The Australian government is going all in on quantum computing. After investing more than $100 million on “quantum technology” in 2021, it is now reportedly considering spending up to $200 million on purchasing a “quantum computer” from a US company.
Is this a sensible decision? You might think so, if you read reports from media, industry and government predicting that quantum computers will revolutionise many fields of science. Two common examples given are drastically accelerating the design of better batteries and drug discovery.
Given the scale of investment, from governments around the world and also private companies, you might think quantum computers are a sure bet to reach these amazing goals. Unfortunately, in the words of US quantum computing theorist Scott Aaronson, the reality is “much iffier”.
What’s so iffy about quantum computing?
In a recent perspective article in the Proceedings of the National Academy of Sciences, French physicist Xavier Waintal warned of weaknesses in “the quantum house of cards”. Waintal notes that “a simple task such as multiplying 3 by 5 is beyond existing quantum hardware” and that a useful quantum computer might “require an improvement by a factor of one billion” on the error rate of current devices.
Skeptical voices such as Waintal’s are growing louder as success still seems a long way off, despite huge investments of time and effort. While companies like IBM and Google are still spending on quantum computing, China’s tech giants are dumping their own quantum computing labs.
It’s possible that a chain of breakthroughs could occur over the next few years, leading to useful quantum computers. We have seen other technologies, such as traditional computing chips, make huge improvements in short amounts of time.
However, improvements in traditional computing have resulted from massive investment over many decades. Before we can decide whether such a large investment is worth it for quantum computers, we need a clear understanding of their applications.
What would quantum computers really be good for?
One application that first drew attention to the idea of quantum computers (in the 1990s) is their ability to break some kinds of encryption commonly used to store and transmit data. However, new encryption methods have since been developed that would be safe from quantum computers.
Now attention has moved to the potential ability of quantum computers to solve problems in biology and chemistry, such as drug discovery and battery design. The idea is that biology and chemistry are governed by the same laws of quantum mechanics that control the workings of quantum computers.
This argument seems plausible, but it has some problems. One is that, although chemistry and biology do follow the laws of quantum mechanics, in many cases their behaviours are almost indistinguishable from non-quantum ones.
In fact, there is no guarantee that quantum computers will be able to outperform current computers when applied to problems in biology and chemistry.
It’s possible that once we have built a quantum computer we will be able to find ways to make it solve problems in biology and chemistry faster than a normal computer, but it’s far from guaranteed.
Can AI outdo quantum computers?
Quantum computing advocates are not alone in wanting to better simulate chemistry and biology. Many other scientists are working on this problem as well.
For example, quantum chemistry and molecular simulation are two very active research fields. These scientists are making rapid progress on solving many of the problems that supposedly justify the development of quantum computers.
Most excitingly, these fields are taking advantage of recent developments in artificial intelligence to massively improve the scale and accuracy with which they can simulate biology and chemistry. In one recent example, researchers trained an AI algorithm on a huge dataset and used it to study a large range of chemical and biological systems with impressive accuracy and speed.
“Useful” quantum computers are still some distance away, if they ever eventuate. And even if they are built, they may not be as useful as their advocates hope.
So while it’s reasonable for our government to invest in quantum computing research, we should be realistic about what we hope to get out of it. And we shouldn’t neglect other avenues in the quest to understand chemistry and biology at the most fundamental levels.
Just as a smart investment strategy is to diversity, we should do the same with our research funding, backing many different potentially exciting technologies. We should be humble about our ability to know which research directions are the most promising, as the future is incredibly hard to predict. If it wasn’t, we wouldn’t need a quantum computer in the first place.
Do you know more? Contact James Riley via Email.