Saturday, May 11, 2019

NY Times joins the time reversal hype

The NY Times reports:
Using an IBM quantum computer, they managed to undo the aging of a single, simulated elementary particle by one millionth of a second.
...
The algorithm almost always worked. It succeeded in returning the qubits to their youthful states 85 percent of the time when the calculation involved two qubits, but only half the time when three qubits were used.
This result as already been overhyped elsewhere, and Scott Aaronson had debunked it:
Incredibly, the time-reversal claim has now gotten uncritical attention in Newsweek, Discover, Cosmopolitan, my Facebook feed, and elsewhere — hence this blog post, which has basically no content except “the claim to have ‘reversed time,’ by running a simulation backwards, is exactly as true and as earth-shattering as a layperson might think it is.”
Dennis Overbye gives this explanation of time reversal:
On paper, the basic laws of physics are reversible; they work mathematically whether time is running forward or backward. But if time is just another dimension of space-time, as Einstein said, it’s a strange one-way dimension. In the real world we can climb out of the subway and turn left or right, but we don’t have the choice of going forward or back in time. We are always headed toward the future. ...

But the arrow of time takes its direction not only from big numbers. According to quantum theory, that paradoxical body of rules governing the subatomic universe, not even a single particle can reverse its own course through time. ...

The wave function extends throughout space and time. The law describing its evolution, known as the Schrödinger equation, after Austrian physicist Erwin Schrödinger​, is equally valid running forward or backward. But getting a wave function to go in reverse is no small trick. ...

“It remains to be seen,” the team wrote in their paper posted online in February, “whether the irreversibility of time is a fundamental law of nature or whether, on the contrary, it might be circumvented.”
This is what a lot of physicists say, but it doesn't make much sense. How is it that the basic laws of physics are reversible, but you cannot actually reverse anything? I think it is more accurate to say that the basic laws of physics are not time reversible.

He gives this explanation of quantum computers:
Unlike regular computers, which process a series of zeros and ones, or bits, quantum computers are made of so-called qubits, each of which can be zero and one at the same time. A quantum computer can perform thousands or millions of calculations simultaneously, so long as nobody looks to see what the answer is until the end.

Many of the largest tech companies, including Google, Microsoft and IBM, are racing to build such machines, which eventually could solve problems that regular computers can’t, such as breaking currently unbreakable cryptographic codes. Some scientists argue that nature itself is a quantum computer, and that the greatest utility of such a computer will be in simulating and exploring the paradoxes of quantum weirdness. ...

The IBM computer they used represents a baby step in the direction of what theorists call “quantum supremacy.”
Aaronson would say that this explanation is wrong, because the power of quantum computers comes from probabilities being negative, not qubits being zero and one at the same time.

I think that it is more accurate to say that quantum computers get their power from time reversible operations. All of the quantum algorithms depend on qubits, and on electronic gates that do time reversible operations on qubits.

But we can't do any time reversible operations. Not significantly. According to the above paper, a state-of-the-art quantum computer can do it for a millionth of a second on two qubits 85% of the time.

This is like saying someone invented a perpetual motion machine that works for a millionth of a second 85% of the time.

Quantum supremacy has not been achieved. Everyone now acknowledges this. The above paper is supposed to be a baby step toward that end. On the contrary, it shows how difficult the task is.

2 comments:

  1. No, Roger, I think you got it wrong when you said:

    ``I think that it is more accurate to say that quantum computers get their power from time reversible operations. All of the quantum algorithms depend on qubits, and on electronic gates that do time reversible operations on qubits.''

    No, they don't. What they instead count on is: the linearity of QM. Never forget this part. Grant them the ``inherent'' etc. linearity of QM, and they go wild---because you granted them it. They never did concern themselves with the qubits in the first place (i.e., way back in the 1950s and earlier, in the late 1920s).

    I am writing this rather directly, and I will let you know, after a drink and before a dinner. But I also do think that I am right on money on this one.

    Let them live in their ``linear'' world, erected for them by their own mathematicians. Just know that the real world isn't that.

    More, later. (And sorry for keeping you waiting, but in all fairness, I guess I made my point when I wrote that Outline document and posted it at iMechanica and all... )

    But yes, a more detailed response is necessary, and I shall post it once I land a job in Indian IT industry. A very well-paying one, in fact.

    Bye for now.

    Best,

    --Ajit


    ReplyDelete
  2. It all falls apart in the first sentence:

    "Using an IBM quantum computer, they managed to undo the aging of a single, simulated elementary particle by one millionth of a second."

    What a complete and utter waste of electricity. A single "simulated elementary particle"... does not exist in actual time in any way, IT ISN'T REAL, ergo, it can't age or de-age except as a fiction. Here's another profound revelation: pressing the play followed by the rewind button on a VCR doesn't make 'simulated' time go backwards either. Who knew.

    ReplyDelete