peripitea

Improving the world's access to information is probably the biggest one? I would guess Google Search is, if you could magically add up all the net positives/negatives, one of the best things to happen to humanity in the last few decades. Having grown up before it existed, I can't imagine how much I would personally pay just for Search to come back if Google all of a sudden disappeared. I would probably give up my car before I gave up Search.

I know I replied to a different comment of yours, but this paper also illustrates how this point, while matching common medical wisdom, is not necessarily correct when accounting for altitude: https://www.thelancet.com/journals/langlo/article/PIIS2214-1...

(Summary: Healthy children at higher altitude breathe more rapidly.)

This is only true for people who live close to sea level. At higher altitudes, saturation even down into the high 80s can be normal: https://www.thelancet.com/journals/langlo/article/PIIS2214-1...

From what I understand, this is more because having that low of oxygen at rest is highly likely to be a symptom of a deeper problem, not because the 90% oxygen by itself is a huge imminent problem. Afaik you can go days/weeks at that level and be mostly fine if you are otherwise healthy. For example, this paper suggests that some people who live at high altitudes have perpetual SpO2 of <90%: https://www.thelancet.com/journals/langlo/article/PIIS2214-1....

For those of us living near sea level, temporary drops below 90% should be ok if you are otherwise healthy (up to a certain point of course), because as someone else in the thread suggested, they will rebound to >=95% as soon as you return to rest/normal breathing.

(Disclaimer: I'm not a medical expert, but have had to study breathing a fair amount. Happy to have corrections from anyone with more knowledge on this!)

In my experience this isn't that much better from a discomfort perspective, because (I would guess -- not an expert on physiology) the lower available oxygen in the lungs also leads to a more rapid increase in blood CO2. It might be a bit better depending on your particular body though, so yes it's worth trying both out.

Actually I just tried it and yeah for me at least the full-lungs version is obviously longer but takes less discomfort for my SpO2 to start to drop (like well before contractions start), whereas on exhale I get more discomfort (around 15s before contractions) before a similar drop.

In my experience you need to hold your breath for a while (like 2-3 minutes) for blood O2 to start decreasing. Certainly doable (and safe as long as you're lying down away from water) but might take some practice getting used to the discomfort. You can hyperventilate a bit before to make it easier (though never do this in any other breath holding scenario).

Please don't put words in my mouth. We were paying well over market rate. There simply are not enough qualified Americans to do all of the highly skilled jobs available.

O visas are much harder to get than H-1B. I've hired a bunch of people via H-1B that we could not have hired via O. None of them were displacing American workers -- we would have hired just as many (or more) qualified American workers if they had been available too. So it seems disingenuous to imply that O covers the intended purpose of H-1B. (For the record, I agree that companies like Tata abuse H-1B and that sort of situation should be banned.)

For those interested in data, the COVID Tracking Project (http://covidtracking.com/) doesn't have county data, but it does have test data (both total, positive, and negative). Very helpful given how relatively useless all of these case counts are without knowing how many tests were conducted.

Among others, the numbers are way under-reported. For example, in Iran there's good reason to believe that 1-5% of the population already have it: https://www.theatlantic.com/ideas/archive/2020/03/irans-coro...

Even if you look at Italy, though, and if you assumed that their numbers were accurate (highly unlikely), 20/100k is only 5-6 doublings away from 1% of the population, or as little as 3-4 weeks if not properly contained.

Is there any indication this would apply to all ISO/NSOs? Can't read the WSJ article (no subscription), but articles referencing it seem to be saying that it's targeted at highly-paid executives. Obviously this would be disastrous if it applied to all stock options. Just trying to get the facts straight before getting too riled up.

I actually disagree in both directions. Have you played or coached many sports? I have, and I'm not trying to be an asshole with the question, but it just seems so clearly untrue to me that I wonder if you haven't or if we're somehow talking about different things. For some sports/positions, you can become 95th percentile even with below average athleticism, and for others it's essentially impossible even if you have average or even slightly-above-average athleticism. It's all on a spectrum depending on how reliant that sport or position is on innate vs trainable attributes.

Let me know if you disagree with any of these, or if you think the examples I'm choosing are not within the scope of your argument: -No matter how much you drill, you will never be a 95th percentile center in basketball if you're 5'8" or whatever the average male height is. -Slightly further down the innate vs. trainable spectrum, I have never seen an average athlete become even close to a 95th percentile sprinter, long jumper, football wide receiver, or soccer winger. Speed is too important for these, and speed is largely innate. -I have seen below-average athletes become 95th percentile power lifters, football linemen, archers, grapplers, climbers, long distance swimmers, etc. That's because these sports/positions depend on more trainable attributes like strength, flexibility, endurance, technique, and mental focus/resilience.

Yeah I could be wrong on basketball. I don't play it much but now that I think about it you can be a great shooter without being a great athlete, and that alone can make you quite valuable.

There are plenty of sports for which I find this still holds true. A lot of it comes down to how much the required skills to be great are around more innate skills (e.g. speed, height, jumping ability) vs acquirable skills (technique, flexibility, problem solving, mental fortitude). It's very hard to be a 95th percentile soccer, basketball, or track athlete without a decent amount of innate athleticism. But I've seen plenty of not-particularly-athletic people get to 95th percentile in things like climbing, grappling, long-distance running, etc.

