3 ms·

Oh boy, I just wasted my time reading this, and I really wish the headline would have flagged that this is written by a "site reliability engineer" and not an e

by dmode 6y ago

Oh boy, I just wasted my time reading this, and I really wish the headline would have flagged that this is written by a "site reliability engineer" and not an epidemiologist with credentials. Feel like there would be a revolution in HN if this was the other way around - an epidemiologist writing about site reliability engineering after reading a few articles for 2 weeks. Should we not exercise restraint, thoughtfulness and humility instead of writing commentary and articles in areas where we have rudimentary understanding ? It is really an insult to experts in this area who have studied viruses for decades, have considered a variety of factors, have conducted academic studies, and have build complex models, all to be outdone by a "site reliability engineer". As for the article itself, it is so full of misinformation, oversimplification, and hyperbole, where should I even begin. 1) Claiming absolute authority on fatality rate as 3x of influenza, when global experts have yet to agree on a number [https://www.nytimes.com/2020/04/17/us/coronavirus-death-rate.html https://www.nytimes.com/2020/04/17/us/coronavirus-death-rate...] 2) Then basing the article on an absolute fatality rate, when the fact remains that fatality rate depends on a number of different scenarios include health, immunity, age, hospitalization, ICU beds, and may be even weather etc. [2] To lump all of that into a single number simply hides the cost of COVID. Just in Michigan alone, 35% of COVID deaths are within the black community. Not having a grasp on how this impacts along racial and economic lines demonstrate a very rudimentary understanding of public policy. Ask yourself, if the fatality rate is 1% on average, but 10% for black people, will you still call for opening the economy [2] https://fivethirtyeight.com/features/why-its-so-freaking-hard-to-make-a-good-covid-19-model/ https://fivethirtyeight.com/features/why-its-so-freaking-har... 3) Repeated references to comparison with influenza, but not mention of the massive R0 differences between the two [3]. Influenza has a R0 of 1.3, while COVID is estimated to have an Ro of 5.5. And this matters, because now COVID is not only more fatal, but is also spreads faster. Both the denominator and numerator are deadlier for COVID, which is a compounding factor that the article doesn't discuss at all. if 20% of US population get impacted, we will have a death toll of 600K at 1% fatality rate. That will be 10x than that of the worst strain of influenza. [3] https://wwwnc.cdc.gov/eid/article/26/7/20-0282_article https://wwwnc.cdc.gov/eid/article/26/7/20-0282_article 4) There are a lot of reference to negative externalities of lockdown [without much data to back-up], but no mention of the positive ones. Such as plummeting car fatality rates, plummeting crime rates, and plummeting pollution [4] [4] https://www.nationalgeographic.com/science/2020/04/pollution-made-the-pandemic-worse-but-lockdowns-clean-the-sky/ https://www.nationalgeographic.com/science/2020/04/pollution... 5) A strange out of context reference to social isolation and related deaths. The linked article and current situation are very different. People are isolating from others, but they are very much spending more quality time with their loved ones. I don't see any data on spike in depression during this period, and anecdotal evidence suggests that people are refreshing their relationship with their loved ones. 6) Section 4 presents a half baked plan of "opening" up without broader discussion of trade-offs and implications. For example, opening up but encouraging "at-risks" group to self quarantine. This is not how US workplaces work. Let's stay I am a at-risk worker with chronic lung disease, but I also work at a factory floor and my health insurance is through my employer. How am I supposed to self-quarantine when my employer will simply ask me to show up and if I don't I will lose my job and insurance ? And there won't be any government dollars to support 7) There is no discussion on how we can support the re-opening without ability to test people at scale. How will you know that the server at a restaurant or your gym instructor is not infected without massive upgrades to testing infrastructure ? These are some basic observations from the article. I can keep going as every paragraph in this article has some questionable conclusions. I also feel the "soul of the nation" feels very hyperbolic.

