Pages

Monday, September 25, 2017

Microsoft makes play for quantum computer programming

Ars Technica reports:
At its Ignite conference today, Microsoft announced its moves to embrace the next big thing in computing: quantum computing. Later this year, Microsoft will release a new quantum computing programming language, with full Visual Studio integration, along with a quantum computing simulator. With these, developers will be able to both develop and debug quantum programs implementing quantum algorithms.
This is ridiculous. No one will ever have any legitimate use for this.
This ability for qubits to represent multiple values gives quantum computers exponentially more computing power than traditional computers.
Scott Aaronson likes to say that this is wrong. Of course I say that there will never be a quantum speedup.
It will have quite significant memory requirements. The local version will offer up to 32 qubits, but to do this will require 32GB of RAM. Each additional qubit doubles the amount of memory required. The Azure version will scale up to 40 qubits.

Longer term, of course, the ambition is to run on a real quantum computer. Microsoft doesn't have one, yet, but it's working on one.
Wow, that is a lot of memory for a simulator.
One awkward spectre is what happens if someone does manage to build a large quantum computer. Certain kinds of encryption gain their security from the fact that integer factorization ... but if the technology were developed to build quantum computers with a few thousand qubits, these encryption algorithms would become extremely vulnerable. ... That quantum computing future is, fortunately, still likely to be many years off.
That's right, we are fortunate that no one has a quantum computer. It would only cause harm, for the foreseeable future.

Saturday, September 23, 2017

Journals try to deny group differences

Here is a Nature mag editorial:
Science provides no justification for prejudice and discrimination.

Physicians like to say that average patients do not exist. Yet medicine depends on them as clinical trials seek statistical significance in the responses of groups of people. In fact, much of science judges the reliability of an effect on the basis of the size of the group it was measured in. And the larger and more significant the claimed difference, the bigger is the group size required to supply the supporting evidence.

Difference between groups may therefore provide sound scientific evidence. But it’s also a blunt instrument of pseudoscience, and one used to justify actions and policies that condense claimed group differences into tools of prejudice and discrimination against individuals — witness last weekend’s violence by white supremacists in Charlottesville, Virginia, and the controversy over a Google employee’s memo on biological differences in the tastes and abilities of the sexes.

This is not a new phenomenon. But the recent worldwide rise of populist politics is again empowering disturbing opinions about gender and racial differences that seek to misuse science to reduce the status of both groups and individuals in a systematic way.

Science often relies on averages, but it thrives on exceptions. And every individual is a potential exception. As such, it is not political correctness to warn against the selective quoting of research studies to support discrimination against those individuals. It is the most robust and scientific interpretation of the evidence. Good science follows the data, and there is nothing in any data anywhere that can excuse or justify policies that discriminate against the potential of individuals or that systematically reinforce different roles and status in society for people of any gender or ethnic group.
This is really confused. I am not sure that group differences had anything to do the Charlottesville riot, or that there was any violence by white supremacists. I guess Google was using pseudoscience to justify its discriminatory policies, but the point is obscure.

I don't even know what it means to "discriminate against the potential of individuals". How does anything do that?

There certainly is data "that systematically reinforce different roles and status in society".

Nature's SciAm is apologizing for past such remarks:
In 1895 an article in Scientific American — “Woman and the Wheel” — raised the question of whether women should be allowed to ride bicycles for their physical health. After all, the article concluded, the muscular exertion required is quite different from that needed to operate a sewing machine. Just Championnière, an eminent French surgeon who authored the article, answered in the affirmative the question he had posed but hastened to add: “Even when she is perfectly at home on the wheel, she should remember her sex is not intended by nature for violent muscular exertion.... And even when a woman has cautiously prepared herself and has trained for the work, her speed should never be that of an adult man in full muscular vigor.”
We do have separate bicycle races for women; why is that?

That SciAm issue has an article by Cordelia Fine. See here for criticism from a leftist-evolutionist, Jerry Coyne, of her feminist polemic book getting a science book award.

Monday, September 18, 2017

Did Einstein use his own reasoning?

The site Quora gives some answers to this:
Did Einstein get his famous relativity theory from his predecessors (like Galileo, Newton, etc.) or from his own reasoning? ...

The Irish physicist George FitzGerald and the Dutch physicist Hendrik Lorentz were the first to suggest that bodies moving through the ether would contract and that clocks would slow. This shrinking and slowing would be such that everyone would measure the same speed for light no matter how they were moving with respect to the ether, which FitzGerald and Lorentz regarded as a real substance.

But it was a young clerk named Albert Einstein, working in the Swiss Patent Office in Bern, who cut through the ether and solved the speed-of-light problem once and for all. In June 1905 he wrote one of three papers that would establish him as one of the world's leading scientists--and in the process start two conceptual revolutions that changed our understanding of time, space and reality.

In that 1905 paper, Einstein pointed out that because you could not detect whether or not you were moving through the ether, the whole notion of an ether was redundant.
No, Einstein's comments about the aether were essentially the same as what Lorentz published in 1895. Whether the aether is a "real substance" is a philosophical question, and you get different answers even today. Einstein later said that he believed in the aether, but not aether motion.

