I refer to papers announcing that Google has achieve quantum supremacy. You can find links to the two papers in the comments on Scott Aaronson's blog.
I am not conceding defeat yet. First, Google has withdrawn the papers, and refuses to say whether it has a achieved a breakthru or not. Second, outside experts like Aaronson have apparently been briefed on the work, but refuse to comment on it. And those who do comment are not positive:
However, the significance of Google’s announcement was disputed by at least one competitor. Speaking to the FT, IBM’s head of research Dario Gil said that Google’s claim to have achieved quantum supremacy is “just plain wrong.” Gil said that Google’s system is a specialized piece of hardware designed to solve a single problem, and falls short of being a general-purpose computer, unlike IBM’s own work.Gil Kalai says that the Google and IBM results are impressive, but he still believes that quantum supremacy is impossible.
So it may not be what it appears to be.
Aaronson had been sworn to secrecy, and now considers the Google work a vindication of his ideas. He stops short of saying that it proves quantum supremacy, but he implies that the quantum supremacy skeptics have been checkmated.
Probably Google is eager to make a big splash about this, but is getting the paper published in Science or Nature, and those journals do not like to be scooped. The secrecy also helps suppress criticism, because the critics usually don't know enuf about the work when the reporters call.
The paper claims quantum supremacy on the basis of doing a computation that would have been prohibitive on a classical supercomputer.
That sounds great, but since the computation was not replicated, how do we know that it was done correctly?
A universal quantum simulator is a quantum computer proposed by Yuri Manin in 1980 and Richard Feynman in 1982. Feynman showed that a classical Turing machine would experience an exponential slowdown when simulating quantum phenomena, while his hypothetical universal quantum simulator would not. David Deutsch in 1985, took the ideas further and described a universal quantum computer.So we have known since 1982 that simulating a quantum experiment on a classical computer can take exponential time.
At first glance, it appears that Google has only verified that. It did some silly quantum experiment, and then showed that the obvious classical simulation of it would take exponential time.
Is that all Google has done? I haven't read the paper yet, so I don't know. It is hard to believe that Google would claim quantum supremacy if that is all it is. And Google has not officially claimed it yet.
The paper says:
The benchmark task we demonstrate has an immediate application in generating certifiable random numbers ;Really? Is that all? It would be more impressive if they actually computed something.