5 ms·

Just saw the first video, but i found his analogy with mayan priests calculating Venus trajectory correctly, having no clue of the "why" quite disturbing. Ther

by bsaul 11y ago

Just saw the first video, but i found his analogy with mayan priests calculating Venus trajectory correctly, having no clue of the "why" quite disturbing. There is a HUGE difference between knowing "why" and not knowing it. Mayans priests couldn't guess the real "why", because they had no notion of gravity or a correct description of the solar system, but now we do and we know. The fact that quantum effects aren't grasped right now (at least by me :), means a very different thing whether we believe there will be such an explanation one day, with a more complete description of nature, or if such an explanation is inexistant. Which makes me wonder : i've read that there was a proof that the "hidden variable hypothesis" is wrong. Does anyone know of someone explaining that proof in a comprehensible way ? Has this proof been contested by some parts of the scientist working on the field ?

db48x 11y ago
It's called Bell's theorem. Try wikipedia.
yagyu 11y ago
The 'proof' you refer to is likely Bell's theorem and friends: "Bell's Theorem states that the predictions of quantum mechanics cannot be reproduced by any local hidden variable theory.": https://en.wikipedia.org/wiki/Bell_test_experiments https://en.wikipedia.org/wiki/Bell_test_experiments https://en.wikipedia.org/wiki/Bell's_theorem https://en.wikipedia.org/wiki/Bell's_theorem edit: the theorem has been fairly thoroughly experimentally tested and still holds.
m1el 11y ago
> Has this proof been contested by some parts of the scientist working on the field? Here's a story I find interesting. Einstein got his nobel prize for discovering the law of photoelectric effect [1]. This discovery influenced the development of quantum mechanics quite a lot. I'd say that this discovery alone is enough for the theory of QM to emerge. However, Einstein did not like the "probabilistic" nature of QM, "God doesn't play dice with the world."[2], and he argued against it. Albert Einstein along with Boris Podolsky and Nathan Rosen have formulated a paradox[3] which was resolved many years later by Bell's Theorem and corresponding experiments[4]. So yes, there were scientists arguing against the probabilistic nature of QM. > Does anyone know of someone explaining that proof in a comprehensible way? "Comprehensible" explanation in this case mostly depends on the reader. The wiki[4] gives a pretty good explanation of the Bell's Theorem and experiments. [1]: http://www.nobelprize.org/nobel_prizes/physics/laureates/1921/ http://www.nobelprize.org/nobel_prizes/physics/laureates/192... [2]: https://books.google.com.vn/books?id=QXCyjj6T5ZUC https://books.google.com.vn/books?id=QXCyjj6T5ZUC [3]: https://en.wikipedia.org/wiki/EPR_paradox https://en.wikipedia.org/wiki/EPR_paradox [4]: https://en.wikipedia.org/wiki/Bell%27s_theorem#Practical_experiments_testing_Bell.27s_theorem https://en.wikipedia.org/wiki/Bell%27s_theorem#Practical_exp...
- bsaul 11y ago
  Thanks for taking the time. I've read the wikipedia article and think i got the argument better : from what i got, you can't have a "classical" effect underlying QM, because it would mean faster than light transmission of information. And yet, experimentally, Bell's experiment has been done, and it is working the way it is predicted by QM (the famous "spooky action at a distance"). Would that be correct ?
  - mirimir 11y ago
    Yes, comprehension = getting the math. Comprehension <> classical visualization. That's what I eventually accepted, anyway.
qwertyboy 11y ago
I think Feynman's point is that to be a good scientist, you must keep in mind the difference between the map and the territory. You preform observations. You think of explanations ("why"s) and try to fit them in a model. You make testable predictions using your model. You test them. If the experiments support your model, then it's a good and useful model. People will use it to achieve cool and terrible things. But it's still just a model. And you have to be ready at a moment's notice, as soon as the empirical data demands it, to drop your model like a hot potato and start looking for a better one. To do otherwise - to believe that you already know the "why" - is to abandon the scientific method. "Knowing" is the opposite of "learning", and the antonym of scientific progress.
- bsaul 11y ago
  With your description i would say what feynman suggests is abandoning the model altogether. Whenever i hear physicist say "we don't understand it and it doesn't matter", or speak of "spooky action at a distance", you can't say that they've built a "model" of anything that can be proved or disproved by any "better" model. They are, in fact, computing numbers, just like mayan priests, and not even try to put a "god" or a "magic number theory" behind it (as mentioned in Feynman's speech). PS : this kind of debate reminds me a bit of the debate between chomsky and norvig, with norvig saying that numbers and results are all that matters, and chomsky arguing this isn't even science.
  - Retra 11y ago
    Capability matters, and so long as you aren't exploring avenues that could expand your capabilities, you have something more to do.
  - Ollinson 11y ago
    You don't have to understand a model for it to bear fruit. I'm sure you could find plenty of traders on Wall Street with models of the market who would say "we don't understand and it doesn't matter" as the money flows into their account.
  - tagrun 11y ago
andrepd 11y ago
The result is Bell's theorem. The original paper is only 5 pages long, so it should be readable by anyone with a background in QM. The wikipedia article is also very nice and in-depth with discussions of its ramifications. http://www.drchinese.com/David/Bell_Compact.pdf http://www.drchinese.com/David/Bell_Compact.pdf https://en.wikipedia.org/wiki/Bell's_theorem https://en.wikipedia.org/wiki/Bell's_theorem
effie 11y ago
> i've read that there was a proof that the "hidden variable hypothesis" is wrong. Where did you read that? I do not think that's true, since there are infinitely many unexplored theories in which "hidden variables" may be at work. True, there is the Bell theorem, which says that a particular kind of theory that uses local hidden variables cannot give the same results as quantum theory. However, that's very far from a proof that hidden variables hypothesis (which is basically the idea that quantum theory with its psi -based description isn't the final word) is wrong. What's important for a new theory is not to reproduce all quirks of the old theory, but to explain the experiments in a better way. There may be hidden-variable theories Bell did not think about so his theorem does not apply to them, and there may also be non-local hidden variable theories that would be consistent with experiments. I hear Bohm's mechanics is such a non-local theory whose credibility is unaffected by the Bell theorem.

