4 ms·

There is a "sermon" that is given at the beginning of almost every QM intro lecture, and as it was given by Feynman more than a half century ago, it went someth

by tbabb 9y ago

There is a "sermon" that is given at the beginning of almost every QM intro lecture, and as it was given by Feynman more than a half century ago, it went something like: > "Quantum mechanics is weird. Particles are not like baseballs; leave your classical intuitions behind." Which is all well and good. But it has subtly shifted to: > "Quantum mechanics is weird. Our human brains are not wired to understand it intuitively, so don't try, because you will fail and get the wrong answer. Just turn the math crank and get results." And IMHO that attitude is both (a) flatly wrong and (b) mind poison. With regards to (a), the human brain is a turning machine-- it can simulate any rules. Learn the rules and work with them enough, and they will become intuitive. If you can learn DiffEQ or Fourier transforms, you can learn QM. Actually, the rules for the "spooky bits" are probably simpler than both of those two things. For (b), all these undergraduates grow up into research physicists and still believe the creed they were solemnly lectured on Day One about how to think about quantum mechanics (which is to NOT think about it). As such, they are all very bad scientists because they have stopped wondering about what makes it all tick. Case in point: A good fraction of scientists still believe in "wavefunction collapse", which blows my mind, because it's so self-evidently rejectable[1]. Science went from a Plum Pudding atom to quantum field theory in about 30 years. It's been nearly a century since then, and we don't really understand things much more deeply, and I think this is because we are too busy banging particles together and not asking "why" enough. I (boldly) posit that if undergrads didn't get the "don't ask" lecture, quantum physics would be decades ahead of where it is today. We can understand, and we should try, and I would love it if undergrads were told that instead. The kind of thing in this article is what more serious physicists should be doing. [1] This is a controversial statement, but that's kind of my point. Explaining Why Not Copenhagen is probably a topic for a separate diatribe.

