convexfunction

You might wonder whether it's actually regressive for people who have college degrees (or who at least attempted to get one) -- overwhelmingly not the neediest people in our society, even if I do sympathize with how terrible our student loan system is -- to be targeted for help from the government!

Ah, I think I get what you were saying, company A is billing company B hourly so employee of company A needs to track their time even if overtime-exempt. Yeah, not necessarily "rare" in tech, mea culpa.

Just to be clear -- frequently incorrect even among HN readers? I would think one or more of these exemptions apply to just about everyone likely to read these comments who has a job, but perhaps I'm misunderstanding them.

Yes, that is one of the complications, some roles are "salaried nonexempt". :)

Edit: the downvotes would seem to mean my "often accurate" generalization here is not accurate often enough to be worth saying, which I'm willing to believe. No level of detail in a HN comment can substitute for hours of research and talking with a lawyer, of course!

A good indicator is "do you have to fill out a timesheet". If you do, that means you're probably paid an hourly wage, and you should be paid overtime; if you don't, you're probably salaried and overtime-exempt. As usual, it's more complicated than that if you get into the weeds (e.g. in the news article linked in the comment you're replying to), and having a lawyer interpret your employment contract is probably the most authoritative way to figure that out if you feel your situation is unusual.

Homes are being used as investments by institutions and nonresidents, in the West broadly, because there's a widespread belief (especially among homeowners) that homes should have a large ROI rather than being more of a consumption good, and so policies have slowly gravitated toward furthering that end -- NIMBYism and restrictive zoning laws, basically, but various other stuff too.

Call me optimistic, but perhaps if more voters appreciated that "homes should be a good investment for the people owning and living in them" and "homes should be a good investment even if you're not living in them" and "the rent should be too damn high forever" all describe basically the same state of affairs, absent exceptionally well-designed policy that makes those not the same state of affairs, we'd see a slow drift over time away from this regressive nonsense.

It should've been (and almost surely was) some specific person's job to sign off on this thing being at least reasonable before deploying it, which they failed to do. Successfully redirecting attention from that specific person's failure to do their job (and their manager who failed to ensure they were doing their job, etc) to the abstract idea of Algorithms is some exceptional accountability judo.

Seconded -- I don't have much interest in reading the guy's own writing at this point, but I'll gladly read a concise summary that leaves out all the boasting since I assume there might be some actual "content" buried in there.

Would be cool if this is an exception, but yeah, there do seem to be far more "minor invents brilliant new X" news stories (surely thousands by now) than there are cases of these things actually seeing widespread adoption...

Nicotine gum surely does have the same teeth-cleaning function as regular gum. And, now that I look it up, it seems there's basically nothing online that claims nicotine gum itself causes staining (results are mostly just a study or two that establish that nicotine gum helps remove staining from prior smoking, which yeah that's probably true). I, or someone else talking to me, might've gotten confused about a claim originally made about smokeless tobacco products and thought it applied to gum too.

Probably goes without saying but vaping the nicotine would be the 80/20 approach here (or perhaps just 100/20), and I'll second that oral administration is also good. Lozenges in particular seem to have no downsides as long as you can find a flavor you like. (gum is great but stains your teeth) (EDIT: the gum probably does not stain your teeth, or at least I can't dig up anyone claiming it does)

I've been subscribing to Monthly Applications in Strength Sport for over a year now -- it's a research review primarily intended for competitive powerlifters/bodybuilders and their trainers, right at the intersection of "we actually have skin in the game and lots of practical experience managing body composition for ourselves and others" and "we know how to interpret studies and perhaps more importantly not over-interpret studies" -- and below is a writeup I did on the topic based on what I've learned from them (and applied to myself very successfully):

- forget about the idea of a "diet"; pretty much every popular "diet" doesn't result in sustained fat loss for most people when tested experimentally. if you want to lose body fat and keep it off, what you're trying to do is modify what you habitually eat "ad libitum" (eating freely) such that you end up completely satiated despite eating fewer calories than you do now. (yes, it just comes down to calories; some otherwise-sharp people will misinterpret a few studies to conclude that it doesn't, and those people are wrong.)

- there are lots of factors that go into satiety, but one of the most important ones is just having a lot of stuff in your stomach. so, the "caloric density" of what you eat, as in energy per mass, is good to know: if you want to eat less energy, then you want to eat more foods that have few calories per gram, and less foods that have a lot of calories per gram. this mostly reproduces the generic conventional wisdom on what you should be eating -- fruits, vegetables, unprocessed meats, cottage cheese and yogurt, and starches with high water content (potatoes, rice, pasta) are generally your best bets in the former category, while desserts and fast food and so on fall in the latter category -- but there are some surprises you might observe if you actually calculate that "calories per gram" ratio for the stuff you eat, e.g. nuts are ridiculously energy-dense and you should probably eat less of them if you'd benefit from losing body fat. (this doesn't apply to liquids, pretty much every liquid besides milk has negligible effect on satiety.) that said, you shouldn't plan to completely cut out whole categories of food, since that's much harder to adhere to than just eating more of the types of food that will make you full more efficiently.

