HNHacker News
TopNewBestAskShowJobs

knzhou

5,522 karma · joined September 2, 2019

I'm a physics graduate student. You can also find me on:

Physics StackExchange: https://physics.stackexchange.com/users/83398/knzhou

Personal website: https://knzhou.github.io

submissionscomments
knzhou··on Let's hear scientists with different Covid-19 views, not attack them
> what specifically leads to such bold conclusions (e.g. "owe an apology", "wasted time and effort", "statistical error", "screwed up", etc.)?

They used a test with a false positive rate of about 2%, and found a positive rate of about 2%. It is impossible to use this information to establish a lower bound on the number infected. However, by screwing up the computation of the confidence interval, they claimed a huge lower bound.

knzhou··on How much power can we expect from rainfall?
To evaluate stuff like this, you just need basic physics knowledge (enough to do dimensional analysis with) and estimation skills. One fun read along these lines is "The Art of Insight in Science and Engineering".

For stuff like the bad social science you mentioned, often you can catch it if you know basic statistics along with the common tricks people use to p-hack. See, for example, "The garden of forking paths" by Gelman.

Even if the statistical analysis is not given explicitly, or the issue is more subtle, for social science that touches on everyday life, you can usually tell if it'll hold up just with common sense. Try yourself using this quiz: https://80000hours.org/psychology-replication-quiz/

knzhou··on Ask HN: What scientific phenomenon do you wish someone would explain better?
I answered this here: https://physics.stackexchange.com/questions/287101/where-doe...

The basic answer is that the extra energy that goes to the rocket comes from harvesting the kinetic energy that the fuel itself had by virtue of being in the moving rocket.

knzhou··on Ask HN: What scientific phenomenon do you wish someone would explain better?
"Tidal effects" are defined in terms of having different gravitational fields in one place than another (i.e. the tidal bulge near to the moon occurs because the moon's field is stronger there).
knzhou··on Ask HN: What scientific phenomenon do you wish someone would explain better?
You mean dark matter annihilation? There are some hints: an excess of gamma rays detected by Fermi-LAT, an excess of X-rays detected by Chandra, excess ionization detected by EDGES, and excess positrons detected by AMS-02. However, all of these anomalies are contentious and could well be explained by systematic effects or unknown astrophysical sources.

If none of these hold up, it could be that dark matter just doesn't annihilate. That's not weird or anything -- I mean, your room is full of matter right now, and it's not annihilating itself either.

knzhou··on Ask HN: What scientific phenomenon do you wish someone would explain better?
This entertaining article lists what happened to the early LSD researchers: https://slatestarcodex.com/2016/04/28/why-were-early-psyched...
knzhou··on Ask HN: What scientific phenomenon do you wish someone would explain better?
> If light is emitted at a constant wavelength independent of the stretching of the universe, doesn't that imply light is traveling through a higher spatial dimension, otherwise the emitter itself would be stretched with the universe and we'd never be able to observe differences in the speed of light? If I understand this paper, once light is emitted, it's "stuck" to space and will stretch along with it. But if the emitter wavelength stays constant doesn't that imply it's waving through a higher dimension?

I'm not totally sure what you mean by a higher dimension. The properties of the emitter (which is, e.g. a laser cavity) aren't affected by the gravitational wave because the emitter is a rigid body, which doesn't get stretched. (It's the same thing as described here: https://news.ycombinator.com/item?id=22990753 ) So it puts out light of a given frequency.

By contrast, LIGO is not a rigid body, because the mirrors at the ends of the arms hang freely, hence allowing gravitational waves to change the distance between them.

> What's baffling to me is everyone who has tried to explain the LIGO detector doesn't even realize this question exists. I've independently thought this question and when people start explaining LIGO to me, and I take the time to spell out the question, they realize they don't understand LIGO either.

