Hacker News Reader

tprice7on George Floyd Protest – police brutality videos on Twitter

https://en.wikipedia.org/wiki/List_of_killings_by_law_enforc...

The rate of police killings is vastly higher in the US than any other Western country. So something is going on there that is not a universal phenomenon.

tprice7on Dolphins at popular spot miss tourists and keep leaving 'gifts' on shore

I think that was the point. They are trying to show how that sort of logic is flawed.

tprice7on Spotify signs ‘The Joe Rogan Experience’ to an exclusive multi-year deal

"It's not like they're kicking math off of YouTube."

Actually, a funny thing happened to me recently that is sort of in this vein. I made a simple app for tuning theory nerds, for examining how a given equal temperament approximates different whole number ratios. I couldn't link to it when posting on a Facebook group; I was told it violated Facebook community standards (I think they just completely blocked the 1mb.site domain that I hosted it on).

tprice7on So much of academia is about connections and reputation laundering

It's one of the most famous Japanese films, top ten perhaps.

tprice7on Ask HN: What scientific phenomenon do you wish someone would explain better?

Ok fair enough. I admit I regret this comment and think I posted it a bit hastily, I tried to delete it a few minutes ago but it appears to be too late now. I edited out the part about dinosaurs though.

tprice7on Ask HN: What scientific phenomenon do you wish someone would explain better?

1. Carbon dating. Sure, I get that carbon decays over time and this changes the proportion of isotopes. But why does this give you any information? That carbon didn’t come into existence just to be in that bone, it was made in the sun billions of years before that, so why does the age of the carbon tell us anything about organic matter? The key fact, which I think is not emphasized enough, is that the ratio of isotopes in atmospheric carbon is kept at a constant equilibrium by cosmic rays. So you can use carbon dating to tell roughly when the carbon was pulled out of the atmosphere. Without this additional fact, the concept of carbon dating makes absolutely no sense.

2. The tides. The explanation I was given is roughly something like “the tides happen because the moon’s gravity pulls the water toward it, so you have high tide facing the moon. There’s also a high tide on the opposite side of the earth, for subtle reasons that are too complicated for you to understand right now and I don’t have time to get into that.”

The first problem with this explanation is this: gravitational acceleration affects everything equally right? So it’s not just pulling on the water, it’s also pulling on the earth. So why does the water pull away from the earth? Shouldn’t everything be accelerating at the same rate and staying in the same relative positions?

The second problem is that, when viewed correctly, the explanation for why there is a high tide on the opposite side of the earth as the moon is equally simple to why there is a high tide on the same side as the moon.

The resolution to both these problem is this: tides aren’t actually caused by the pull of the moon’s gravity per se, but are actually caused by the difference in the strength of the pull of the moon’s gravity between near and far sides of the earth, since the strength of the moon’s gravitational pull decreases with distance from the moon. The pull on the near water is stronger than the average pull on the earth, which again is stronger than the pull on the far water. So everything becomes stretched out along the earth-moon axis.

3. This one isn’t so much a problem with the explanation itself, more about how it’s framed. I remember hearing about why the sky is blue, and wondering, “ok, more blue light bounces off it than other colours. But isn’t that essentially the same reason why any other blue thing is blue? Why are we making such a big fuss about the sky in particular? ” A much superior motivating question is “why is the sky blue during midday, but red at sunrise / sunset”? I was relieved when I saw this XKCD that I’m not the only one who felt this way:

https://xkcd.com/1818/

tprice7on John Conway has died

That article was published on April Fool's day.

tprice7on The 2,500 year old history of why Python’s all([]) returns True

Another way to make it seem more intuitive: "all of these are true" is the same as "none of these are false" (assuming values must be booleans), and clearly nothing in an empty list is false.

tprice7on Mathematics for the Adventurous Self-Learner

I emphatically agree. There is something that walking does to your brain that really helps you see the big picture.

tprice7on More bosses give four-day workweek a try

"This is what they already do with their spare time, there's no reason to believe there wouldn't be more of it."

I can see one reason: the benefit that you get from an activity is a nonlinear function of how much time you invest in it. This is why people tend to specialize in certain domains. If you barely have any time to practice painting for example, maybe it's not worth learning to paint in the first place. Conversely, if you have more free time, it could be worthwhile to find more creative things to do with it.