As a historical matter, Einstein's 1905 paper did not change our understanding of time, space and reality.
If you wanted to live longer, you could keep flying to the east so the speed of the plane added to the earth's rotation.
Einstein had a similar comment in his 1905 paper, but it was wrong because it fails to take gravity into account.
This unease continued through the 1920s and '30s. When Einstein was awarded the Nobel Prize in 1921, the citation was for important -- but by Einstein's standards comparatively minor -- work also carried out in 1905. There was no mention of relativity, which was considered too controversial.
No, there was no controversy about the 1905 special relativity. Special relativity became widely accepted in about 1908 because of theoretical work by Lorentz, Poincare, and Minkowski, and because of experimental work that distinguished it from competing theories.

No one wanted to give Einstein the Nobel prize for special relativity because no one thought that he created the theory or the experimental work.

Some of the other answers mention Lorentz and Poincare as having discovered special relativity.

Friday, September 15, 2017

Video on Bell's Theorem

The YouTube video, Bell's Theorem: The Quantum Venn Diagram Paradox, was trending as popular. It is pretty good, but exaggerates the importance of Bell's theorem.

The basic sleight-of-hand is to define "realism" as assuming that light consists of deterministic particles. That is, not only does light consist of particles, but each particle has a state that pre-determines any experiment that you do. The pre-determination is usually done with hidden variables.

Bell's theorem then implies that we must reject either locality or realism. It sounds very profound when you say it that way, but only because "realism" is defined in such a funny way. Of course light is not just deterministic particles. Light exhibits wave properties. Non-commuting observables like polarization cannot be simultaneously determined. That has been understood for about 90 years.

There is no need to reject locality. Just reject "realism", which means rejecting the stupid hidden variables. That's all.

Thursday, September 14, 2017

tHooft advocates super-determinism

Gerard 't Hooft was the top theoretical physicist behind the Standard Model of elementary particles. He proved that gauge theories were renormalizable, so then everyone worked on gauge theories.

He just posted Free Will in the Theory of Everything:
Today, no prototype, or toy model, of any socalled Theory of Everything exists, because the demands required of such a theory appear to be conflicting. ...

Finally, it seems to be obvious that this solution will give room neither for “Divine Intervention”, nor for “Free Will”, an observation that, all by itself, can be used as a clue. We claim that this reflects on our understanding of the deeper logic underlying quantum mechanics. ...

Is it ‘Superstring Theory’? The problem here is that this theory hinges largely on ‘conjectures’. Typically, it is not understood how most of these conjectures should be proven, and many researchers are more interested in producing more, new conjectures rather than proving old ones, as this seems to be all but impossible. When trying to do so, one discovers that the logical basis of such theories is still quite weak. ...

Is humanity smart enough to fathom the complexities of the laws of Nature? If history can serve as a clue, the answer is: perhaps; we are equipped with brains that have evolved a little bit since we descended from the apes, hardly more than a million years ago, and we have managed to unravel some of Nature’s secrets way beyond what is needed to build our houses, hunt for food, fight off our enemies and copulate. ...

Our conclusion will be that our world may well be super-deterministic, so that, in a formal sense, free will and divine intervention are both outlawed. In daily life, nobody will suffer from the consequences of this.
I guess he is trying to say that we will sill be able to copulate, even if we have no free will.

It is rare to see any intelligent man advocate super-determinism. This is an extreme form of determinism where things like randomized clinical trials are believed to be bogus. That is, God carefully planned the world at the Big Bang in such detail that when you think that you are making random choices for the purpose of doing a controlled experiment, God has actually forced those choices on you so that your experiment will work according to the plan.

Super-determinism is as goofy as Many Worlds theory. It is something you might expect to hear in a philosophy class, where the professor is listing hypothetical tenable beliefs, to which no sane person would subscribe.

I don't want to call anyone insane. If I did, the list would be too long.

'tHooft attempts to detail how Alice and Bob can do a simple polarization experiment, and think that they are making random choices, but their choices are forced by the initial state of the universe, and also by God or some natural conspiracy to make sure that the experimental outcomes do not contradict the theory:
The only way to describe a conceivable model of “what really happens”, is to admit that the two photons emitted by this atom, know in advance what Bob’s and Alice’s settings will be, or that, when doing the experiment, Bob and/or Alice, know something about the photon or about the other observer. Phrased more precisely, the model asserts that the photon’s polarisation is correlated with the filter settings later to be chosen by Alice and Bob. ...

How can our model force the late observer, Alice, or Bob, to choose the correct angles for their polarisation filters? The answer to this question is that we should turn the question around. ... We must accept that the ontological variables in nature are all strongly correlated, because they have a common past. We can only change the filters if we make some modifications in the initial state of the universe.
This argument cannot be refuted. You can believe in it, just as you can believe in zillions of unobservable parallel universes.

These arguments are usually rejected for psychological reasons. Why believe in anything so silly? What could this belief possibly do for you?

How do you reconcile this with common-sense views of the world? How do you interact with others who do not share such eccentric beliefs?

Here is what I am imagining:
Gerard, why did you write this paper?