Yup, it generally is the case in physics that over 95% of people who claim they can explain any given thing don't actually understand it! But the professionals are aware. I even know a LIGO guy who goes to popular talks armed with a pile of copies of the paper I linked.

knzhou··on Ask HN: What scientific phenomenon do you wish someone would explain better?
I guess, but it depends on how one parses the poster's setup. You're correct that for particles interacting under a 1/r^2 force, the energy turns out finite in quantum mechanics. My comment was referring to the fact that once you quantize the field that gives rise to that force, the infinities return, but for a different reason.
knzhou··on Ask HN: What scientific phenomenon do you wish someone would explain better?
I'd say that's a perfectly fine explanation. And in fact the light does get stretched. For sufficiently low frequencies, that's fine because there's enough time for new (unstretched) light to enter and exit the apparatus. At higher frequencies this causes the sensitivity of LIGO to fall.

Part of the popular confusion around how LIGO works is the freedom in coordinates: there are different, perfectly good definitions of space and time you can use, and the explanation sounds different in each one. So people can get them mixed up. For example, my previous paragraph makes sense in "transverse traceless gauge", but not in others.

I'm not sure what GP was referring to with "higher spatial dimensions".

knzhou··on Ask HN: What scientific phenomenon do you wish someone would explain better?
This is definitely related, though not the whole story. Quantization did get rid of some infinities, but as GP kind of states, it also introduced others. My comment focuses on what we do for those.
knzhou··on Ask HN: What scientific phenomenon do you wish someone would explain better?
The issue with these animations is that they're getting rid of all the thermal noise. In reality, single proteins are flying around the whole length of the cell many times a second, just from their thermal motion. And when processes like DNA transcription happen, they're not like a regular conveyor belt -- a fraction of the time the machine will even accidentally run steps in reverse! However, if any of this were shown, the animations would become impossible to understand.
knzhou··on Ask HN: What scientific phenomenon do you wish someone would explain better?
The official way this is handled is called renormalization.

Basically, we just declare that we have no idea what is going on at such short distances, and put in some regulator by hand to get rid of the infinity. One very crude regulator (which nobody uses, but which is suitable for demonstration) would just be to say that particles are simply not allowed to get any closer than some fixed tiny distance.

But what about the effects that occur when particles actually do get that close? Well, in most theories, whatever is happening can be parametrized in terms of a few numbers (e.g. it could shift the observed mass of the particles, or their charge, etc.). Our ignorance of what is actually happening prevents us from computing these numbers from first principles. But we can still make scientific progress, because we can treat them as free parameters and measure them -- and after that measurement, we can use the values to crank out perfectly well-defined predictions.

Repeating this process through several layers was crucial to building the Standard Model, which currently has about 20 free parameters.

knzhou··on Ask HN: What scientific phenomenon do you wish someone would explain better?
> If space is expanding why aren’t the radii of fundamental particles and their orbits and molecules also expanding? And if that we’re the case we couldn’t notice space expanding.

> Does space only expand somewhere else? Only between me and the Andromeda galaxy, and not _within_ me and the Andromeda galaxy? How would it know to do that?

If you start with expanding space in general relativity, and then carefully take the limit where you get back to Newtonian gravity, then it just corresponds to a classical force, specifically a very tiny force that weakly pulls everything apart, growing with distance.

This doesn't expand small objects, because they're rigid. It's the same reason that I can't make my laptop get bigger by gently pulling on the ends. On the other hand, it would pull apart two independent, noninteracting objects (such as the Milky Way and Andromeda).

knzhou··on Ask HN: What scientific phenomenon do you wish someone would explain better?
Well, if you're uncomfortable with LIGO, you're in good company. Once a whole decade went by in physics where people couldn't agree if it would work in principle or not. And it is true that a lot of common explanations are bad (e.g. "it's just a ruler" is not complete by itself because "why doesn't the light get stretched too?"). Nonetheless, today we have a variety of independent explanations.

