lagrangian
1yr ago
Maybe I'm missing it, but isn't Kulak making the basic mistake that Social Security (etc) are, by design, an immediate transfer from the young to the old? I will pay so much more than my 633k in taxes over my career. I get that most people pay less tax, and it doesn't all go to Social Security. Still, the analysis seems to fall short of showing that amount can't be paid off per person.
My prior of "nothing interesting happens" strongly objects, e.g. to
Sometime soon, the economic weight will become non-viable, like a broke gambler doubling his bet each time he loses, the debts will go exponential then asymptotic… and instead of a a lifetime to find a way to squeeze all that money out, there will be a crisis, and the American, probably “provisional” government at that point, will have to decide:
lagrangian
1yr ago
As others have pointed out: it is definitely causing some /priors updating/ that anybody takes this data even remotely seriously. This would be like polling a bunch of redditors about religion and then reporting on it as if this was a meaningful sample.
Right, that killed me. From Scott, emphasis mine:
I was surprised how certain people were that poly relationships were disasters that couldn’t work, compared to how little of a sign there was of that in the data. I like Aella’s explanation that most mono people’s experience of poly people is mono people “experimenting” with “opening up their relationship”, which is a natural danger zone. An alternative is that Aella got a bad sample (but her sample ought to be much more representative than mine), or that poly people lie / misremember / have a hard time answering surveys.
More representative, maybe, but still not even vaguely representative. "We only polled the people at the back of the church! How religious could they be??" I think Scott's ~asexuality just means he is typical minding really hard on these subjects.
lagrangian
1yr ago
We were unable to raise at my startup, after 4 years of hard work, and I'm starting to wonder whether these interviews are just accurate representations of my real abilities.
Most startups fail. Your startup going well enough for you to not want to pursue FAANG money for four years is a huge win. You should, unsarcastically, feel entitled to brag about it forever.
If you ever want a practice interview/some leetcode tips/to vent, DM me. (Current FAANGER).
lagrangian
1yr ago
Almost nowhere I've ever interviewed gives feedback. It's a legal/reputation risk for no gain.
lagrangian
1yr ago
Already tried that, got an offer >2x FAANG, doing cool work. Ain't worth it to me. I bought a house I could happily live in, indefinitely, in shorts. Hence, in OP:
something fully remote
lagrangian
1yr ago
Yup. Blog post if you're curious: https://blog.janestreet.com/why-ocaml/
Their podcast, Signals and Threads, is awesome. https://signalsandthreads.com/
lagrangian
1yr ago
I'm not an expert on C++ or HFT, but I think in short: yes. Very much so. The main thing. When I've interviewed in the past (Citadel, Two Sigma, etc), C++ is a huge topic. Outside finance, I get language agnostic leetcode type things (and usually answer in Python). In finance, details of C++ are the majority of the interview topics. E.g., implement your own shared_ptr class.
I think that while there are finance jobs where Python (no doubt wrapping Pandas or similar, and therefore C++) is adequate, something faster/more customizable/more predictable is usually needed. Golang is garbage collected, so the pauses will ruin your tail latency. "Web friendly" is not relevant at all to HFT.
Rust is an option - it should in general have equal control and performance to C++. But, the institutional knowledge and ecosystem are just not in the same league as C++. The basic rust tradeoff vs C++ is that you get more safety, but it's harder to get anything done. Maybe a sufficiently good Rust programmer can be actually more productive, since the safety lets you avoid thinking about certain things, but there's vastly more adequate C++ programmers than Rust ones.
Idk if "too low IQ for FAANG" was a joke or not, but I think "medium IQ, high conscientiousness, high willingness to be bored for money" is more the requirement. Sure, there are people doing cool shit who need all the IQ they can get, but it ain't most of us. For every engineer optimizing the company-wide database engine, there's thousands shuffling bits from A to B.
lagrangian
1yr ago
I do have those recruiters, yeah. I've even successfully gotten an 2x FAANG offer, but turned it down because it wasn't remote. I still think there's things I don't know about how to optimize the process, e.g. what pay/companies to aim for, beyond "trust your recruiter."
Software side of quant trading roughly, yeah. I also don't really know what the options are. I don't want go in the SRE/ops direction, probably - my only experience with that at FAANG was more stressful and less interesting than SWE. I have a math bachelors from a top program, which maybe opens up my options slightly. If nothing else, it seems to make companies happier to talk to me.
lagrangian
1yr ago
Does anyone have advice for transitioning from FAANG -> finance/HFT?
I know some c++, but not a ton, so "learn more c++" is the obvious first step. Beyond that, unclear to me what to do. I'd like something fully remote that pays better than FAANG and is less soul crushingly boring. Something with high performance code instead of just tons of business logic.
Relatedly, I'm about halfway from (1-2 years away from) promotion to senior. Worth sticking around to get it, or jump ASAP? Leaning "stick around" to hedge my bets and have time to study.
Great question, I'd love to know.
In a broad sense, being stoned is less impairing than being drunk. Not categorically - one (standard) beer is less inebriating than several dabs (especially sans tolerance). But, for typical consumption, I think it's clearly the case. The asymptotic inebriation is much greater for booze - people can drive blackout drunk, incapable of telling your their name. Even a hardcore alcoholic is still very fucked up at a certain amount of alcohol. I'd much rather be driven by the typical pothead who hits the bong every ten minutes than by the typical drunk who polishes off a fifth (15 shots) a day, or even the average person after a few drinks.
