I guess it depends on what you mean by 'understanding'. I personally feel that you haven't really grasped the math if you've never used it to solve an actual problem - textbook will do, but ideally something not designed for solvability. There's a certain hard-to-convey Fingerspitzggefühl, intuition, feel-for-the-problem-domain - whatever you want to call it - that comes only with long practice. It's similar to debugging computer programs, which is a somewhat separate skill from writing them; I talk about it in some detail in this podcast and these slides.

That said, I would say you can get quite a good overview without any math; you can understand physics in the same sense I understand evolutionary biology - I know the basic principles but not the details that make up the daily work of scientists in the field.

Those two questions are completely unrelated. Popular physics books just aren't trying to convey any physics. That is their handicap, not the math. Greene could teach you a lot of physics without using math, if he tried. But there's no audience for such books.

Eliezer's quantum physics sequence impressed me with its attempt to avoid math, but it seems to have failed pretty badly.

I don't think it's going to be practical this century. The difficulty is that the same properties that let you cut the latency are the ones that make the detectors huge: Neutrinos go right through the Earth, and also right through your detector. There's really no way around this short of building the detector from unobtainium, because neutrinos interact only through the weak force, and there's a reason it's called 'weak'. The probability of a neutrino interacting with any given five meters of your detector material is really tiny, so you need a lot of them, or a huge and very dense detector, or both. Then, you can't modulate the beam; it's not an electromagnetic wave, there's no frequency or amplitude. (Well, to be strictly accurate, there is, in that neutrinos are quantum particles and therefore of course are also waves, as it were. But the relevant wavelength is so small that it's not useful; you can't build an antenna for it. For engineering purposes you really cannot model it as anything but a burst of particles, which has intensity but not amplitude.) So you're limited to Morse code or similar. Hence you lose in bandwidth what you gain in latency. Additionally, neutrinos are h... (read more)

There isn't a unified "string theory controversy".

The battle-tested part of fundamental physics consists of one big intricate quantum field theory (the standard model, with all the quarks, leptons etc) and one non-quantum theory of gravity (general relativity). To go deeper, one wishes to explain the properties of the standard model (why those particles and those forces, why various "accidental symmetries" etc), and also to find a quantum theory of gravity. String theory is supposed to do both of these, but it also gets attacked on both fronts.

Rather than producing a unique prediction for the geometry of the extra dimensions, leading to unique and thus sharply falsifiable predictions for the particles and forces, present-day string theory can be defined on an enormous, possibly infinite number of backgrounds. And even with this enormous range of vacua to choose from, it's still considered an achievement just to find something with a qualitative resemblance to the standard model. Computing e.g. the exact mass of the "electron" in one of these stringy standard models is still out of reach.

Here is a random example of a relatively recent work of string ... (read more)

I don't do formal Bayes or Kolmogorov on a daily basis; in particle physics Bayes usually appears in deriving confidence limits. Still, I'm reasonably familiar with the formalism. As for string theory, my jest in the OP is quite accurate: I dunno nuffin'. I do have some friends who do string-theoretical calculations, but I've never been able to shake out an answer to the question of what, exactly, they're calculating. My basic view of string theory has remained unchanged for several years: Come back when you have experimental predictions in an energy or luminosity range we'll actually reach in the next decade or two. Kthxbye.

The controversy is, I suppose, that there's a bunch of very excited theorists who have found all these problems they can sic their grad students on, problems which are hard enough to be interesting but still solvable in a few years of work; but they haven't found any way of making, y'know, actual predictions of what will happen in current or planned experiments if their theory is correct. So the question is, is this a waste of perfectly good brains that ought to be doing something useful? The answer seems to me to be a value judgement, so I don't think you can resolve it at a glance.

What on earth is the String Theory controversy about, and is it resolvable at a glance like QM's MWI?

I wonder how you resolve the MWI "at a glance". There are strong opinions on both sides, and no convincing (to the other side) argument to resolve the disagreement. (This statement is an indisputable experimental fact.) If you mean that you are convinced by the arguments from your own camp, then I doubt that it counts as a resolution.

Also, the Occam's razor is nearly always used by physicists informally, not calculationally (partly because Kolmogorov complexity is not computable).

As for the string theory, I don't know how to use Bayes to evaluate it. On one hand, this model gives some hope of eventually finding something workable, since it provided a number of tantalizing hints, such as the holographic principle and various dualities. On the other hand, every testable prediction it has ever made has been successfully falsified. Unfortunately, there are few other competing theories. My guess is that if something better comes along, it will yield the string theory in some approximation.

Proton decay has not been observed, but even if it happens, it needn't be an obstacle to life, as such. For humans in anything remotely like our present form you need protons, but not for life in general. Entropy, however, is a problem. All life depends on having an energy gradient of some form or other; in our case, basically the difference between the temperature of the Sun and that of interstellar space. Now, second thermo can be stated as "All energy gradients decrease over a sufficiently long time"; so eventually, for any given form of life, the gradient it works off is no longer sharp enough to support it. However, what you can do is to constantly redesign life so that it will be able to live off the gradients that will exist in the next epoch. You would be trying to run the amount and speed of life down on an asymptotic curve that was nevertheless just slightly faster than the curve towards total entropy. At every epoch you would be shedding life and complexity; your civilisation (or ecology) would be growing constantly smaller, which is of course a rather alien thing for twenty-first century Westerners to consider. However, the idea is that by growing constantly s... (read more)

I took a physics course in my International Baccalaureate program in high school - if you're not familiar with IB, it's sort of the European version of AP - and it really resonated with me. There's just a lot of cool stuff in physics; we did things like building electric motors using these ancient military-surplus magnets that had once been installed in radars for coastal fortresses. Then when I went on to college, I took some math courses and some physics courses, and found I liked the physics better. In the summer of 2003 (I think) I went to CERN as a summer student, and had an absolute blast even though the actual work I was doing wasn't so very advanced. (I wrote a C interface to an ancient Fortran simulation program that had been kicking around since it was literally on punchcards. Of course the scientist who assigned me the task could have done it himself in a week, while it took me all summer, but that saved him a week and taught me some real coding, so it was a good deal for both of us.) So I sort of followed the path of least resistance from that point. I ended up doing my Master's degree on BaBar data. Then for my PhD I wanted to do it outside Norway, so it was basically ... (read more)

Well, if we do find it there are presumably Nobel prizes to be handed out to whoever developed the correct variant. If we don't, I most earnestly hope we find something else, so someone else gets to go to Stockholm. In either case I expect the grant money will keep flowing; there are always precision measurements to be made. Or were you asking about practical applications? I can't say I see any, but then they always do seem to come as a surprise.

'Observation' is a shorthand (for historical reasons) for 'interaction with a different system', for example a detector or a human; but a rock will do as well. I would actually suggest you read the Quantum Mechanics Sequence on this point, Eliezer's explanation is quite good.