Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
etangent
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
14 ms
·
91.
▲
by
etangent
6y ago
There is a (mistaken?) tendency to see OOP as a step beyond functional programming, probably because it allows for more complicated abstractions. But it seems to me a more natural scale to use is how the state is managed. In procedural prog
92.
▲
by
etangent
6y ago
I have not seen a single useful model from professional epidemiologists so far -- at least nothing that would guide me better than "look at China and Italy and consider that it might also happen here."
93.
▲
by
etangent
7y ago
Your entire post seems to be about #6. And therefore it's entirely wrong. Given that "people with background in infectious diseases" have largely failed, as a group, to warn us in January-February about this pandemic (see thi
94.
▲
by
etangent
7y ago
I don't understand how an n=36 study got into Nature in this day and age.
95.
▲
by
etangent
7y ago
Regardless of the other three points, it is generally rational to mistrust authority because many authorities have repeatedly failed in the past to protect lives during emergency scenarios.
96.
▲
by
etangent
7y ago
Imagine thinking one's choices are completely independent of status games played in the society.
97.
▲
by
etangent
7y ago
now imagine being someone who dislikes being around lots of people, and especially dislikes being around city people...
98.
▲
by
etangent
7y ago
> would you consider that an argument against the sanctions levied against, for example, doctors who lie and fake data to stoke fears about vaccines and autism? Yes. I never said authoritarianism is bad. In fact, I probably consider myse
99.
▲
by
etangent
7y ago
> Maybe, but it's also the impulse of people who rightly care about the future of our shared world. Literally no authoritarian says "i want to suppress people because I'm evil mwahaha." Every single authoritarian says
100.
▲
by
etangent
7y ago
The desire to "correct" other people who are wrong is the fundamental impulse of authoritarianism.
101.
▲
by
etangent
7y ago
> One advantage of this is that it forces you to name your variables such that they don't overwrite each other. Further down the line this enables sophisticated comparisons of states (e.g. dataframes) before and after (something dat
102.
▲
by
etangent
7y ago
You don't reload every time you write a line of code. Nobody's insane like that. You reload every two hours or so. This is good enough for most except most extreme data sets.
103.
▲
by
etangent
7y ago
No, that's not the workflow. You work in the notebook as normal but from time to time (say every two hours) rerun the whole thing. One advantage of this is that it forces you to name your variables such that they don't overwrite
104.
▲
by
etangent
7y ago
> insofar as that being outraged over what appears to be a more precise weapon makes very little sense. There is a rational sense to be outraged over more precise weaponry: precision incentivizes super-states to use it more, which in tur
105.
▲
by
etangent
7y ago
There is a recent Nature paper that uses a "relatedness disequilibrium regression" [1] that claims to show that twin studies overestimate heritability [2] [1] https://www.nature.com/articles/s41588-018-0178-9
106.
▲
by
etangent
7y ago
Sure, but for someone who does not know what an SVM is, the big-picture takeaway is that a particular version of "soft" statistical learning (deep learning) had succeeded where "hard logic" approach failed.
107.
▲
by
etangent
7y ago
Willing to bet that none of these experiments will replicate if conducted on a larger sample.
108.
▲
by
etangent
7y ago
The amount of Apple shilling on this board is incredible. $5999 gets you a base configuration that is outperformed by any $2000 Dell.
109.
▲
by
etangent
9y ago
I'm noticing you are applying this motte-and-bailey tactic, first claiming "it is all about politics" (bailey), but when people confront you, you retreat to "it is all about sociology" (motte) that is a lot easier t
110.
▲
by
etangent
10y ago
> “Men,” she writes, “have cannily figured out how to sidestep the implication of moral failing that attaches to mental illness—as well as the specific criticism of self-indulgence that is attributed to more introspective accounts of thi
111.
▲
by
etangent
10y ago
A possible semi-Marxian explanation is that what we call leftism has always consisted of two tendencies: a popular one and an elitist one. The alliance between the two is not natural per se but only held together while the elitist tendency
112.
▲
by
etangent
10y ago
> Why Trump isn't good for poor working class America - Assuming there's a large tax on every good coming into the country, then the cost of groceries will increase dramatically. The cost of cars will increase. The cost of cons
113.
▲
by
etangent
10y ago
A lot of valid points, but then there is this sentence: > It goes without saying that the universal ignoring of the psychic action of technology bespeaks some inherent function, some essential numbing of consciousness such as occurs unde
114.
▲
by
etangent
10y ago
I hope you're being sarcastic :D
115.
▲
by
etangent
10y ago
That's one of my favorite sci-fi novels and it's eerily ahead of its time IMO. Among many difficult questions, it also deals with "is it good for humankind to be ruled by an invisible technocratic elite?" "are we un
116.
▲
by
etangent
10y ago
In cultural-psychological sense, team sport sports absolutely are.
117.
▲
by
etangent
10y ago
Before man walked on the moon, it wasn't long before man first flew in a heavier-than-air apparatus, and before that it wasn't long before nobody has ever set foot on the South Pole. You can keep going. And do define "cope&qu
118.
▲
by
etangent
10y ago
> On the scale of tragedies that are currently occurring or for which there is a significant threat of occurring in the near future, the possibility that there might be a period of time with no living humans who have walked on the moon i
119.
▲
by
etangent
10y ago
People have been applying ML to genome analysis ever since it was first sequenced.
120.
▲
by
etangent
10y ago
> Every specialization has their own jargon to grok. True, but a more subtle observer might point out that the jargon can serve more than one function simultaneously. In fields like organic chemistry which is famous for its compound name
More ›