josquindesprez

I strongly agree with the sentiment of this, and naturally gravitate towards avoiding any such loaded language.

But writing posts like the ones he's referring to doesn't seem like a great solution to me. They come off as a little disingenuous, like they're trying to show off how detached they are from the Big Moral Questions while still shunting you towards a Big Moral Opinion.

His thoughts seem to be shaped by having been exposed to things defined by the contours of contentious terms. He's writing the posts to add a rational, more persuasive voice to a broader conversation, while intentionally alienating himself from the broader conversation. It's a worthwhile persuasive strategy...so I would've liked to have seen a bit more on how to get over the hurdle of tone.

My personal experience is that most people in the language learning community have trouble fitting SRS into their long term workflow because they don't stop to think much about the natural spacing of repetition they get in their other learning processes.

The people that popularize things like Anki the hardest are those that built and stuck with their own systems from the start. That's more of a referendum on their personality type than on the utility of their way of using Anki.

My hunch is that most people end up struggling because a lot of the most popular systems end up doing the exact opposite of spaced repetition once you consider the complete universe of foreign language input a learner is getting.

I've found that for me, Anki works really well for the first 1000 words, letting me jumpstart early vocab while most of my other time is spent on basic grammar.

For words 1000-5000 or 10000 in frequency, it's easy to get stuck in a trap I see a lot online: every time you encounter a new word, add it to Anki. This is a great way to burn out. You'll encounter most of these words with a natural spacing as you read native materials, if they're in the higher frequency bands. Doing reviews becomes excruciating, since you're losing the efficiency benefits of using the SRS unless you adjust review frequency based on the native input you consume (a good machine learning side project, perhaps?)

I've had a lot more success using Anki for words off in the long tail that I wouldn't otherwise have a chance to remember.

> In a free society, you penalize the people who can't act like responsible adults, not the people who can.

If the ills of social media are, as you say, (and I also believe), caused by the higher-order effects of many individuals "failing to exercise common sense and critical thinking," then who are we to punish?

A phenomenon like QAnon might have been started by a single nefarious person, and amplified by a small group of misinformation lovers, but it's only because of social media's leverage that millions of people have had their moral framework and way of interacting with the world corrupted so heavily.

Punishing the irresponsible was a reasonable solution for all of history where the irresponsible had a reasonable amount of societal leverage. We're no longer there.

Perhaps banning social media isn't the way. But focusing on the humans that find leverage points in the system to amplify bad messages also doesn't solve the fact that the system is designed to maximize ordinary people's ability to create self-sustaining societal doom loops.

Banning guns might be overreach. But banning something that's the functional equivalent of distributing assault rifles to millions of toddlers seems like a necessity in a free society.

I'd honestly have a hard time making a case that Fox News is less net bad for society than Facebook as of today. But Fox News feels like the "end-stage" version of whatever you'd call that particular flavor of classic corruption and human-brain-hijacking, whereas Facebook has been visibly improving its algorithmic behavior control abilities, has already invented new flavors of human-brain-hijacking, and seems to be getting better and better at making us worse every day.

We've learned how to handle the old style of human-brain-hijacking. Those lessons were written in blood and war. Facebook has few parallels, only distant rumblings of far worse lessons to be learned.

These are the same word games that got us tired arguments like "guns don't kill people, people kill people." Looking only at people as individuals limits you to seeing first-order effects. The insidious effects of social media come about because large scale algorithmic optimization has found high leverage points for influencing society. While that's ultimately a consequence of individual people, you can't solve many complicated problems using a lens that can't see higher-order effects.

People seem to be fundamentally incompatible with social media, much in the same way they're incompatible with bullet wounds and drug addictions. Blame-passing word games just get us farther from taking that truth and starting to fix the world.

A lot of good insights here, particularly on how Substack's network effects and incentives push authors towards a milquetoast homogeneity.

But the author's self-admitted unfamiliarity with Substack leads the article away from pointing out the true value of the platform, getting stuck instead in a thicket of contradictory takes:

> [Substack] can become a full fledged aggregator, build network effects and community, personalized content and so forth, but risk the moral purity of being one of the last ad-free algorithm-free corners of the internet [...]