Maybe you'll find this paper helpful: https://aapt.scitation.org/doi/10.1119/1.18578

knzhou··on Ask HN: What scientific phenomenon do you wish someone would explain better?
> If your explanation is that gravitational waves stretch and squash spacetime so light takes different amount of time to bounce from an emitter back to a measurer, then you don't have even a slight understanding of how it works.

Well, I'll bite. I'm a physicist and I understand LIGO. What's your alternative explanation?

knzhou··on Ask HN: What scientific phenomenon do you wish someone would explain better?
You're right that this is glossed over in popular explanations, but the point you make is exactly the starting point for all formal courses and textbooks.

The first part of the argument is that for single point particles falling, the effect of gravity is the same for all particles. This suggests that we should model gravity as something intrinsic to spacetime itself, rather than as a field living on top of spacetime, which could couple to different particles with different strengths.

The second part of the argument, which is what you point out, is that gravity can have nontrivial tidal effects. (This had better be true, because if all gravitational effects were just equivalent to a trivial uniform acceleration, then it would be so boring that we wouldn't need a theory of gravity at all!) This suggests that whatever property of spacetime we use to model gravity, it should reduce in the Newtonian limit to something that looks like a tidal effect, i.e. a gradient of the Newtonian gravitational field. That leads directly to the idea of describing gravity as the curvature of spacetime.

So both parts of the argument give important information (both historically and pedagogically). Both parts are typically presented in good courses, but only the first half makes it to the popular explanations, probably out of simplification.

knzhou··on Tiers of answers to half-baked questions
> To be clear, you're saying fields are essentially virtual mathematical notions from which emerges actual stuff?

Perhaps, but I think that's a bit unfair. There's no dividing line between "mathematical notions" and "actual stuff". For example, when you poke a bowl of jello, it springs back, so it feels like "actual stuff". But in terms of the mathematics, it's because the configuration of fields that corresponds to compressed jello has a higher energy value than when uncompressed. In other words, while we don't assign intuitive properties like "solidity" and "elasticity" to fields, the relation between the fields and those properties in macroscopic objects is quite direct!

> How and where _do_ fields "exist"? Where are they in the standard model? Intuition says maybe they reside in spacetime near particles. I have a hunch from QFT that the fields are technically "everywhere" and are merely highly excited into particles when energy transforms the field in certain spots.

Indeed, they're technically "everywhere". You commonly hear that for the Higgs field, but it's true for all others as well, such as the electron field, whose excitations are electrons.

knzhou··on Teenagers are better behaved and less hedonistic (2018)
I went to a school with similar dynamics, and if it's any comfort, I can assure you that the bullshit doesn't pay off. Those of us that pursued our interests deeply had our pick of college, those obsessed with simple metrics like AP count and SAT score got to what you call second-tiers if they were lucky, and those chasing leadership positions in do-nothing clubs got sweeping rejections. The system is not fair by any means, but it's not stupid, it can recognize people trying to game it.
knzhou··on Tips for Mathematical Handwriting (2007)
While I'm with you on the Greek, bra/ket notation is one of the best anywhere in physics. Here [0] are some examples of why it generally looks simpler, and here's [1] a derivation which would be a nightmare to write without it.

0: https://physics.stackexchange.com/questions/488903/advantage... 1: https://physics.stackexchange.com/questions/533523/confusion...

knzhou··on Tiers of answers to half-baked questions
"Field" is the abstraction we use in physics to say "there is a value associated with each point, and we're just not going to say whether it's representing the displacement of some invisible jelly, or the rotational velocity of some invisible tiny gears, or if it doesn't even make sense to ask the question". When we say the electromagnetic field wobbles, we just mean some abstract value is going from positive to negative and back.