- i believe that macronutrient-focused "diets" typically work for people to the extent that they cause you to accidentally eat a diet with lower caloric density. both carbohydrate intake and fat intake seem to be relatively uncorrelated with how much you eat ad libitum, but the average caloric density of what you eat is a much better predictor of ad libitum caloric intake. (protein is probably good for satiety though independent of caloric density.)

- rigorously logging your caloric intake for at least a few months is a good idea for multiple reasons, but i think the most important one is that it helps you develop an intuition for which foods are "worth it" in terms of satiety per calorie and result in you effortlessly hitting your calorie target, especially when fuzzier factors besides caloric density are at play (e.g. your own eating habits around specific foods). when i say "rigorously" i mean weighing absolutely everything you eat or drink unless you have nutrition facts for exactly how much of the thing you ate (i.e. single-serving packaged foods and food from restaurants that give nutrition facts), then looking up how many calories you ate based on how many grams of each thing you ate, and actually logging every single thing with calories that you put in your mouth, not just 80% of it. not everyone has the temperament to do this, and that's fine if you don't, but if you do then i'd highly recommend it. personally i just keep my food log in workflowy (you can use any note-taking app), but you could also look into "noom", which has a nice educational program in addition to the food and weight logging, or "macrofactor", which doesn't have the educational program but does do you the favor of actually telling you your macro intake and estimating your long-run TDEE for you.

- even if you're solely trying to lose body fat, regular exercise of some kind is a good idea because your satiety in relation to your energy balance becomes "dysregulated" when you're completely sedentary -- with at least a little exercise, you'll tend to match your caloric intake to your caloric expenditure, but with none at all, you'll be more prone to overeating. as a corollary, exercising a lot for fat loss is probably not a good use of time unless you intrinsically enjoy it, because you'll likely end up compensating for most of the energy you used by eating more.

Very weird to frame crypto as contrasting with "an open source culture that, at least aspirationally, built software around ideas of abundance, post-scarcity, universal access, and equality". Were the "older grizzled engineers" all literal communists or what?

As usual, the reporting I've seen hasn't involved any provably false claims that I've noticed, but it's been framed and selectively reported in such a way (again, probably mostly-sincerely) that most readers end up very misled. e.g., I think the internal study that found that a small fraction of teenagers reporting suicidal ideation said that the first trigger that came to mind was stuff they saw on Instagram on a survey was just insanely over-interpreted to mean that a world without Instagram would have less misery in teenagers, and I think that FB's decision to decline to publish a finding of an internal study that'll be predictably over-interepted in the media isn't really described well as a "cover up".

"Hate" in the sense of "ideologues being haters as usual", not like "KKK members".

At this point I think the hate from journalists toward social media and tech more broadly is better described as "cultural" than literal economic self-interest. If you're a journalist, most of your colleagues believe everything they read in the news and therefore sincerely think FB is somewhere between "at least as evil as its size alone would predict given the typical progressive perception of large companies" and "just cartoonishly evil", so unless you're unusually indepedent-minded that's probably what you'll believe too.

Did the studies show a negative impact in any meaningful sense, though? "Some teens report that Instagram contributes to or triggers to mental health issues" is my understanding of the most supposedly-damning finding; I understand why this can be turned into a high-engagement news story with (at best) genuine ignorance on the part of journalist and editor about how hard it is to defensibly turn survey results into interesting causal psychological claims or (at worst) mild deception, but I really don't think it tells us much about whether a world with any plausibly different Instagram or no Instagram at all would have less extreme or less frequent mental health issues in teens overall, or even in many of the specific people who report these problems.

I don't think declining to publicize the results of an internal study that will be predictably grossly over-interpreted by media and therefore by the public is actually wrong, and I read the FB response you posted an archive link to as saying "this research has been grossly and selectively over-interpreted", not "this research isn't very good".

Not that you explicitly claimed otherwise, but the focal point of the moment's discourse seems to be a small fraction of teens who reported suicidal ideation citing content they see on Instagram as the first thing that triggers suicidal ideation for them, and I very, very strongly doubt that the relevant content is often ads, or even content that could be argued to result from the fact that the platform is ad-supported.

The current discourse seems to be based mainly on insinuations that FB/IG actually harms people on the basis of a small fraction of struggling kids reporting that they feel worse when they look at Instagram. But, would anyone have predicted otherwise? That literally no teen who reports suicidal ideation would attribute their issues in part to what they see on social media?

Or perhaps that's a strawman -- would we not have bothered with this discourse if the reported fraction of teens reporting suicidal ideation who cited Instagram were 2% instead of 6%? 1%?

Given how many and how much young people use Instagram, what fraction there would represent "no effect size", where you can't get away with insinuating that Instagram has any causal impact on mood compared to any other activity that kids could be doing? It's probably not 0, right?