by bsaul 11y ago

db48x 11y ago

It's called Bell's theorem. Try wikipedia.

yagyu 11y ago

The 'proof' you refer to is likely Bell's theorem and friends: "Bell's Theorem states that the predictions of quantum mechanics cannot be reproduced by any local hidden variable theory.": https://en.wikipedia.org/wiki/Bell_test_experiments https://en.wikipedia.org/wiki/Bell_test_experiments https://en.wikipedia.org/wiki/Bell's_theorem https://en.wikipedia.org/wiki/Bell's_theorem edit: the theorem has been fairly thoroughly experimentally tested and still holds.

m1el 11y ago

> Has this proof been contested by some parts of the scientist working on the field? Here's a story I find interesting. Einstein got his nobel prize for discovering the law of photoelectric effect [1]. This discovery influenced the development of quantum mechanics quite a lot. I'd say that this discovery alone is enough for the theory of QM to emerge. However, Einstein did not like the "probabilistic" nature of QM, "God doesn't play dice with the world."[2], and he argued against it. Albert Einstein along with Boris Podolsky and Nathan Rosen have formulated a paradox[3] which was resolved many years later by Bell's Theorem and corresponding experiments[4]. So yes, there were scientists arguing against the probabilistic nature of QM. > Does anyone know of someone explaining that proof in a comprehensible way? "Comprehensible" explanation in this case mostly depends on the reader. The wiki[4] gives a pretty good explanation of the Bell's Theorem and experiments. [1]: http://www.nobelprize.org/nobel_prizes/physics/laureates/1921/ http://www.nobelprize.org/nobel_prizes/physics/laureates/192... [2]: https://books.google.com.vn/books?id=QXCyjj6T5ZUC https://books.google.com.vn/books?id=QXCyjj6T5ZUC [3]: https://en.wikipedia.org/wiki/EPR_paradox https://en.wikipedia.org/wiki/EPR_paradox [4]: https://en.wikipedia.org/wiki/Bell%27s_theorem#Practical_experiments_testing_Bell.27s_theorem https://en.wikipedia.org/wiki/Bell%27s_theorem#Practical_exp...