This is especially true if you consider frame-switching overhead. The amount you can get done in 6 hours of some demanding creative activity is probably more than 3 times the amount you could get done in 2 hours. And obviously people would have more long uninterrupted blocks of free time if they had to work less overall.

Also, with most schools being the way they are, I think children in general get very little opportunity to pursue their creative impulses; it's always about following the teacher's instructions. Maybe adults would have more curiosity and creativity if these traits weren't snuffed out in their formative years.

Finally, even in our current society, being just a lazy consumer in your free time is far from universal, for example: https://www.namm.org/news/press-releases/gallup-organization...

tprice7on The Rosetta Stone of Mathematics (2018)

Reflecting on this more, my best guess is that you think they are similar because they are both a "grand unified theory of mathematics". To explain the difference extremely briefly then, I'd say it's a massive exaggeration to call either category theory or the Langlands program a "grand unified theory of mathematics." They each demonstrate connections between some areas of mathematics, but they are different connections, and fall very short of unifying everything. If you want to know more, I think you'd be better off learning more about category theory and the Langlands program independently of each other rather than looking for a comparison between them.

tprice7on The Rosetta Stone of Mathematics (2018)

They are different enough that it's frankly a bit difficult to make sense of that question. I think in general, when asking "what is the difference between X and Y" questions, it helps if you elaborate a bit on why you think X and Y are similar / difficult to distinguish. That gives the person who answers the question a bit of a starting point: they can talk how the similarities you perceive might not be the whole story, or might be illusory (and you might end up answering your own question while clarifying your thoughts on the matter). Otherwise it might not be clear how to even begin answering the question.

tprice7on My story as a self-taught AI researcher

I found it to take much less discipline actually. I find an unstructured and curiosity-driven approach to learning math to be much more enjoyable and effective than the typical school approach. You are right about the solitude issue, although I’m unsure about whether this approach to learning is an intrinsically lonely pursuit or if there’s a possible society where it’s not.

I know I’m just speaking from my own experience and what works for me doesn’t necessarily work for everybody. But my claim isn't that everyone should do as I did, my claim is that you're wrong that a self-taught ML researcher would necessarily only be able to make superficial contributions because they are bad at math.

tprice7on My story as a self-taught AI researcher

This is logically independent from any claim about the value of formal education. I speak from experience that an undergraduate degree is not necessary in order to gain a firm grasp of undergraduate level math. Happy to elaborate if that is desired.

tprice7on My story as a self-taught AI researcher

"The usual corollary (that ML should "therefore" be able to learn with a few examples) may only apply, as I see it, if we somehow encode previous "learning" about the problem in very the structure (architecture, hardware, design) of the model itself."

Yes, and they do. They aren't choosing completely arbitrary algorithms when they attempt to solve a ML problem, they are typically using approaches that have already been proven to work well on related problems, or at least are variants of proven approaches.

tprice7on Clojure: `every-pred` and `some-fn`

Are there more recent features of Clojure that you think have made every-pred and some-fn less useful? I'm curious what they are if that is the case.

tprice7on The Hottest New Thing in Seasteading Is Land

The argument isn't that we shouldn't colonize other planets because it's expensive, it's that we shouldn't colonize other planets because it's Pareto inferior to making orbital habitats.

tprice7on Jimmy Wales has quietly launched a Facebook rival

Hacker News starts to severely rate-limit replies in deeply nested threads so if you want to continue discussing this you can email me (email in profile).

In regards to the question, the latter. The source corresponds roughly to the E vector of the pagerank whitepaper. It's a set of websites that's axiomatically good. The pagerank paper suggested the Netscape home page and John McCarthy's homepage as examples, or alternatively making every website a source, although the last option is more prone to abuse. The sink is harder to find a good analogy for. I think a reasonable choice would be to add a new node to the graph representing the sink, and then connect every website via a directed resistor. Then websites could be ranked by how much current flows through the resistor to the sink.

Edit: By the way, upon thinking about this more, I've just realized that I'm not sure about the exact mechanism by which link farms boost pagerank. As far as I can tell there are two possibilities:

A. The original "stuck in a loop" explanation, by which you can illegitimately magnify legitimately earned rank.

B. With the every-page-as-source strategy, you can boost the rank of a website by creating a new website and linking to it, even if nothing links to the new website. Then the other extra links in the link farm just exist to avoid detection of this sort of thing. So you are illegitimately creating rank from scratch, rather than magnifying existing rank.

