Quantum computer, you You may have heard that the magical super machine will quickly cure cancer and global warming by trying all possible answers in different parallel universes. For 15 years, in my blog Elsewhere, I opposed this cartoonish vision, trying to explain the more subtle but ironically more fascinating truths I saw. I see this as a public service, almost my moral responsibility as a quantum computing researcher.Alas, this work feels terrible: Over the years, the daunting hype about quantum computers will only increase, because companies and governments have invested billions of dollars, and as the technology evolves to programmable 50-qubit devices , (On some artificial benchmarks) can indeed provide the world’s largest supercomputer Running for their money. Just like in cryptocurrency, machine learning, and other fashion fields, scammers also come.
However, at the moment of reflection, I understood. The reality is that even if you eliminate all bad incentives and greed, without mathematics, quantum computing is still difficult to explain simply and honestly. As Richard Feynman, the pioneer of quantum computing, once said about his work on quantum electrodynamics that won the Nobel Prize, if it can be described in a few words, it will not be worthy of the Nobel Prize. prize.
Not that this prevents people from trying. Since Peter Shor discovered in 1994 that quantum computers can crack most of the encryption that protects Internet transactions, people’s excitement about the technology is not just out of curiosity. In fact, developments in this field are often reported as business or technical stories rather than scientific stories.
If a business or technical reporter can tell the reader truthfully, “Look, there are all these deep quantum things under the hood, but all you need to understand is the bottom line: physicists are about to build faster computers and change everything.”
The problem is that quantum computers will not completely change everything.
Yes, they may one day solve some specific problems in a few minutes, which (we think) will take longer than the age of the universe on classic computers. However, most experts believe that quantum computers are also of little help to many other important problems, if any.In addition, although Google and other companies have recently reliably claimed that they have achieved artificial quantum acceleration, this only applies to specific, esoteric benchmarks (those I Help development). A sufficiently large and reliable quantum computer can outperform classical computers in practical applications such as cracking codes and analog chemistry. This may still have a long way to go.
But how can a programmable computer be faster for certain problems? Do we know which ones? In this case, what does a “big and reliable” quantum computer mean? To answer these questions, we must study in depth.
Let’s start with quantum mechanics. (What can go deeper?) The concept of superposition is difficult to express in everyday language. Therefore, it is not surprising that many authors have chosen a simple method: they say that superposition means “simultaneously”, so a qubit or qubit is just a bit that can be “0 and 1 at the same time,” while the classical bit can It is one of them. They continue to say that quantum computers will achieve their speed by using qubits to try all possible solutions in a superposition-that is, trying all possible solutions simultaneously or in parallel.
This is the fundamental error in the popularization of quantum computing that I began to think, and the error that caused all other errors.From here to quantum computers can quickly solve problems such as Traveling salesman problem By trying all possible answers at once-almost all experts think they can’t do this.
The problem is that for a computer to be useful, at some point you need to look at it and read the output. However, if you look at an equal superposition of all possible answers, the rules of quantum mechanics say that you will only see and read random answers. If this is what you want, you can choose one yourself.