bsaul 11y ago
Thanks for taking the time. I've read the wikipedia article and think i got the argument better : from what i got, you can't have a "classical" effect underlying QM, because it would mean faster than light transmission of information. And yet, experimentally, Bell's experiment has been done, and it is working the way it is predicted by QM (the famous "spooky action at a distance"). Would that be correct ?
- mirimir 11y ago
  Yes, comprehension = getting the math. Comprehension <> classical visualization. That's what I eventually accepted, anyway.

qwertyboy 11y ago

I think Feynman's point is that to be a good scientist, you must keep in mind the difference between the map and the territory. You preform observations. You think of explanations ("why"s) and try to fit them in a model. You make testable predictions using your model. You test them. If the experiments support your model, then it's a good and useful model. People will use it to achieve cool and terrible things. But it's still just a model. And you have to be ready at a moment's notice, as soon as the empirical data demands it, to drop your model like a hot potato and start looking for a better one. To do otherwise - to believe that you already know the "why" - is to abandon the scientific method. "Knowing" is the opposite of "learning", and the antonym of scientific progress.

bsaul 11y ago
With your description i would say what feynman suggests is abandoning the model altogether. Whenever i hear physicist say "we don't understand it and it doesn't matter", or speak of "spooky action at a distance", you can't say that they've built a "model" of anything that can be proved or disproved by any "better" model. They are, in fact, computing numbers, just like mayan priests, and not even try to put a "god" or a "magic number theory" behind it (as mentioned in Feynman's speech). PS : this kind of debate reminds me a bit of the debate between chomsky and norvig, with norvig saying that numbers and results are all that matters, and chomsky arguing this isn't even science.
- Retra 11y ago
  Capability matters, and so long as you aren't exploring avenues that could expand your capabilities, you have something more to do.
- Ollinson 11y ago
  You don't have to understand a model for it to bear fruit. I'm sure you could find plenty of traders on Wall Street with models of the market who would say "we don't understand and it doesn't matter" as the money flows into their account.
- tagrun 11y ago

andrepd 11y ago

The result is Bell's theorem. The original paper is only 5 pages long, so it should be readable by anyone with a background in QM. The wikipedia article is also very nice and in-depth with discussions of its ramifications. http://www.drchinese.com/David/Bell_Compact.pdf http://www.drchinese.com/David/Bell_Compact.pdf https://en.wikipedia.org/wiki/Bell's_theorem https://en.wikipedia.org/wiki/Bell's_theorem

effie 11y ago

> i've read that there was a proof that the "hidden variable hypothesis" is wrong. Where did you read that? I do not think that's true, since there are infinitely many unexplored theories in which "hidden variables" may be at work. True, there is the Bell theorem, which says that a particular kind of theory that uses local hidden variables cannot give the same results as quantum theory. However, that's very far from a proof that hidden variables hypothesis (which is basically the idea that quantum theory with its psi -based description isn't the final word) is wrong. What's important for a new theory is not to reproduce all quirks of the old theory, but to explain the experiments in a better way. There may be hidden-variable theories Bell did not think about so his theorem does not apply to them, and there may also be non-local hidden variable theories that would be consistent with experiments. I hear Bohm's mechanics is such a non-local theory whose credibility is unaffected by the Bell theorem.