Admittedly the electrical circuit thing only addresses mechanism A. Mechanism B can be addressed by being more discriminating about which websites are in the source. That applies whether you use pagerank or the electrical thing.

tprice7on Jimmy Wales has quietly launched a Facebook rival

“The question is whether or not you can determine the interconnectedness.” I’m not sure what you mean by this. The algorithm I'm describing doesn't require any data beyond what is already used by PageRank.

I have a feeling that you think the electrical potential must be defined by some ad-hoc method before applying an electrical algorithm, and this requires fancy techniques like cluster analysis and the like? This is not the case. Let me re-emphasize that you only need to specify the network of resistors (which is the link graph) and the source and sink, and then the potential is defined automatically in terms of those things (the same way as in physical circuits).

tprice7on Jimmy Wales has quietly launched a Facebook rival

"Or is that rephrasing your description?"

I think so? There is an obvious circuit equivalent where links are interpreted as one-way resistors. You get a charge potential automatically given the circuit and a choice of source / sink nodes (which you need to decide on anyway to apply pagerank).

tprice7on Jimmy Wales has quietly launched a Facebook rival

If I may go on a tangent, I’ve thought a bit about the problem of link farms and how it might be addressed.

The problem is essentially this: since pagerank is basically the probability that a random walk through the link graph will end up on your site, linking back to yourself, and no other websites, gives a big boost to your pagerank because a random walk will get stuck on your site. Of course it’s easy to just ignore self-links, but you can get essentially the same effect through clique-like groups of websites and this can be more difficult to detect.

What’s interesting is that an algorithm based on how electrical current flows (so a link is a one-way resistor, i.e. a resistor in series with a diode) would not have this problem. Attaching a conductive loop to some point in a circuit does not change how current flows. Electrons don’t get stuck in loops because they don’t drift around randomly, they move from lower voltage to higher voltage.

tprice7on How to Turn Physics into an Optimization Problem

You need a convex Lagrangian for the Legendre transform to work so I think it's more than a convention. Of course you could redefine the Legendre transform to work for concave functions instead of convex functions but in either case you'll want the convexity or concavity of Lagrangians to be uniform.

tprice7on How to Turn Physics into an Optimization Problem

"or maximizing"

The critical points of the action are either local minima or saddle points, never maxima.

tprice7on How to Turn Physics into an Optimization Problem

Also, a simple example of the action being a saddle point: a moving object constrained to the surface of a sphere, with no potential, which will move around a great circle.

tprice7on How to Turn Physics into an Optimization Problem

> Strictly, realizable paths are an extremum (maximum, minimum, or inflection point) of the action.

At a realizable path, the action can be a local minimum or a saddle point but it's never a local maximum.

Tangentially, while "principle of least action" is a misnomer, it's not too severe of a misnomer, because realizable paths always minimize the action on a short enough time scale.

A reference for both these claims: http://www.eftaylor.com/pub/Gray&TaylorAJP.pdf

(I haven't read it to be honest but the abstract confirms what I'm saying)

tprice7on Scala: A software tool for microtonic, macrotonic and other alternative tunings

Yeah I checked that too. It's puzzling as I am 100% certain it's older than 2007.

tprice7on Scala: A software tool for microtonic, macrotonic and other alternative tunings

Also see Sevish's scale workshop:

https://sevish.com/scaleworkshop/

It has the most useful bits of functionality of Scala, with a simpler UI, and as a web rather than desktop application.

tprice7on Scala: A software tool for microtonic, macrotonic and other alternative tunings

edit: weirdly the parent comment of my comment has disappeared and now my comment has been moved to a top level comment and appears to be apropos of nothing. Oh well.

I'm pretty sure this predated the programming language. I can't find a source on when Scala (the microtuning one) first appeared but it existed when I first got into this stuff around 2004, and I didn't have the impression it was new at the time.

tprice7on Do It Yurtself

I think Portland is pretty temperate compared to Mongolia, where such dwellings are traditional. Also, they better insulated than the mobile kind of tent and therefore are not Pareto-inferior.

tprice7on A visual proof that neural nets can approximate any function

No.