jonex

Such an important subject and they touch on important topics within that subject, but ultimately end up confusing things more and ultimately probably making things even harder for those who would be in service of advice on how to manage their weight (most of us nowadays).

Losing fat when overweight, provides very clear health benefits, we don't need to look at observational studies for this, the fact that gastric bypass operations have such positive impact on health is enough to show this.

It's true, that dietary advice rarely can make any lasting dent in population level body composition, however, this should not be confused to mean that a motivated individual can't. A very simple and robust method is to count calories and drop the amount until the desired level of weight loss is achieved, then simply revert back to a comfortable level when the goal is reached. It's just that this takes a level of motivation that most people don't have. This is not surprising as many of the negative effects of being overweight are rather subtle and takes time to appear.

This means that many people will have to live with their overweight and ideally not feel severe anxiety from it, which is the constructive point the article is making. The problem is when they try to cloud the evidence to achieve this goal.

This makes people believe that staying at a healthy weight or losing fat when overweight is not meaningful at all and you can be just as healthy while being obese as long as you take some walks and eat your greens. Which in turn means that we don't get motivation to do the things we need to keep our weight in check.

Now, this doesn't, of course mean that keeping a healthy weight is the ONLY thing that matters. But it's a pretty important one, in particular for those who are obese.

"Beyond the stuff like style, conventions, and obvious bugs/problems ..." Catching the obvious issues is IMO one of the main benefits of never skipping code reviews. Even trivial changes often have those (at least when I write them) and I've most likely saved many hours of debugging by a reviewer catching stuff like that. But this does require more than simple rubber stamping. I encourage my reviewers to read my code with the assumption that I make errors, basically looking for the bug that's fairly likely to be there.

The other thing I think is easy to do is simply see if you understand the code. If the reviewer struggle to do understand what's going on, chances are that other people will struggle as well. In this case, your lack of knowledge is an asset, the original author is likely to not see how it looks for someone not as well versed with the problem.

Based on the abstract, they did not show anything useful, they produce a null result, this just means that there weren't any effects large enough to be statistically significant for their sample. Weirdly, the still seem to draw a conclusion based on this null result which is really not correct. To draw the conclusion that they did they would have to show that the hypothesis that there is a clinically significant change can be discarded. Which is not mentioned in the abstract that they do.

Not claiming that the study itself is useless, it's important to understand how interventions affect completely healthy people, but my worry would be that it triggers eating disorders in high-risk group, that would affect wether we want to widely encourage calorie counting or just recommend it to those who specifically need it. The thread OP helpfully points out the fact that this study doesn't do much to answer that question.

The candidate probably does not know that though, so they might be writing an exhaustive list in the belief that you want to see everything, or a very short list in the belief that you only want to get a quick overview and will ask detailed questions in interview. Obviously you get to choose what criteria you use for hiring, but I can't imagine you won't be putting quite some noise into the process if you think of the resume as a personality test rather than their attempt to present things in a way they believe companies want.

In particular, the viewpoints that are downvoted differs from those that aren't. There's probably a lot of engineers that, like me, have learnt that HN isn't very fond of posts arguing for other perspectives in relation to big tech, and therefore will refrain from presenting them. Even if they don't, their posts will be hidden from view quickly enough to prevent them from affecting the overall tone.

I started wall climbing and bouldering with my GF. The time investment is that we go there two mornings a week for ca 2 hours, and some weekends for longer sessions. Living reasonably close to the gym makes it rather low effort.

I like it as it combines multiple skills such physical strength, climbing specific ability, balance, problem solving endurance etc. It feels like there's always new things to develop. Indoor climbing has the advantage of being less weather and season dependent than outdoor sports. Bouldering has the nice property of not requiring someone to do it with and you don't need to book anything in advance or follow fixed schedules.

That seems like a reasonable approach if you have the right skills for it. My problem with such an approach would be that I'd not feel comfortable to make an overall evaluation of a person based on a short interaction. Having a clear cut process with well defined metrics provides me with more confidence that I'm evaluating candidates fairly. Without it I'd be worried that my rating would be too biased by superficial stuff like how they dress etc. Something that at least in psychology research seem to mbe a lot more common than many people think. In general, looking too much at cultural fit would feel I'd risk "hiring my friends" rather than based on what value they could contribute.

