+1. I was excited for it and it sounded nice in theory but in practice it was just too much of a slog for me. I picked it up because I have a non-traditional math background and while I more or less know calc + LA, I do lack fluency with the concepts. Perhaps I am wrong, but at my level I feel that I don't need to go through 10 trivial variations of the same theme and many foundational concepts could be massively condensed. I am instead progressing much faster just picking up a textbook and referring back to precalculus books to fill in holes where needed.
Seriously? This article is obviously poorly written marketing copy and is just one step above the average InfoWars article.
"Caffeine anhydrous was first developed by the Nazis"
OK, so what? Not to mention the fact this is completely wrong. The first total synthesis of caffeine was described by Emil Fischer in 1895.
"Do the soda drink makers have to state how much caffeine is in the can?"
I have no idea if there's a law mandating it but literally every can of soda or energy drink I've ever had lists the amount of caffeine in it.
I have no idea if the assertions about the quality of the Chinese plants is true or not but there are so many things that are off the about the article The general alarmist tone, rehashing of standard chemophobic tropes (woah, _SYNTHETIC_ caffeine!), inserting some industrial mishaps related to the manufacturing of caffeine products (no logical connection is drawn between these mishaps and well, anything else in the article...) etc.
PC has spoken admirably of Koch Industries and he thinks there's some merit to remaining private. That's probably why. I don't have the source at hand but probably one of his podcasts.
I would like to second this recommendation. For a long time I thought their content was low quality edutainment because of the generally 90s website design and the generic stock photos used for courses(admittedly, that's...a pretty shallow line of reasoning). The courses are usually taught by university professors, have a well thought out lesson plan and have helpful accompanying guidebooks (you can also request course transcripts, I believe). For example, the course linked above is taught by Paul Zeitz based on his book, "The Art and Craft of Problem Solving", which is a classic for learning Olympiad-level problem solving techniques. I've also learned a lot from their history courses.
Nothing in our previous understanding suggests that they increase release of aerosols or aerosolization. Multiple previous studies have shown that they don't (see Asadi, S., Cappa, C.D., Barreda, S. et al. Efficacy of masks and face coverings in controlling outward aerosol particle emission from expiratory activities. Sci Rep 10, 15665 (2020), for one example; there's also older studies but you can look them up yourself).
The other arguments are of the "seatbelts cause more accidents" variety, which honestly aren't worth refuting. But I'll just note that social distancing has much weaker evidence of affecting virus transmission than even masks (there is no real justification for the 6 feet number and mostly unlikely it makes much of a difference). Hell, social distancing is even much more difficult to comply with than wearing masks so your argument makes no sense at all.
Yes, but they are a pretty good proxy for psychological discomfort levels/rates. And my point was that you can't really compare acne to covid as the latter is much worse (let me emphasize that you _don't_ have to agree with me on this, all you need to agree with is the fact that exponential risk >>>>>>> additive).
Furthermore, there is nothing to suggest that masks run the risk of worsening transmission. I see from your other comments that you're familiar with the mask clinical trials (presumably you're referring to the Danish study). That study was inconclusive (also note that it was _severely_ underpowered), and considering that _nothing_ in our previous knowledge suggests that they worsen infection rates, your reasoning is once again flawed.
Your logic is flawed because additive risks like that cannot be compared to exponential risks (like the spread of a virus). There's an asymmetry in the outcomes and the potential upside is enormous; especially when your stated risks are minimal to (say the least), like acne or in some cases, have been proven to be wrong by data (re: psychological "discomfort"--suicide rates decreased last year).
Honestly, this is a problem with everything. Snobs are sort of a double edged sword. They are great because they are very passionate and likely spend a lot of time on the topic at hand (could be anything. Maybe it's running, classical music, Buddhism, beer etc.). You can learn a lot from them. On the downside, since they are so opinionated they can be frequently insufferable. I think you have to learn to not take their opinions too seriously and continue to have fun.
Nobody disagrees with holding judges accountable but from the context, what you're advocating for sounds like a blood feud and we all know how those tend to turn out.
Yeah, basically this. I assume HN has a higher number of people who work in ML jobs in fields like finance etc. If you're working in any sort of social/public health research, then most new methods seem to be implemented as R packages. I'm thinking of things like new methods for propensity score, sequential trial designs etc. Also seems to be the preferred language on the Stats Stack Exchange posts.
Very good, something I have been thinking about as well. Inequality is bad because of the massive power differentials. How do you propose we "solve" that while still being able to reap the benefits of scale?
Any pointers or books to read about this topic? Admittedly I haven't looked too hard.
Why wonder? You can read in the original press release that there were ~5000 patients under the age of 65 with chronic diseases such as diabetes, severe obesity and cardiac disease. There were also ~7000 patients over the age of 65. That seems reasonable to me.
Sorry but this is ridiculous. Your solution to inmates having crappy lives is to impose even more arbitrary restrictions on them? This would make sense only if they could voluntarily go outside for a smoke whenever they want or go to a special smoking room or whatever.
Why shouldn't we have great recall of the details of books we read? All of the things you describe about actively engaging with the works you read are obviously correct (taking notes, summarising main points etc.). But often, I (and I dare say most people) fail to follow up with them effectively. As I see it, the author's question is: can we create a new medium which affords such active engagement with the works we read more effectively? Incorporating spaced repetition is a very good first step.
This is not a strong signal. The adjusted odds ratio is 0.03 (95% CI: 0.003 - 0.25). For randomised trials, it makes no sense to say that the difference in baseline is not significant because we have already observed the outcomes. It is better to use ASDs and in such small trials, small differences in baseline matter a lot. In this case it's pretty obvious the effect of vitamin D would easily be non significant if you shift around some variables. Reeks of p hacking imo
But in practice, directing animosity at advertisers and platforms is an effective way to remove content you don't like. It doesn't matter if the platform/advertisers says that they don't necessarily endorse the content--there's nothing stopping users from pressuring them regardless.
Yep. What you're getting at in a sense is "Knightian uncertainty", the presence of unknown unknowns (aka "black swans"). And the heart of the matter is certainty, which we cannot hope to achieve. So what exactly do these tools for handling risk help us achieve? I think this is still a question of active investigation and at any rate, if good answers do exist for it, they certainly haven't filtered down to the common masses.
Others have already raised this point... but let me try to reiterate. The problem of getting priors is not just one of "acquiring more information". In many cases it's not even clear what such a probability means. For example, you believe that Trump is the 45th POTUS...and you assign it a prior of 0.8...what does the probability mean in this case? In the case of rolling dice it's clear what each probabilities mean, but not in this case. And in any case, how much should you update your probabilities for any given piece of prior? All of these questions (how to assign priors, how much to update them etc.) are the _crux_ of Bayesianism and Bayesianism itself has little to say about it. The founders of Bayesianism itself were aware of these issues. For a more substantive critique read the following.
Bayes is a good _tool_ but to me it's a very small one and it doesn't and _cannot_ do most of the heavy lifting of how to live my life. Suppose I want to decide what I should do next week, Bayes is close to useless for that. And that is certainly something "critical thinking" should help me with.
Keywords to search for "small worlds vs large worlds Bayes"
Can't help but feel that people here are missing the point and perhaps too hung up on the specific example in the wikipedia article. The broader point is about the misuse and misapplication of technical/"rational" methods in real life i.e. not thinking about whether the model can actually be usefully applied in a certain real world situation. For example, blindly following rote statistical methods/not understanding their assumptions is what got us into the whole psychology replication crisis mess.