Quantum computers, which could one day blow past the limits of today’s machines, have been a distant dream in the computing world for decades. The challenges of harnessing the weird properties of subatomic particles for computing has made them more intriguing science project than practical technology. But what if, finally, workable quantum computers are nearly within reach?
Recent technical advances have led companies like Google and IBM to predict they will be able to build full-scale quantum systems by the end of this decade. Sensing that the technology may mature years earlier than it had expected, the Pentagon has begun a study to see if any of the commercial efforts could produce results by 2033.
Quantum systems, when they arrive, will remain tools for specialists rather than the masses. But, in the fields where they have most impact, they could usher in a period of considerable discontinuity. The most widely used forms of encryption would be vulnerable, meaning that anyone hoping to protect today’s information from future prying eyes should already be switching to new forms of quantum-proof cryptography. The technology would put new tools into the hands of scientists who could bring breakthroughs in materials or pharmaceuticals, opening up new processes and markets or spelling doom for old ones. In finance, it could bring a better understanding of complex risks and more efficient pricing in markets.
Yet it has been hard even for the companies with the most at stake to know how seriously to prepare for a technology that has always seemed beyond the horizon. Some have made it a focus of advanced research for years, without seeing results.
Like any new technology, timing is key. Investing too early, besides being a waste of resources, risks a backlash. The history of artificial intelligence has been punctuated by “AI winters” or periods of disillusionment that followed bouts of excessive optimism. The intense competition between companies racing to build the first workable quantum computers has made this a natural breeding ground for tech hype, raising the risks of overpromising and disappointment.
Advances in fundamental science have brought the first prototype quantum machines, setting the stage for an attempt to turn these designs into full-scale systems. But there are still considerable engineering challenges to be overcome, injecting uncertainty into the companies’ ambitious timetables. Jensen Huang, CEO of Nvidia, poured cold water on the industry earlier this year when he predicted practical machines were probably still 20 years away, though he later suggested he might have been too negative.
The interplay between quantum technology and AI complicates things further. Rapid advances in AI could make today’s computers far more capable, reducing the need for quantum machines, as Sir Demis Hassabis, head of AI at Google, has suggested. Most in the industry, though, predict a symbiosis between the two technologies that will eventually give a lift to both.
Despite the uncertainties, recent advances in quantum computing should still be a spur to action. This includes working on the tools and the skills that will be needed to make the most of workable quantum machines. On the tools side, new algorithms will be needed to take advantage of the particular properties of quantum machines and make them useful in tackling a wider set of problems.
A bigger skilled workforce will also be essential. Quantum experts have already been in high demand, even before the technology reaches industrial scale. Even if it is difficult to predict precisely when the quantum era will dawn, it’s not too soon to start preparing.