Last January, IBM announced at the Consumer Electronics Show in Las Vegas that they were introducing the world’s first integrated quantum computer for business and research purposes, to be made available later this year. This kicked up the usually breathless press coverage about how quantum computers are going to change everything. The truth is, we aren't really sure what quantum computing will change, but that doesn’t mean that its all hype. It might not "change everything", but it's likely that nothing will remain unaffected by it, even if indirectly.
What Is Quantum Computing and What Does It Do?
Quantum computing is when a computer uses the quantum superposition of particles to store data the way a bit does in a classical computer. Qubits, as they're called, exist in an indeterminate state inclusively-bounded by 1 and 0. This means they can be 1, 0, or both 1 and 0 because the universe hasn't decided yet which one it wants to be.
This allows us to perform several calculations at once by exploiting the superpositions of these qubits, opening the door to solving classes of problems that might require hundreds if not thousands of years to solve using a classical computer.
However, there is a catch. Quantum computing is an exceptionally delicate thing since maintaining a suspended quantum particle in a superposition can only be done for about 100 microseconds. It also needs extremely cold temperatures and superconductors, not exactly things that are going to fit in your iPhone. This kind of hardware makes quantum computers highly specialized equipment that is only really practical for very specific tasks right now, things like prediction modeling and optimization problems on complicated systems with a large number of variables.
And even then, their use will initially need to be rationed for a while, just as the old UNIVACs used to be. The CDC will be using quantum computers to model the upcoming flu season, so local hospitals that aren't major research centers are going to have to wait in the middle towards the back of a very long line until they're done. That is unless Goldman Sachs shows up needing to do financial forecasting before you get to the head of the line.
So What Will Change With Quantum Computing?
First off, the one thing we know for sure is that modern RSA encryption is toast. RSA encryption relies on the practical impossibility of a classical computer finding the right prime factors of a very large integer, like 500-digits-long kind of large. This could take hundreds of years using our most efficient algorithms on a classical computer, and thousands of years if you just tried to brute force your way to an answer.
The reason why this is solvable with a quantum computer is that it was already solved by Peter Shor in 1994. Shor's algorithm, as his solution is called, needs a sufficiently powerful quantum computer to run on in order to break RSA encryption though, and one doesn't yet exist. Soon it will, and the way we secure data will be as effective as using a hook and loop latch on your front door is at securing your home. We will have to invent an entirely different way of securing all of our existing data, that much we know.
As for the things we think quantum computing will change, the first candidate is the way we organize macro-level systems like telecommunications infrastructure and roads. How to make these systems optimally balanced between cost and utility is one example of the kind of problem quantum computers might be able to solve thanks to the quantum superposition of qubits.
The same optimization problem plagues the global supply chain, and this is no small thing. The transportation sector wastes an unspeakable amount of money—we’re talking possibly hundreds of billions of dollars here—due to the kinds of hidden inefficiencies that optimization will identify.
And not just governments and business are likely to benefit from quantum computing, but medicine, astronomy, and other sciences are excellent candidates for transformative advances. Astronomers hunting down exoplanets have mountains of petabytes of data that needs to be processed in order to glean scientifically relevant insights from it, and this is the sort of data processing that we think quantum computing will change in a transformative way.
In medicine, quantum computing can possibly accelerate the pace of medical breakthroughs at an incredible rate by processing the kinds of multiple-variable problems that make research in these fields such a challenge. By exploiting the quantum superposition of qubits to model the kinds of analysis used when researching diseases and developing new medicines, we might discover all kinds of new medicines and treatments that no one ever would have thought of.
The most significant development might be in physics, where researchers exploring superconductivity hope to one day use quantum computing to identify a material that is superconductive at room temperature by using qubits to model different compounds and test their characteristics.
If such a superconductor is found, then that will change everything, honestly. It would allow us to eliminate energy loss from the transmission of electricity and transform our energy grid, reducing the power generation required to power everything to a fraction of what it is today. If we get our power requirements down enough and we could power the world on renewable energy very, very soon.
So What Can't Quantum Computers Do?
Quantum computing strength is in the superpositions of its qubits. What we do with that data, however, won’t be helped by quantum computing at all, at least not in any way we can see now. A quantum computer won’t “run a program” the way our computers do today; quantum computers would have to run a program's instructions line by line just like a classical computer would have to do. Superposition doesn't generally help make this faster.
It’s conceivable that after physicists and chemists discover a room-temperature superconductor that they could be made small enough to fit into network servers or home computers, but if we ever use quantum computing in our day-to-day computing tasks, its more likely that our classical computer reached a point in a program that requires the kind of task that is best handled by a quantum computer, like the factorization of an integer, and it would either connect to a cloud-based quantum computer to process this problem for a result, or use a built-in quantum chip, a QPU, to solve it the way today's CPU's offshore graphics processing to the GPU.
Performance increases of a million-fold that some are anticipating though will really only happen on poorly-written programs that repeatedly get bogged down performing unnecessarily complex algorithms that get passed off to a QPU. No well-written program is going to implement a brute force factorization algorithm in any program any of us are likely to run in our day to day lives.
We Simply Don't Know the Limits of Quantum Algorithms
That said, no one thought the Mark 1 or the UNIVAC would develop into anything beyond a room-size calculator. They could not predict the kinds of uses that those calculations would be put to.
Today, we’re building room-sized quantum calculators and that is all we can really see them as; someday, someone will show us how we were thinking much too small. After all, it wasn’t until the developers of the BASIC language included an INPUT command during the languages third revision that anyone realized they could now write a program that functioned as a game and run it on a mainframe computer.
Even the most advanced video games or just 1s and 0s. Algorithms and I/O devices like monitors and keyboards make those 1s and 0s more than just calculations.
Ultimately, it's just too soon to know anything for certain without looking like an idiot in 10 years time. Theoretical computer scientists and mathematicians, who often develop the algorithms that get incorporated into the programs we use in our day to day lives, are only now starting to explore what kinds of quantum algorithms can run on a quantum computer. After they develop these, it will be up to others to implement those new quantum algorithms into different kinds of programming.
What we can say, however, is that things will change with quantum computing and the subsequent development of advanced quantum algorithms and that this change is coming soon.