In Sweden a significant portion of tech workers are unionized. It doesn't work well for creating leverage in term of wages though.

If anything, the impression I got at my last job where I was in the union was that the union negotiations mostly created an excuse for not increasing individual pay. "Sure, you perform really great and all, but if we give you more we'll have to reduce pay for someone else based on the agreed upon average increase."

For high-in-demand jobs such as SWE, the leverage of "if you don't maintain competitive pay I'll probably start look around for other work" is much more effective than union negotiations when no one is paid bad enough that they'd be willing to go on strike to get better terms.

Education is good and free enough that the supply of qualified labor would be hard to keep down without governmental intervention.

I think the objection is about the phrasing that can be interpreted as "I will only be wrong in at most 5% of significant results", while what we mean is "if there is no effect, we will only get so large differences in at most 5% of cases".

It's not uncommon that people fail to notice the nuance here and it's quite important for correctly interpreting reults.

Could you explain the calculations that lead to the claim that the result is not significant? From what I can tell, if we assume that clicking the ad is a weighted binary variable, eg. what is modelled as a "proportional distribution" there's a statistically significant difference between the two results. It's even pretty strong at P=0.006 (per https://epitools.ausvet.com.au/ztesttwo), but I might be missing something?

In other words, if I'm doing the math right, for him to p-hack this by rerunning the experiment in the case of no difference, he'd have to run it more than 100 times to get a 50% chance of getting as good or better significance.

There's of course plenty of other things that could be wrong outside of the simple statistical test, he could be making the numbers up, the groups might not be properly randomized etc.

I wouldn't hold anything against a person wanting to end their own life as I haven't been in their shoes. Barring that caveat, coupling your life to a dream does seem pathological - I would worry that someone ending their life for something like that are throwing away a decent chance to still end up with a overall good, fully lived life. It's important to consider that you can have rough patches in live but recover well. The largest tragedies are when people react on short term mental pain by giving up, where they would have changed their mind if just holding out for just a bit longer. The person who might feel like there's nothing to live for after having to give up their dream might just have their true love around the corner or maybe just a few session with a therapist would have given them a new perspective, finding joy in the moment, etc.

While I didn't downvote - my guess would be that your message was received as support towards the idea that suicide would be a reasonable and healthy reaction to not reaching your dreams - an idea that in itself could make someone consider the option.

I think sometimes "need more data" has a similar connotation as "citation needed". It's a way to dismiss an idea you don't believe in without being fully dismissive. In other words, it can be a way of saying "I don't think it's obvious that your idea will float, but I'm willing to accept it if you can provide evidence for it".

Of course there's a risk for bias here, you'll be more inclined to trust a close team mate than someone with just as good arguments outside your team which can lead to unlucky territorial behavior, even when intentions are good. But at least it leaves an opening for challenge that "nah, I think it won't work and I own the code, sorry" does not.

Very negative view in the article, but as long as they are open about it to the listener it doesn't seem like it has to be a bad thing. It will provide some incentive for artists to pick tracks that will be popular enough to compensate for the reduced per play-profit, if they do this well, it's also a track that listeners are likely to appreciate. Basically the incentives here align, and as long as they don't start removing the existing mechanisms for discovery, there's nothing forcing artists to opt in to this.

The "Rendering Effects" section describes it in some detail: "Once segmentation is complete, we use OpenGL shaders for video processing and effect rendering" and some info on what that covers. (OpenGL parts runs on GPU)

It is done on the CPU instead of the GPU. GPU would seem like the natural choice for a convolution heavy model but was not used here for the mentioned reasons.

It would probably be easier to read, but as a one-man project, my guess is that readability is less of a concern than for the average code base. Splitting code up in files requires some extra work and makes refactoring harder as you have to spend a lot of effort on trivial tasks such as moving dependencies between files and updating includes all over the place.