> [I]t is still worth speaking the obvious truth that having someone who is not your spouse feed their thoughts to you 5-days-a-week, thoughts that they themselves have only had a day to work on, thoughts which would likely go refined or unexpressed in a publication with longer-time horizons, is probably not good for your brain. That’s not a “weekly update”, it’s a parasocial relationship, it’s the same poison you get from every other one-sided social media platform.

I can't wrap my head around the idea that Substack is both good (because it avoids many of the things that poison the well of social media) and bad (because it's social media and thus still poison). Not all social media is bad, just the forms we've invented to date. Some of the positive things I see in Substack are shaped by the same incentives that could lead to a future, better generation of social media:

Because it's missing the algorithmic distortion of social media, even with Substack acting as a forcing function towards topic-of-the-day Twitter-chasing homogeneity, it's very hard to find things in the world that let one do:

  from smartfriend import mental_models

for only $10 a month! To me, it's a potential antidote to social media: not only is there the mildly pleasurable pornlike imitation of socialization, I can get trained to think like a variety of intelligent people whose social circles I otherwise wouldn't be able to break into. That's a form of digitally enhanced socialization (read: social media) I'm happy to pay for.

Even with the homogeneity on the platform today, there's still a lot of great thinkers you can watch in action. Perhaps the author of this blog is intelligent enough to come up with these mental models without watching others think. I'm not so lucky.

Tech has already given us assortative mating via online dating, and people seem to like that. But for everything else, we get trapped in filter bubbles. Substack lets me finetune my social simulacra to deliver maximum benefit.

I wish they also realized this potential themselves.

I imagine for popular posts, you'd get way too many low-value bumps: lots of leaves in the comment tree would be low-effort back and forth replies of uninteresting content (e.g. Reddit). If there's a voting system, the top-level thread would get bumped a lot but it'd be difficult to find that low-value content. The signal that the bump is trying to convey (there's new and good stuff here!) wouldn't match the value of the information that the user gets (the scattered dregs of many conversations). If there isn't voting and you sort by newest content, the experience would probably be like 4chan in slow motion: bump ordered blasts of low-value replies.

In a classic forum, since posting something not only moves the thread to the top, but puts your post in a highly visible place (the end of the thread), it's a lot harder to have too many deeply branching side conversations with low value replies: there's a social norm against polluting threads.

For something like HN where comment trees don't get too deep and the replies are always meaningful, I think threaded bump ordered ranking could work.

This appears to be a mediocre piece of undergrad "research", which is to say, the theoretical material is entirely plagiarized without citation. Even the applied analysis (Beethoven's 9th, 2nd mvt. mm. 143-172) is directly lifted from another work.

The application of group-theoretical ideas to music theory (in particular, the rules underpinning voice leading) is quite interesting, however! People have been interested in the link between symmetry and musical beauty since the time of the Pythagoreans. Looking through the modern lens of group theory shows a delightful simplicity: if you look at the world of musical operations in this way, the ones that sound best are often small deviations away from maximum symmetry.

There are far better places to start, if you're interested, covering much of the same (plagiarized!) material:

https://www.math.drexel.edu/~dp399/musicmath/algebraicmusict...

https://sites.math.washington.edu/~morrow/336_09/papers/Ada....

https://alpof.wordpress.com/category/music/math-music/neo-ri...

http://www-personal.umd.umich.edu/~tmfiore/1/mathmusiccolloq...

https://www.maa.org/sites/default/files/pdf/upload_library/2...

Or, of course, follow the citation chains on Wikipedia:

https://en.wikipedia.org/wiki/Neo-Riemannian_theory https://en.wikipedia.org/wiki/Transformational_theory

> The other article has examples like "how do I get the subset of an array in JS". In my experience if I encounter this type of question it either means that I end up having to hack a JS program but it's a one-off I don't intend to actually learn the language (in which case I don't really need to bother to learn the language) or I'm actually learning the language but then I will probably encounter this situation enough times that I'll have it memorized soon enough.

