On December 17, 1903, in Kitty Hawk, North Carolina, Orville Wright flew an aircraft he built with his brother Wilbur. Twelve seconds later, Orville experienced a bumpy landing, his journey was over, and the airplane era began.
Original title: Google’s low-key claim that “quantum supremacy” was hit in the face, how far is humanity from the quantum age?
But for newspaper readers at the time, they may not have realized what had happened to the historic event. An article about the Wright brothers was widely reprinted with the headline “Appearing to be a success” and said the news was “exciting”, but it did not see Orville’s brief flight as a landmark moment for humanity. Even a few years later, another story described the Wright brothers as “the inventor of the most successful aircraft to date”.
Speaking of which, I can’t help but think of the near-cold treatment Google received in October, when it published a research paper saying that its team in Goleta, California, used an experimental quantum computer with a processor called Sycamore that can perform a random number generation in 200 seconds. According to it, even today’s fastest traditional supercomputers take 10,000 years to do the job.
Google’s scientists claim that they have achieved “quantum supremacy” by building a quantum system that can accomplish tasks that traditional computers simply cannot accomplish – unless you are willing to wait 10,000 years for a result. For years, quantum watchers have been waiting for a moment when they talk about it, often referring to the Wright brothers.
But when Google officially announced its achievements (the news leaked a month ago), its quantum computing rival IBM has claimed that quantum supremacy has not yet arrived. The company says the state-of-the-art 200-petaflop IBM supercomputer installed at Oak Ridge National Laboratory can perform Google’s benchmark missions in two and a half days. That’s far less than Sycamore’s speed, but it’s enough to be a reasonable choice.
For most people, ruling that Google and IBM are at odds is an unrealistic task. But the idea of quantum supremacy, and how it corresponds to the defining moments of technological progress in the past, is worth thinking about.
There are some technical details to consider here.
From 1956’s UNIVAC 1 to your pocket smartphone to today’s most powerful supercomputer, “traditional computers” embrace everything based on traditional 1 and 0 digital computing. In contrast, quantum computers use completely different concepts, physics, and techniques, and the basic component is a qubit or qubit — it can represent 1, 0, or both. This gives them the potential to perform difficult computational tasks at speeds that traditional computers cannot match.
Even if you agree with IBM’s position that its supercomputer could have completed Google’s benchmark in two and a half days, rather than 10,000 years, the fact that Google’s Sycamore completed the test in just 200 seconds should spark interest in the potential of quantum computing.
But the word “quantum supremacy” itself has an epoch-making finale, as if its realization would immediately usher in the end of traditional computers. Google’s experiment in Santa Barbara, even in the best way possible, failed to do so. As IBM’s counter-argument points out, the future is more likely to involve the work of traditional and quantum computers.
The term “quantum supremacy” was originally coined and explained by John Preskill, a professor at the California Institute of Technology, who said in a 2012 paper that quantum supremacy was not an inflection point, but the key to a new era.
The whole goal of quantum computer research at Google, IBM and elsewhere is to create computers that can handle tasks that traditional computers simply can’t handle, and that have world-changing implications, such as modeling complex molecules.
By contrast, Google’s random numbers exercises are relatively flat, except that only 200 seconds. Sergio Boixo, the head of the project’s theoretical team, calls it the ‘Hello World program for quantum computers’, the equivalent of a programmer doing super-simple programming exercises in a new language, just to make sure everything is going as expected. “It’s the first thing we want to do,” he explains.
Although Google has won itself the highest prize in quantum, it also acknowledges that there is still a lot of work to be done before quantum machines are widely used.
In the field of quantum computers, there are two ultimate problems to be solved, after all, if these two problems are not solved, everything is all in vain. One is the problem of quantum precision, due to various complex reasons, quantum computer error rate is very high, may calculate 100 times 10 plus 10 such a simple problem, there will be a mistake, but it is worth noting that Google, which has been using quantum supremacy as a gimmick, has never publicly stated a breakthrough in this regard, therefore, The accuracy of Google’s quantum computer may be worrying.
Another problem is quantum extensibility, the higher the quantum bit of quantum computers, the more accurate the overall operation will decline, for now, it is not clear how much Google’s quantum computer is, or whether Google really solved this obstacle to human progress.
Marissa Giustina, a member of the hardware team, drew a timeline with a whiteboard, showing that quantum computers, fault-tolerant and error-correcting, were less than half way through becoming the mainstream reality.
At the end of the day, we shouldn’t judge Google’s quantum computer by whether it’s eliminated by other calculations, just as we shouldn’t be wondering if Wright killed hot air balloons.
At Google’s event in Goleta, Neven proposed another analogy related to human conquest of the sky: the Soviet Union launched the first man-made satellite, Sputnik 1, in 1957. “One of the criticisms we’ve heard is, ‘Google, you’ve made up a human-made benchmark question – it hasn’t done anything useful yet,'” he said. “The satellite didn’t work much that year. It surrounds the earth and beeps. But this is the beginning of the space age. “
Although this is true, in 1957 people didn’t care. U.S. newspapers have paid great attention to the launch of the satellite, but tend to focus on whether the Russians intend to use their new technology to spy on or drop bombs on the United States. It is only when the United States takes sputnik seriously that people figure out why it is so important – but that’s the afterthought.
Now, the most interesting historical allusion is a blog post by UT’s Aaronson on his own website. Thinking carefully about Google’s research paper and IBM’s response to this, he concluded that it might be right that IBM could match Google’s Sycamore calculations on its best supercomputers.
But he still sees historic google’s achievements. When he mentions the Apollo 11 moon landing in 1969, it’s not that Google did something like this, it’s that quantum supremacy isn’t the same at all: it was clear from the start that quantum supremacy wouldnot be as landmark as it was on the moon, which could be achieved in a flash, and then everyone could know it at any time. It’s more like eliminating measles: it can be achieved, then temporarily impossible, and then done again. Quantum hegemony, as the name suggests, refers to defeating something, traditional computing, which can at least fight back for a period of time.
Human progress is rarely achieved in a minute, and everyone knows it. So waiting for the best before assessing Google’s full impact is the best way to do it – even if we’re not yet in the era of quantum computing.