sgt101 9y ago
I'm certain that human brains aren't Turing machines; we know that the entire set of human brains halts! Also human brains do not have infinite tapes or an approximate equivalent. So it can't be said that it can simulate any rules.
- tbabb 9y ago
  > ackchyually.jpg ;) I think you got my point, though? The rules of QM are not a busy beaver function. They are simple and they can be learned because the human mind is flexible. Yes, manipulating the algebra can be very complicated indeed, but that is because we are analyzing how the simple rules carry out and looking for an exact answer. It's like playing checkers vs. solving checkers; or learning the rules of Conway's Life vs. actually running a simulation. We can understand the rules and gain meaningful intuitions about both. If we want an exact answer it gets complicated.
  - Lambdanaut 9y ago
    If I were to be bold enough to rephrase what you so elegantly said. We're looking at the universe as the pre-copricornians were. Seeing chaos and logging the data, believing that the order behind the masterpiece was all locked away in God's noggin, hidden from man's grasp. Instead we should be looking at the universe as Copernicus did. Re-frame our base assumptions. Among all this chaotic data is a simpler pattern. It may make no sense whatsoever to our human assumptions about reality, but if we are to know reality, then we must evolve within reality.
  - sgt101 9y ago
    I looked the jpg up, and you are a bit mean!!! I am trying not to cry real tears ;) I get the busy beavers thing, humans can't do that, nor computers - it says nothing much! But... are there things that are not computable that aren't formalised either - for example the choosing to determine when shifting system is needed to prove something that may or may not be proved in a particular system?
    - tbabb 9y ago
      Sorry, maybe that was a bit harsh. :) [Recurse Center](https://www.recurse.com/manual https://www.recurse.com/manual) (explained in the section 'social rules') has a social rule against "well-actuallys" which is intended to discourage nitpicking. If someone does it, you just point out "hey, that's a 'well-actually'", and they go 'oh sorry' and you move on. That was my way of pointing out a well-actually. :) > the choosing to determine when shifting system is needed to prove something that may or may not be proved in a particular system? Sorry, I am having a hard time parsing this. What do you mean?
mikekchar 9y ago
A very long time ago, I went to university and took CS. I found it too easy, so I switched to Physics. I found that too hard, so I switched back to CS ;-). I didn't get "the lecture". In fact, just the opposite. We were taught physics and then told that it was up to us to learn whatever math we needed to help us. Not being particularly fluent in math, I just couldn't keep up at the time. One of the best things about my brief foray into Physics was that it taught me the importance of learning math -- not as a prerequisite for something, but as a tool. Having said that, I think you are wrong in your supposition. QM is hard to reason about because it doesn't manifest itself particularly visibly in our everyday lives. The framework that you have to work with comes from your own imagination (and mathematical models). There is no real problem with that, except that if you get the axioms wrong, you will necessarily get the conclusions wrong as well. There are an infinite number of ways that it could work and virtually all of them are incorrect. With classical physics, you can reason based on your lifetime of observations of everyday objects. Of course, we still get it wrong. The more we abstract away from the ordinary, the easier it is to jump to strange conclusions. QM is abstracted so far away that it is incredibly hard to test anything intuitively. That's why math is attractive. It gives you a logical system for testing your ideas. Math is just a language, though. It's a way of expressing an idea. If you have an idea and express it in English, it is no better or worse than expressing it in Math, and no better or worse than expressing it in pictures (Expressing it as short presentation with hand puppets, is obviously the best way, but that should go without saying). Similarly, if you are working through a problem by talking to yourself, or doing math, or drawing diagrams, or imagining scenarios with hand puppets -- it's all equivalent. The representation of the thing is not the thing. And therein lies the problem. Many people do not really understand things well. They think they do because they can picture a representation of it. Let me give you an example. I once showed a colleague one of those ring-within-a-ring magic puzzles where you have to unlink the rings. I showed him how to do it. I got him to do it 4 or 5 times and then I asked him, "Do you understand how this works now"? He said, "Yes, it is very simple". Then I took the magic rings away and gave him some straight wires. "Please build the magic rings", I said. Of course, he couldn't because he didn't understand them. We see this problem all the time in programming. People think they understand the problem because they can write code. The code that they write gives them an illusion that they understand the problem. But the code is not the problem. They start focusing on the code and forgetting the problem. Eventually they have written a huge amount of code and they have not solved the problem. You're just noticing the same phenomenon in Physics. It doesn't matter if a person builds a model with math, or diagrams, or English description, or hand puppets -- most people will start to focus on their language and their tools and become dissociated from the problem. They create their own religion around the model into which they have poured their creative energy. That becomes their reality and they have no idea that it doesn't match a more objective reality. The more abstract the system, the easier it is to tumble into fantasy land (which is why programmers are so susceptible IMHO). The lecture I wish people would get is: "You are wrong. Now tell me why. If you don't know why, find out." That would help people a lot better, I think ;-)

by tbabb 9y ago

sgt101 9y ago

I'm certain that human brains aren't Turing machines; we know that the entire set of human brains halts! Also human brains do not have infinite tapes or an approximate equivalent. So it can't be said that it can simulate any rules.