I know that sounds unsatisfying, but there's good historical reason for doing this. Maxwell, for example, had a rich intuitive understanding of the electromagnetic field as a set of many invisible jellies and gears pushing on each other. His original equations were also 5x longer than the modern form and impossible for anybody else to understand, plus all the extra ingredients made tons of predictions that played poorly with relativity. The sheer butchery one needs to make intuitive pictures like this compatible with relativity is so high that we've basically collectively decided to give up. We now say fields make up real jellies and gears, they're not made of jellies and gears. They're just fields.

knzhou··on Tiers of answers to half-baked questions
In theoretical physics, "why X" stands for absolutely nothing more than "please derive X from some other principle". Some people will use Y to derive X. Others will say that's not a real explanation, you really should use Z to derive X. And yet others will flip the derivation around and say that X really is the fundamental principle, and use it to derive Y and Z! (The route one takes depends on what one views as "more fundamental", a metaphysical notion that nobody agrees on.)

It's completely fair if you don't think any of the answers were satisfying, but it's not like there's another option on the table.

A lot of laypeople think that a truly satisfying answer must be mechanistic -- something like "there's an invisible jelly everywhere and it wobbles at a finite speed, and that is the speed of light" -- but the 20th century taught us that this completely natural desire is actually counterproductive, because you have to add piles of epicycles to hide the jelly's other effects.

knzhou··on Ultraviolet Irradiation of Blood: “The Cure That Time Forgot”? (2018)
They don't. All of the "evidence" here is cherry-picked clips where Biden stumbles for 2 seconds within an hour-long speech. Even now, Biden's a far better speaker than all the people snidely passing these clips around.
knzhou··on YouTube bans coronavirus-related content that directly contradicts WHO advice
This is one of those completely false things that people only believe is true by repetition. Go back and actually read the full set of WHO statements in mid-January. They have a bunch of statements saying that nations should get prepared, one saying that specific studies haven’t yet found hard evidence for person-to-person transmission (because at that point most of the cases they’d managed to find were tied to the market). The WHO never, ever said that it can’t be transmitted, and they absolutely never said that people should do nothing about COVID-19. They were urging nations to act for months before they actually did.
knzhou··on YouTube bans coronavirus-related content that directly contradicts WHO advice
The central kind of thing that this ban is meant to prevent is people posting on Youtube to drink bleach as a COVID cure and laughing over the ensuing deaths. We've disapproved of that kind of speech for centuries -- this slope isn't nearly as slippery as you think it is.
knzhou··on YouTube bans coronavirus-related content that directly contradicts WHO advice
There are constant new outbreaks, at the rate of several per year. (If you want to try to keep up, see here: https://promedmail.org/ ) The vast majority fizzle out, and for the rest, travel restrictions are not often useful or necessary. Recent examples of this include SARS, MERS, swine flu, Zika, several hemorrhagic fevers, and resurgences of diseases like measles and plague. If the WHO responded to all of these with a call for a global travel ban, travel would never be allowed.
knzhou··on YouTube bans coronavirus-related content that directly contradicts WHO advice
There are legitimate arguments that travel restrictions are counterproductive, so you should argue against those, not against some perceived double standard that doesn't actually exist.
knzhou··on YouTube bans coronavirus-related content that directly contradicts WHO advice
No, that's wrong. WHO essentially never encourages travel restrictions, their whole philosophy is that open international collaboration is crucial (that's why the W stands for "world"!). At the height of the Ebola epidemic, they didn't call for travel restrictions to Africa either.
knzhou··on YouTube bans coronavirus-related content that directly contradicts WHO advice
It doesn't work that way. It's trivial to build a rigged tree of evidence that makes any claim look good. Remember gish gallops from 90s internet arguments?
knzhou··on YouTube bans coronavirus-related content that directly contradicts WHO advice
There's an endless flood of these bad comments. I'm just going through them (a new one posted every 2 minutes) and making the same rebuttal. People are stupid -- especially people who think they're smart contrarians.
knzhou··on YouTube bans coronavirus-related content that directly contradicts WHO advice
The ban is meant for 4chan-based messages like "drink bleach/rubbing alcohol/methanol to cure yourself now!", which have killed plenty of people.
← PreviousPage 6 of 27Next →