Personally I tend to want to split up code written by others so I can focus on the parts that go together without having to read through irrelevant parts, but I can easily navigate larger files/classes I've authored myself. I mainly split it up for the sake of future readers.

A very small fraction of experimental science is about just doing something and seeing what happens without a strong prior. You are expected to do your homework first. As a scientist, your role is typically to read and understand all the relevant prior work in the area first, then using this theory to derive a hypothesis for the outcome of your experiment. If you are lucky, the way the experiment unfolds might be outside your understanding, given that your understanding of the prior theory was correct, you have made a new scientific discovery which will be used to create new theory.

The fragment of uncertainty in this is typically very small and the process is very far from guessing your way through, hoping to find something new (whatever you found using such a method will most likely be either already known or incorrect).

I was inclined to write something similar myself as I agree in general. Mathematics is much more about the thinking than the doing, programming is more about the doing.

However, I do think that programming can be an excellent complement and aid when understanding certain concepts. Sometimes, writing a numerical simulation of a hard to grasp concept can make things fall into place. In particular in probability theory it can be quite helpful to generate a lot of random samples to validate and clarify some counterintuitive concepts.

I've used ZeroMQ lightly as a socket-replacement and for simple IPC and I'd say the key thing it solved was connection establishment. It turns out that opening a two way connection between two apps on the same network without a clear starting order can go wrong in many ways and become quite messy to get right. With ZeroMQ it more or less just magically worked and it allowed me to focus on the actual problem I was working on.

The wide range of language support was also nice. It allowed me to test the interface using a Python notebook which came in handy for debugging.

This happened to me on my last hire, I asked what I should work on and didn't get much direction at all. I think I wasted like a year doing the wrong things and strongly considered quitting during that period.

Then I figured that just as I got rather vague direction on what I should do, there also wasn't anyone telling me what I should not do. So I started doing things in the intersection of being interesting and beneficial for the product.

So my advice would be: If there's no concrete problem to solve, find one that you like and be glad that you were given the freedom to do so.

For people with vitamin D deficiency I would not be surprised if there's a connection to depression and I guess standard advice applies: If you think you might be at risk, get yourself tested. If nothing else, you can avoid experiencing the bone related issues that a vit D deficiency might cause.

This was the study I was thinking of (they did not target people with depression) https://jamanetwork.com/journals/jama/article-abstract/27689...

Weirdly I couldn't find any studies on the effects of supplementing with vitamin D for depression, even though there's plenty of studies on the correlation, and even one study where they used it for treatment, without any control group...

Great that you feel better! There are several potential explanations for this, for example: * It could be pure chance and selection bias, for you it seems unlikely, but consider how many took the same advice and saw not effect, they wouldn't post about it. This is why we need larger sample sizes and a control group to compare to. * It could be placebo, maybe your expectations made you notice and appreciate the happiness, maybe also strengthening it. This is why we need to make studies blind by using a placebo group. * There was indeed an effect because you were vitamin D deficient.

Recently a large study was published where they gave a large group vit D to look for depression incidence and they could not find any effect. There's a lot of parameters that could affects the results so it doesn't mean there is none, but probably that it's not as simple as "give everyone vit D and we get rid of depression".

[Assuming good faith despite some inflammatory language.]

In smaller teams everyone talking to each other and building a shared vision can be quite sufficient. It's indeed common to have startups that have very flat organisation, they might have a designated CEO that's in no actual way above the others. They can be plenty capable of producing amazing results despite the small team size.

However such a small group can't handle larger challenges, like designing and building a commercial jet, they will need many more workers. 1000 workers can't just come together and agree on a shared vision, even ignoring the square law of communication paths just giving each person 10 minutes to present their view would take weeks. Hence, in the very least we will need some communication hierarchy to allow workers to do their job while still synchronising with relevant other teams.

However, just communication is not sufficient, coordination is also important, if the jet is going to have the appropriate range, fuel tanks, engine efficiency and packing density needs to be designed to provide that range. Again, this would take quite some time if everyone should be involved in the process.