This is a great post, and it's unfortunate that so many people are missing the point. You may not like the writing, or the layout, or the title, and you may be able to pick holes in the argument, but who cares? That's an easy, boring game to play. Take it from a reformed cynic -- you will grow so much more from trying to figure out what's right about someone's argument rather than what's wrong.

The key insight in the article is: For many skills, becoming highly proficient is easier than most people realize. That is both true and highly under-appreciated, in my experience. Any time you can combine those two attributes, you're writing something important. Keep the "ridiculable" posts coming Dan!

The point is about improving skills.

SAT performance is (largely) determined by intelligence, which is (largely) innate. If you're relatively smart and don't practice at all, you will score higher on the SAT than your less smart friend no matter how long they practice.

This applies even more so to height.

With many learned skills, though, if you invest time and are thoughtful about how you approach your learning, you can get to 95th percentile fairly easily, despite that seeming quite impressive. Many participants don't put the time/effort in, and many of those who do don't do it in a remotely efficient way. If you do both, you are virtually guaranteed to get to a high percentile across a wide range of skills, mostly independent of your innate ability.

Putting aside whether this is ok for a second: This should not be a surprise to anyone. Of course they are doing this. I would almost argue that it would be foolish of them not to collect this data. The lifetime expected value of the data is so much higher than the marginal cost of transferring and storing it.

I can't think how many times on the product side of services I've run that we've been grateful to have this kind of data. Sometimes it's for obvious reasons that we always knew, but often it's for suddenly crucial reasons that we never could have anticipated when we first started collecting the data.

Again, that doesn't mean it's OK and I strongly support GDPR-style data privacy legislation in the US. But in the meantime I guarantee you that just about every service you use is gathering data like this and a whole bunch more.

>but damn is it frustrating to see so many vulnerable students saddle themselves with so much debt in pursuit of making a stable living.

It sounds like you're talking about traditional American colleges with this criticism, not Lambda school. I'm confused.

Yeah you're probably right that VP, rather than SVP, is the more common cutoff.

This seems to be more a concern on the terms rather than the instrument. You can structure any financial instrument (ISA, loan, whatever) to be predatory if you're willing to play with the variables sufficiently.

I don't know if those are real terms in your example, but they do seem pretty bad. (Although still better than a predatory loan IMO, since they are contingent on the student actually having a somewhat successful outcome).

By contrast, something like Lambda's ISA ($30k max payment vs $20k upfront tuition; 15% of income for two years; min $50k/y salary as software engineer) strikes me as not predatory in any way -- I would guess they make less money from their ISA students on average than from their tuition-paying students.

Edit: For context, his argument is basically: "If independent learning is not your forte then that's a much larger problem, because you will need to do significant independent study regardless of the bootcamp."

It's not a great argument. Yes, being able to learn/research independently is an important skill that everyone should have, but it's a spectrum. Teaching yourself programming in six months through self-study alone is VERY FAR along on that spectrum. I would bet there are far more people who can get a job with the structure/forced deadlines/etc. of a bootcamp + a heavy amount of self-study to improve their skills, than there are who can do the entire thing themselves.

Also, self study is far from the most efficient way to learn most things. Having someone to help unblock you when you have a bug you can't figure out, or when there's a concept you're not quite grasping can make a huge difference in your overall rate of learning.

I know multiple people who could not have attended a bootcamp if not for ISAs. I also know one who could afford to pay but still chose to go the ISA route. They are all happy-to-thrilled with the model. I would love to know your reasoning for describing them as "just as bad if not worse than predatory student loans". I can't conceive of a worldview where that would be true, so I think I'm missing something in your perspective.

Yes, that's exactly what he's saying -- the "just as well" part is key. Sure, there's an IC track that goes up to the SVP level at many companies, but it's statistically much harder to climb pretty much everywhere.

>The risk-adjusted $ compensation is the same at a startup and at bigco. Startups just have more beta in the comp.

I work with startups, and while this may have been true 10-20 years ago I can't imagine how you would think it's true now. As the sibling comment says, I'd love to see your work, i.e. a remotely plausible example case with numbers illustrating how startup EV would approach bigco EV.

Yep. It continues to become more intense as you go deeper. You can wreck your lungs going to even 20-30m if you haven't trained them to handle the pressure.

This is because you are artificially pressurized by the air you're breathing when scuba diving that deep. If you held your breath and went that deep, you would definitely feel the pressure.

90 seconds is uncommon for complete novices, but most people can get to 2-3 minutes within a day or two of training. It's almost entirely psychological up to that point; blood oxygen levels barely drop for the first few minutes you hold your breath while at rest. You just need to learn what the sensations are, that they are normal, that you aren't at risk (assuming you've done some basic amount of safety education), and to relax into it.

That's very helpful. Thanks!

How do you handle progressive overload? Do you have a bunch of different kettlebells of various weights? Seems like that would defeat the space savings. Or do you cycle out the lower weights or put them in storage once you've progressed past them? Curious to see if I could make this work for me, as I also have time/space constraints.

Yes, that seems like an important problem, but one separate to what they're describing in OP's article. (Again, assuming I'm understanding this right.) Their constrained RL approach is still relying on our ability to enumerate and assign costs to the undesirable behaviors, right? From reading the article, I get the impression that they are focused on addressing that scenario, and leaving the problem of how to enumerate all undesirable behaviors to separate research.