This is a double edged sword: it's easy and reasonable to say "don't drink (preferably any, certainly more than ~2 units) and drive." But, since THC is generally less inebriating, people are more likely to be stoned frequently/all the time, and this almost requires driving to participate in society. Similarly, I think it's much less acceptable to show up drunk to work than stoned.
A further difficulty is the lack of THC tests for current level of inebriation. It's hard to enforce stoned driving laws when all you can tell is "this person has consumed THC in the last few weeks."
I don't have a policy proposal here - just observing how tricky the situation/comparison to alcohol is.
lagrangian
1yr ago
So does this just end up in the supreme court, then get reversed? I think we should just make it ~impossible to prosecute presidents, past or present, and skip all the theatre.
lagrangian
1yr ago
I thought the implication was that the art-hoe was more dangerous to others. I picture the depressed incel not leaving the basement; the BPD, out keying cars over imagined slights.
lagrangian
1yr ago
a bit more expensive
5-10x roughly - about $3
I mean, you can reasonably infer, but just in case - I bought them on Amazon and put them on my penis and that into the woman.
lagrangian
1yr ago
Re sheep skin condoms: 1) they're actually sheep intestine, 2) they're great
That said, I don't know how old their use is, or if DIY sheep condoms were common, or if they would have been as surprisingly not gross as the kind you can buy today.
lagrangian
1yr ago
Frankly, I would pay double to live in a neighborhood of likeminded people who agree that barking, smoking, and subwoofers just don't belong in a shared building at all
If you've got the budget it for it, and like the other aspects, you've just described much of suburbia. If you'd buy the house from the new house flipper guy just down my street, mine would go back to being one.
lagrangian
1yr ago
Oh wow, source? My prior is "immigration (both rate and total) is large and bad", fwiw - I'd enjoy having solid stats.
lagrangian
1yr ago
You seem to be assuming that kids will know to find all the resources on their own, and generally do the executive function things a parent would likely be much better at. Parental encouragement, purchase of supplies (robot, pencils), and setting up the house comfortably for the hobbies (desks, quiet space) all matter greatly on top of what you can get by googling "learn to code"
lagrangian
1yr ago
Hm? That sounds like the opposite - OP is going to choose cheaper swings; your wife is going to choose fancier school.
lagrangian
1yr ago
Texas is currently being forced to accept more people showing up at their borders than are actually being born in the entire United States
While I agree with the spirit of your point, I don't think the above specifically is accurate. I see 3.7M births/year for the country, vs 1.8 M total population of illegal immigrants. The rate of "border encounters" appears to be 150-250k/month = 1.8-3M/year, but that's for the whole country.
And, while "border encounters" is technically in line with your phrase of "people showing up at their borders," the more reasonable comparison is to people actually net staying here. That must must be far lower, unless the illegal immigrant population was zero until about six months ago, which it wasn't, which makes the point feel weaker.
lagrangian
1yr ago
Unsarcastically, are there really so many Indian law cases ever that you would predict this to be the most useful second axis on any time scale? I would think the principle components would start with the political compass dimensions (economic/social left-right), and "opinion on Indian law cases" would be very far down the list.
lagrangian
1yr ago
Overall I have a very high opinion of Aella's integrity and have no reason to believe she's intentionally duplicitous,
I feel oppositely.
Her entire brand is attention seeking behavior through discussion (and sale) of sex. She got attention and advertising by redefining a thing in a way that you found interesting enough to podcast and effort post about.
Just did some reading of actual studies. E.g. this meta-analysis from NCBI:
Research conducted among hospitalized older adults found that pharmacological (e.g., benzodiazepines) and non-pharmacological (e.g., diphenhydramine) medications resulted in an 18% and 22% respective increased rate of delirium among the sample.14 Also, research conducted among community-dwelling older adults found that non-pharmacological medication use was associated with lower cognitive function scores as compared to those not taking these components over a 10-year follow up period.15
Why should I find this plausible, rather than making the standard "correlation is not causation" point? Surely people with issues sleeping are in general less healthy, physically and/or psychologically. I don't see anything in there to indicate they controlled for anything.
Some references also seem to do nothing more than ask if people took any sleep aid, lumping together everything from melatonin (presumably very safe, maybe placebo) to daily benzos (clearly neither very safe nor a placebo).
Edit: I am bad at reading (maybe it's the doxylamine). They did control for things. But my question re causation stands. I feel like TheMotte is usually very skeptical, and I find myself surprised by the strength of multiple posters' convictions here.
while controlling for demographic covariates, including age (0= 65 up to 75 years of age; 1= 75 years of age and above), sex (0=male, 1=female), race (0=white, 1=non-white), and relationship status (0=widowed, single, or divorced, 1=living with partner or married). Third, we examined the relationship between sleep medication use and incident dementia while controlling for Model 2 demographic covariates and health conditions (0=no chronic conditions; 1=heart attack, 2=depression, 3=hypertension, 4=stroke, 5=diabetes). Depressive symptoms were measured using the Patient Health Questionnaire-2.34
lagrangian
1yr ago
Fascinating that you say melatonin is dubious in usefulness. I swear by my 300 micrograms a day (dosage at Scott's recommendation).
My favorite zeugma:
Fact: In Pastafarian heaven, there is a beer volcano and a stripper factory.
Joke/zeugma: In hell, they're both flat.
I think Zeugma is a broader term than OP is looking for, though OP's examples are zeugmas. I define zeugma as "a word used once with two meanings as connected to different parts of the context."
More options
Context Copy link