tbabb 9y ago
> ackchyually.jpg ;) I think you got my point, though? The rules of QM are not a busy beaver function. They are simple and they can be learned because the human mind is flexible. Yes, manipulating the algebra can be very complicated indeed, but that is because we are analyzing how the simple rules carry out and looking for an exact answer. It's like playing checkers vs. solving checkers; or learning the rules of Conway's Life vs. actually running a simulation. We can understand the rules and gain meaningful intuitions about both. If we want an exact answer it gets complicated.
- Lambdanaut 9y ago
  If I were to be bold enough to rephrase what you so elegantly said. We're looking at the universe as the pre-copricornians were. Seeing chaos and logging the data, believing that the order behind the masterpiece was all locked away in God's noggin, hidden from man's grasp. Instead we should be looking at the universe as Copernicus did. Re-frame our base assumptions. Among all this chaotic data is a simpler pattern. It may make no sense whatsoever to our human assumptions about reality, but if we are to know reality, then we must evolve within reality.
- sgt101 9y ago
  I looked the jpg up, and you are a bit mean!!! I am trying not to cry real tears ;) I get the busy beavers thing, humans can't do that, nor computers - it says nothing much! But... are there things that are not computable that aren't formalised either - for example the choosing to determine when shifting system is needed to prove something that may or may not be proved in a particular system?
  - tbabb 9y ago
    Sorry, maybe that was a bit harsh. :) [Recurse Center](https://www.recurse.com/manual https://www.recurse.com/manual) (explained in the section 'social rules') has a social rule against "well-actuallys" which is intended to discourage nitpicking. If someone does it, you just point out "hey, that's a 'well-actually'", and they go 'oh sorry' and you move on. That was my way of pointing out a well-actually. :) > the choosing to determine when shifting system is needed to prove something that may or may not be proved in a particular system? Sorry, I am having a hard time parsing this. What do you mean?

mikekchar 9y ago

A very long time ago, I went to university and took CS. I found it too easy, so I switched to Physics. I found that too hard, so I switched back to CS ;-). I didn't get "the lecture". In fact, just the opposite. We were taught physics and then told that it was up to us to learn whatever math we needed to help us. Not being particularly fluent in math, I just couldn't keep up at the time. One of the best things about my brief foray into Physics was that it taught me the importance of learning math -- not as a prerequisite for something, but as a tool. Having said that, I think you are wrong in your supposition. QM is hard to reason about because it doesn't manifest itself particularly visibly in our everyday lives. The framework that you have to work with comes from your own imagination (and mathematical models). There is no real problem with that, except that if you get the axioms wrong, you will necessarily get the conclusions wrong as well. There are an infinite number of ways that it could work and virtually all of them are incorrect. With classical physics, you can reason based on your lifetime of observations of everyday objects. Of course, we still get it wrong. The more we abstract away from the ordinary, the easier it is to jump to strange conclusions. QM is abstracted so far away that it is incredibly hard to test anything intuitively. That's why math is attractive. It gives you a logical system for testing your ideas. Math is just a language, though. It's a way of expressing an idea. If you have an idea and express it in English, it is no better or worse than expressing it in Math, and no better or worse than expressing it in pictures (Expressing it as short presentation with hand puppets, is obviously the best way, but that should go without saying). Similarly, if you are working through a problem by talking to yourself, or doing math, or drawing diagrams, or imagining scenarios with hand puppets -- it's all equivalent. The representation of the thing is not the thing. And therein lies the problem. Many people do not really understand things well. They think they do because they can picture a representation of it. Let me give you an example. I once showed a colleague one of those ring-within-a-ring magic puzzles where you have to unlink the rings. I showed him how to do it. I got him to do it 4 or 5 times and then I asked him, "Do you understand how this works now"? He said, "Yes, it is very simple". Then I took the magic rings away and gave him some straight wires. "Please build the magic rings", I said. Of course, he couldn't because he didn't understand them. We see this problem all the time in programming. People think they understand the problem because they can write code. The code that they write gives them an illusion that they understand the problem. But the code is not the problem. They start focusing on the code and forgetting the problem. Eventually they have written a huge amount of code and they have not solved the problem. You're just noticing the same phenomenon in Physics. It doesn't matter if a person builds a model with math, or diagrams, or English description, or hand puppets -- most people will start to focus on their language and their tools and become dissociated from the problem. They create their own religion around the model into which they have poured their creative energy. That becomes their reality and they have no idea that it doesn't match a more objective reality. The more abstract the system, the easier it is to tumble into fantasy land (which is why programmers are so susceptible IMHO). The lecture I wish people would get is: "You are wrong. Now tell me why. If you don't know why, find out." That would help people a lot better, I think ;-)