Yes, that's what I mean by "the law operationalizes".

Maybe, and the author makes a similar point. I don't think the current discourse, based mainly on insinuations that FB/IG actually harms people on the basis of a small fraction of struggling kids reporting that they feel worse when they look at Instagram, actually gets us any closer to an implementation of greater responsibility with benefits that plausibly outweigh costs.

I'd just love to see how the law operationalizes and the courts interpret something like "banning algorithms".

The clearest take I've seen on this: https://www.piratewires.com/p/bombshell

The money quote, in my opinion:

> The thing is, we’re not really asking for Facebook to control itself, we’re asking Facebook to control other people.

I don't know when or by how much the effect on life expectancy that obesity has is overstated -- funnily enough I think the same is true of smoking -- but one thing I'm more sure of is that obesity is likely to greatly decrease the quality of, say, the latter half of your life. It's just not very fun to have diabetes, "minor" strokes and heart attacks, or the various less-scary but very uncomfortable issues that obesity greatly contributes to. A nice way a friend of mine put it once is that you want to compress all the morbidity you'll experience into the last few years of your life, and being active and lean goes a long way there.

A very important caveat to the basically-true observation that some degree of energy expenditure compensation happens in a caloric deficit is that getting no physical activity appears to dysregulate appetite and make you far more likely to end up in a caloric surplus. See this review: https://pubmed.ncbi.nlm.nih.gov/29289613/

> Energy intake along the spectrum of physical activity levels (inactive to highly active) appears to be J-shaped, with low levels of physical activity leading to dysregulated appetite and a mismatch between energy intake and expenditure. At higher levels, habitual physical activity influences homeostatic appetite control in a dual-process action by increasing the drive to eat through greater energy expenditure, but also by enhancing post-meal satiety, allowing energy intake to better match energy expenditure in response to hunger and satiety signals.

So, if you want to lose body fat, you probably shouldn't focus on getting hours and hours of cardio in (unless you enjoy it), but you really should be doing some amount of physical activity on most days.

A slightly different but related question that hopefully everyone will find interesting, no matter what affect they attach to Amazon warehouse working conditions:

Food service in particular, I think, is having a lot more difficulty hiring and retaining people than they're used to, even with considerably higher wages and hiring bonuses, but if you believe what you read in the news then it sounds like Amazon warehouses aren't (otherwise you'd think they'd ease up a little on being hardasses). These aren't exactly the same type of job, but they're "pretty similar" in many ways. So, what's the difference?

You're betting this definition wouldn't have any serious unintended consequences?

A hypothetical example: suppose it becomes a legal nightmare to have even heavily censored webserver request logs retained for any period of time if your company does any advertising at all. That is, even if you have no intent or even ability to use those logs for advertising purposes, it might be a lot of work to prove that to the law, unless you take the hopefully-easier route of literally never advertising. "Boo hoo, companies have to prove they're not breaking the law", you might say; as is usually the case with these kinds of regulations, demonstrable compliance might be totally practical for bigger companies but a massive barrier for smaller companies, which on the margin means the difference between success and failure for quite a few businesses that would've otherwise created a lot of value.

That specific scenario probably wouldn't happen, I hope, but that's far from the only plausible failure mode! I would like to believe that we can figure out a good definition with relatively little value destroyed in the fallout if a law like this comes into effect, but it's almost certainly not going to be a single sentence.

Re: downvotes: is there good reason to believe this interpretation is definitely wrong? I wasn't able to find any details on where this number came from, and while I'd like to give Israel's public health officials enough credit that they wouldn't make this mistake, I'm hesitant to actually do that without more than a few mystery meat "percent effectiveness" numbers.

After researching a bit and finding this: https://yourlocalepidemiologist.substack.com/p/israel-50-of-...

I'm wondering whether this number is a population-level observational result that has more to do with someone ignoring base rates rather than delta actually being much better at causing infection.

I've lost count of how many times we've seen the same headline for different variants and relating to different underlying results, and this "less effective" headline always seems to get interpreted by people I talk to as "we'll be back to living like it's 2020 again pretty soon here" rather than the much more accurate "kinda sucks for us collectively but if you and all your close contacts get vaccinated (if!) then it's probably not enough additional risk that you should want to change your behavior".

I notice that I'm also curious what their methodology for coming up with these effectiveness numbers is, and even the "original" public source at https://www.ynet.co.il/news/article/rJQ1O5kp00 doesn't seem to make that clear.

I claimed that almost everyone working for these organizations (or to be more pedantic but more accurate, almost everyone involved in content for these organizations) is well-described by the latter, not that almost every article will reflect that perspective in an obvious way or even at all. For someone who tends to see the world differently from the average journalist or editor, the question then becomes: why take even the articles that aren't obviously biased at face value when you have a good idea of how the individuals involved feel about your culture and your priorities?