us today such as climate change. The main barrier holding us back from rapid advances in a variety of different areas is our limit in computation that can produce accurate models and do mind-boggling numbers of calculations in short periods of time. Even if quantum computing might compromise some of the systems we rely on so much now, it will still be ten to twenty years before quantum computing can even be useful given the barriers that exist right now. This gives us a substantial amount of time to find alternatives that fix the vulnerabilities our current systems have. Furthermore, if we can find BQP problems that don’t exist in PH, we will have practical problems that can only be solved by quantum computers, meaning there is no other alternative to the capability of quantum computing. We don’t know if quantum computers will only be used by large corporations and governments or be available to the general public. However, the abilities of quantum computation are undisputable and something we should definitely welcome if we are looking to rapidly modernise our world.

References