> Either you really encounter this word often in your work and you'll have learned it soon enough, or you see it once every other year and then is it really that big of a deal to have a quick google to refresh your memory if you've forgotten what it means?

You've identified a key problem with today's (admittedly justified) cult-like worship of spaced repetition systems: most people don't take the natural spacing they'll encounter into account.

A while back, I tried learning languages by curating and dumping 1000s of sentences with relatively common words into Anki. It was more effective than paper flashcards, but I realized that I was wasting a lot of time and boring myself by continuing to keep words I'd encounter while reading native material in my SRS. I eventually reached the conclusion that it makes the most sense to fill the SRS only with the long tail of words where the natural spacing you'd encounter isn't sufficient to create the natural SRS effect (i.e. words less frequent than, say, the 10,000th or 15,000th most common word).

It's important to take this with a grain of salt. Looking at the fit of the model itself, it seems to be a bit overconfident/overfit to the current configuration of society: https://twitter.com/babeheim/status/1108737930804293632

> There is no evidence that a person with an IQ of 170 is in any way more likely to achieve a greater impact in their field than a person with an IQ of 130.

This feels disingenuous. If other things are bottlenecking human success than intelligence (which isn't even entirely true), computers are generally better at these things, especially in an era of massive low-cost computing resources.

Conquering the Electron: The Geniuses, Visionaries, Egomaniacs, and Scoundrels Who Built Our Electronic Age.

It's a history of electronic invention over the past 150 years. Not the best from a historiographical point of view, but very insightful for getting inside of the minds of those who developed the technology, and getting a sense of the tech itself, the mental steps required to invent it, and the personalities of those involved.

Better yet, this iteratively deepening list of land-lake layers: https://en.wikipedia.org/wiki/Lake_island#Islands_within_lak...

This culminates in:

  The largest island in a lake on an island in a lake on an 
  island is a nameless, approximately 4.0-hectare (10-acre) 
  island at 66.687°N 70.479°W, situated within Nettilling Lake 
  on Baffin Island, Canada.`

There are (extremely untrustworthy) rumors circulating around day traders' Twitters of a potential takeover by Microsoft.

e.g. https://twitter.com/search?src=typd&q=amd%20msft

In my experience, the 'heads down; just let me code and go home' attitude is beaten into a lot of engineers by a series of bad experiences with management. Experiences in which their input is repeatedly ignored in favor of what seems like management bullshit.

> if you want to be involved at the upper levels, nothing is stopping you. you just need to either find that job, or create one for yourself.

I'd imagine many engineers are coming at it from a slightly different perspective. They're used to being able to self-teach. They're used to REPLs, where getting instant feedback is easy. They're used to having the luxury of rapidly iterating on failure before they get it right. They're used to a world of technical decisions where being smart and right is often enough to win.

Trying the same process with management bites them in the ass, and disillusions them.

It sounds like you envy their freedom to put their head down as much as they envy your freedom to _matter_.

Ultimately, you're not totally wrong, but complaining about the symptoms without treating the cause isn't productive.

There's also this take on differentiating datatypes: https://codewords.recurse.com/issues/three/algebra-and-calcu...

Reading this makes a lot of the operations in colah's article feel more intuitive. (To me, at least. I'm no expert here.)

The water sitting around in the pipes is likely colder, so it might be more than just psychological 'coldness'.

On the other hand, this brings my tap water closer to drinkability in the winter, so it's not all bad!

If you haven't seen it already, start with Feynman's appendix to the report on the Challenger disaster: (https://science.ksc.nasa.gov/shuttle/missions/51-l/docs/roge...). Pretty much any of the reports put out by the NTSB would fit the bill. They do a lot of what I like to call 'full-stack debugging' of engineering (both of products and culture), where the stack spans all the way from material science to management science. It's rarely the case that there's a simple mechanical engineering explanation for the big failures that they investigate. It's the confluence of multiple engineering and cultural failures that causes catastrophes.

For a better look at the politics of investigations, try this article on the Columbia disaster, which was on the front page here earlier this week: (https://www.theatlantic.com/magazine/archive/2003/11/columbi...).