The initial state of the universe required that I persuade people to not make so many choices, so I had to tell them that their choices are pre-determined to give the outcomes predicted by quantum mechanics.
His error, as with string theorists and other unified field theorists, is that he wants one set of rules from which everything can be deduced:
Rule #6: God must tell his computer what the initial state is.

Again, efficiency and simplicity will demand that the simplest possible choice is made
here. This is an example of Occam’s rule. Perhaps the simplest possible initial state is a
single particle inside an infinitesimally small universe.

Final step:
Rule #7: Combine all these rules into one computer program to calculate
how this universe evolves.

So we’re done. God’s work is finished. Just push the button. However, we reached a level
where our monkey branes are at a loss.
I do not know whether he is trying to make a pun with "monkey branes". Monkeys have brains, while string theory has branes.

Most of the grand unified field theorists are happy with a non-deterministic theory, as they say that Bell's theorem proved non-determinism. But 't Hooft likes the super-determinism loophole to Bell's theorem:
Demand # 1: Our rules must be unambiguous.

At every instant, the rules lead to one single, unambiguous prescription as to what will happen next.

Here, most physicists will already object: What about quantum mechanics? Our favoured theory for the sub-atomic, atomic and molecular interactions dictates that these respond according to chance. The probabilities are dictated precisely by the theory, but there is no single, unambiguous response.

I have three points to be made here. One: This would be a natural demand for our God. As soon as He admits ambiguities in the prescribed motion, he would be thrown back to the position where gigantic amounts of administration is needed: what will be the `actual' events when particles collide? Or alternatively, this God would have to do the administration for infinitely many universes all at once. This would be extremely inefficient, and when you think of it, quite unnecessary. This God would strongly prefer one single outcome for any of His calculations. This, by the way, would also entail that his computer will have to be a classical computer, not a quantum computer, see Refs. [1, 2, 3].

Second point: look at the universe we live in. The ambiguities we have are in the theoretical predictions as to what happens when particles collide. What actually happens is that every particle involved chooses exactly one path. So God's administrator must be using a rule for making up His mind when subatomic particles collide.

Third point: There are ways around this problem. Mathematically, it is quite conceivable that a theory exists that underlies Quantum Mechanics.[4] This theory will only allow single, unambiguous outcomes. The only problem is that, at present, we do not know how to calculate these outcomes. I am aware of the large numbers of baboons around me whose brains have arrived at different conclusions: they proved that hidden variables do not exist. But the theorems applied in these proofs contain small print. It is not taken into account that the particles and all other objects in our aquarium will tend to be strongly correlated. They howl at me that this is `super-determinism', and would lead to `conspiracy'. Yet I see no objections against super-determinism, while `conspiracy' is an ill-defined concept, which only exists in the eyes of the beholder.
I have posted here many times that hidden variable theories have been disproved, so 'tHooft is calling me a baboon.

To summarize, he has a theological belief that an all-knowing all-powerful God created a mathematically deterministic universe. Because our best theories of quantum mechanics seem to allow for free will, at both the level of human choice and electron paths, they must be wrong. There must be some underlying super-deterministic theory.

No, this is wacky stuff. If common sense and human consciousness and experiences convince us that we have free will, and if our best physics theories of the last century leave open the possibility of free will at a fundamental level, and if all efforts to construct a reasonable theory to eliminate free will have failed, then the sensible conclusion is to believe in free will. 't Hooft's view is at odds with everything we know.

Tuesday, September 5, 2017

Paper on Einstein's inventions

A new paper:
Times magazine selected Albert Einstein, the German born Jewish Scientist as the person of the 20th century. Undoubtedly, 20th century was the age of science and Einstein's contributions in unraveling mysteries of nature was unparalleled. However, few are aware that Einstein was also a great inventor. He and his collaborators had patented a wide variety of inventions in several countries.
The article gives a nice accounts of Einstein's invenstions and patents.

The account of Einstein's life includes the usual myths, such as this account of his most famous paper:
3. On the electrodynamics of moving bodies, Annalen der Physik 17 (1905) 891-921.

This is the first paper on special relativity. It drastically altered the century old man’s idea about space and time. In Newtonian mechanics they have separate identities. In Einstein's relativity, space and time are not separate entities rather one entity called space-time continuum. Continuum because in our experience there is no void in space or time. Identification of space-time as an entity required that bodies moving with velocity of light or near need a different mechanics, relativistic mechanics rather than the Newtonian mechanics. Intermingling of space and time produces few surprises, e.g. a moving clock tick slowly (time dilation), a moving rod contracts (length contraction),
strange laws of velocity addition etc.
Almost all of this is wrong. It was not the first paper on special relativity, as it has little in conceptual advance from Lorentz's 1895 paper. It did not combine space and time. Length contraction was proposed by FitzGerald in 1889, and Larmor discussed time dilation in about 1999 1899. Poincare had the velocity addition formula a couple of months ahead of Einstein.

The author does correctly point out that nobody thought that Einstein's 1905 paper was any big deal at the time. It was considered just an explanation of Lorentz's theory, and special relativity became popular are a result of Minkowski developing Poincare's ideas. Einstein's paper had almost no influence on the development and acceptance of special relativity.