This is mostly in the leadership side of course, but covers those who are attentive to the trade off between security and deadlines. It's not reasonable for the individual engineer to be solely responsible to make that tradeoff. In particular, a sloppy engineer might think security is not important at all, and another engineer might think that no risk at all is acceptable to the point where nothing can ever be produced to achieve their standards. It's leaderships role to make sure that standards are aligned between engineers so we don't get a ten year delayed jet with some key parts made out of brittle plastic.

Now, managers can also have a people management role, and tom me those are the ones responsible to keep me as a worker happy and productive. They are also there to detect when some area is understaffed and find solutions to such issues, eg. talking with other teams for assistance, recruiting, decreasing scope, etc. If they weren't there, this job would fall on each individual worker, who would have to go out hunting for support instead of focusing on their actual tasks at hand.

Well, we don't know that we couldn't if we were able to easily take a blood test of the brain. I think there's two parts to the theory, the stricter interpretation that goes something like "depression is just the symptom of too little serotonin - which can be fixed by SSRIs" is quite weak. But the fact that brain chemistry affects mood, and in particular that increasing some chemicals can induce pleasant feelings is pretty robust - we have drugs like cocaine and heroin to thank for that insight. However translating that knowledge into meaningful treatments of depression is not straightforward as the body has a lot automatic control systems trying to keep things in balance.

So at one level we could claim that indeed depression is caused by "too low" serotonin/dopamine levels, as temporarily increasing them will take away the depression, but on the other hand this insight in itself doesn't provide us with a clear path for diagnosis and treatment, and it doesn't preclude that there's other things that's a root cause, such as thought patterns, neurological setup, etc.

> Certainly "releasing emotions" without a therapist could be somewhat dangerous for some people.

I think this in itself can be a dangerous thought. Agree that some people have dire need for help, an in particular some people might need medication for their condition to avoid getting into self harm behaviors. (Schizophrenia victims for instance.)

However, diving into emotions, trying meditation, and other things is not something to be considered dangerous and to be avoided without professional help. Being afraid of such subjects might in itself lead one to not be able to deal with them. Most people don't have a stand by therapist to consult every time they have some emotions bottled up.

Providing people with a structured approach to meditation/listening to emotions is not like suggesting to someone to do their own surgery. It's more like suggesting to treat a sore muscle with some tradition based massage.

Claiming that it mitigates the need for proper treatment for serious issues would be immoral, but proper treatment isn't always effective (most of the time it's not effective at all) and isn't always available.

They seem to believe in some magical distinction between disorders and non-disorders that I doubt can be established consistently in a meaningful manner and is pretty far from relevant. I think the meaningful way to think about mental disorders is that it simply means that a person have a slightly abnormal psychological condition that affects their quality of life in a significantly negative manner. The classification is helpful to get the right kind of help dealing with the condition.

That some abnormal psychological conditions can be reasonable reactions to stimuli isn't that relevant IMO. A broken leg is a normal reaction to hitting a tree when skiing, but we still need medical care when it happens.

If scientists then wants to discuss whether it should be called unable-to-walk disorder or a broken leg, I guess let them, but that discussion is just about terminology and unless I see concrete proof that it's more important than the treatment I don't think it's clinically relevant.

Could you elaborate? What is the problem with the linked project? Training a slightly faster, smaller and less accurate version of an existing model?

If you vote for politicians that invest in missiles, it's rather hypocritical to criticize people who develop the same missiles. If you happen to be a pacifist it's reasonable to be against all such technology, but the fact that clearly most people aren't makes it odd to single it out as non-grey area. (Unless it's about how it's clearly white-area as it's in service of democratically decided policy.)

Video codecs doesn't work like that, most frames only encode the difference from the previous frame, meaning that if you drop one frame, all frames after that will be impossible to decode correctly.

A technique that's used to get around this is to use special frame reference pattern so some frames depend on a frame two or four frames before, allowing the intermediate frame to be dropped. This comes at a significant cost to the encoding efficiency though. tT use it with a pre-encoded video, it would require you to specifically encode it for this purpose, making the usefulness of such a feature in a client questionable.

Nothing stops you from building a server that does this though, for instance if you want to build a scalable streaming service based on WebRTC.