Alternatively, since you mentioned the blackout, there's a lot to learn from analyzing the government response to Katrina, say, something like (https://www.amazon.com/Disaster-Hurricane-Katrina-Homeland-S...).

Saying that correlation is usually accidental is a stretch. Basing that claim on that (admittedly amusing) website is even more of a stretch: the time series on that website are all highly autoregressive with minuscule sample sizes, which makes spurious correlations extremely likely [1].

Speaking (very) roughly, the fact that they are autoregressive constrains how 'kinky' the shapes can be. Visually speaking, the time series will have a small number of inflection points and will appear interpolated between these points.

This greatly reduces the search space: you just need to line up a couple of kinks in a space of ten samples, or match two smooth and vaguely line-shaped objects, rather than find a convincing relationship for 100s of observations not tied together in time.

[1] http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.611...

Yes, it's a thing. But similarity is no reason for sadness. One could make a similar argument about design patterns.

The important question to ask is, "what ways are there of making music sound appealing to as many people as possible on the first listening?"

Convergent solutions will arise, just as in the case of design patterns.

This is an interesting observation, but it's poorly contextualized, and becomes rather trite in the context of modern trends in melodic construction.

His attempts to contextualize this are not lacking: there's a hint at the minor third's use in schoolyard songs, a quintessential reference to the overtone series (the God of the gaps of armchair music theorists!) an appeal to Bernstein's The Unanswered Question (which is, sadly, for all its visionarity, bordering on irrelevance in the light of the intervening four decades of research on music perception), the quintessential appeal to the overtone series). There is no attempt to understand why the 'whoop' sounds good in songs, or how it relates to other trends in music. The mere identification of a pattern provides little insight into its usage and meaning. But who doesn't enjoy a few digs at the formulaic nature of pop?

Let's look at an example, from the article "Good Time". There's actually far more information density in pop music than is assumed.

Specifically, let's focus on the use of the 'whoop' as a way of delineating chorus activity. The article identifies the 'whoop' at 0:04, but this card-collecting example-finding behavior hides its main usage in the song.

A major 6-5 scale degree pattern predominates in the verses of "Good Time," often to start a phrase ("Woke up"; "What's up") or nested inside of other melodic patterns. Many phrases are ended with a descending 3-1-6, outlining a minor triad (or 1-6-5). All of these patterns avoid closure on the tonic, allowing for smooth elision into the next sentence.

To end the first half of the verse, Owl City uses a 3-5 pattern (the inverse of the 'whoop') at 'inside my head', followed by Jepsen humming 5-3-1, a foreshadowing of the 'whoop' in the chorus.

The 5-3 pattern takes over in the chorus. This fundamentally transforms how the 6-5 patterns at the ends of phrases are perceived. In the chorus, they are no longer a way of avoiding closure during the rapid declamation of the verses. They are fully subsumed as little moments of relaxation in between 'whoops' of freedom, joy, summer...in short, the joys of being a Millennial.

Among the myriad trends and styles that have enjoyed their moments of fame in the history of pop, the most notable is likely the tension between melodies and chord progressions based on the minor pentatonic scale (roughly, blues, rock, etc.) and those based on diatonic scales (arguably, the major pentatonic is stylistically aligned with this side). Different stylistic elements have predominated at various points in time.

There has been a resurgence in the popularity of certain diatonic patterns (especially those based on the diatonic major scale) in the past ten years. The 5-3 minor third is particularly useful as a bridge between the pentatonic and diatonic worlds. The 'Millennial Whoop' is no doubt an outgrowth of that trend. There has also been a trend for choruses to focus on the fifth scale degree, often in a higher pitch range than the surrounding verses (this is not remotely new, but in the past, was more identifiable by the use of V chords rather than melodic construction around the fifth scale degree).

As an example of the 5-3's usage elsewhere, Adele's Someone Like You, with its soulful roots and pentatonic construction, exemplifies many of the trends that are picked up in the 'whoop'. She starts many phrases in the verses with a descending 5-3 minor third, then moves to a higher pitch range for the chorus...with a few anguished (decidedly not feel-good and 'whoopy') minor thirds (e.g. "I beg"). (https://www.youtube.com/watch?v=1CGuafoq3T4)