lbeltrame 6y ago
> I really wish the headline would have flagged that this is written by a "site reliability engineer" and not an epidemiologist with credentials. (Note: not the author). I think any article like this can attract (and even deserve, depending on what's presented) criticism. But I absolutely dislike the "lack of credentials" as an opening argument. It's much better to tear away the points you don't dislike, IMO, without mentioning credentials. It's obvious that being in the field can give you much better understanding of the issues, but not being in it does not mean one will not try to reason over the data (even if wrongly). I'm a biologist (well...half biologist, half computational biologist, and something else thrown in the mix). Does that mean I can not criticize (hypothetically) something on epidemiology because I don't have credentials, or, conversely, not try to figure out something in the data (even if my reasoning is totally wrong) because I don't have the required credentials? I understand the above is a little too black-and-white, but I mean, let's encourage debate, even strong disagreements like your post, without going over what the author is doing in the first place. P.S.: (not a jab at you, but a thought that emerged while writing this reply) I believe there are a lot of models floating on SARS-CoV-2, but I feel we also need loads of basic, bench level experiments to understand it better.
- whatisthiseven 6y ago
  I don't get it. Imagine an HN where random people who never did any software started flooding HN with articles about how the software people are lying to us all and just making big bucks. Software is easy, here is how to make great software. And every single one of them was just random near-garbage pulled from a couple minutes reading slides from Agile experts and the occasional wiki article about "software abstraction". It would be downvoted so fast for being so wrong, and no one would engage with it for being so terribly wrong because disproving all the points is exhausting and not worth the time. Yet, you would think it OK on HN. Maybe you downvote some of the wrong ones, but others get lots of comments, and soon all these wrong posts just get more attention and comments as people are arguing over half-made points in a blog post by a sales guy that "has a few software friends". The long-term changes to HN would eventually push away people that care. But we don't live in that world, maybe not yet. So instead we allow people who get so many basic facts wrong about an epidemic to post on HN, and @dang explained elsewhere why this makes sense (and I agree from a moderating standpoint to keep this content up) and I agree with it. But we really only allow this content because it isn't the bulk of content on HN. If the bulk of content on HN resembled the above, I don't think we would like the community HN becomes. Quite ironically, this is highly related to the virus itself. The virus isn't a problem on its own. It is a problem because it infects hundreds of millions of people very quickly, and we have poor means for slowing or stopping it at all. Very much like misinformation.

by dmode 6y ago

lbeltrame 6y ago

> I really wish the headline would have flagged that this is written by a "site reliability engineer" and not an epidemiologist with credentials. (Note: not the author). I think any article like this can attract (and even deserve, depending on what's presented) criticism. But I absolutely dislike the "lack of credentials" as an opening argument. It's much better to tear away the points you don't dislike, IMO, without mentioning credentials. It's obvious that being in the field can give you much better understanding of the issues, but not being in it does not mean one will not try to reason over the data (even if wrongly). I'm a biologist (well...half biologist, half computational biologist, and something else thrown in the mix). Does that mean I can not criticize (hypothetically) something on epidemiology because I don't have credentials, or, conversely, not try to figure out something in the data (even if my reasoning is totally wrong) because I don't have the required credentials? I understand the above is a little too black-and-white, but I mean, let's encourage debate, even strong disagreements like your post, without going over what the author is doing in the first place. P.S.: (not a jab at you, but a thought that emerged while writing this reply) I believe there are a lot of models floating on SARS-CoV-2, but I feel we also need loads of basic, bench level experiments to understand it better.

whatisthiseven 6y ago
I don't get it. Imagine an HN where random people who never did any software started flooding HN with articles about how the software people are lying to us all and just making big bucks. Software is easy, here is how to make great software. And every single one of them was just random near-garbage pulled from a couple minutes reading slides from Agile experts and the occasional wiki article about "software abstraction". It would be downvoted so fast for being so wrong, and no one would engage with it for being so terribly wrong because disproving all the points is exhausting and not worth the time. Yet, you would think it OK on HN. Maybe you downvote some of the wrong ones, but others get lots of comments, and soon all these wrong posts just get more attention and comments as people are arguing over half-made points in a blog post by a sales guy that "has a few software friends". The long-term changes to HN would eventually push away people that care. But we don't live in that world, maybe not yet. So instead we allow people who get so many basic facts wrong about an epidemic to post on HN, and @dang explained elsewhere why this makes sense (and I agree from a moderating standpoint to keep this content up) and I agree with it. But we really only allow this content because it isn't the bulk of content on HN. If the bulk of content on HN resembled the above, I don't think we would like the community HN becomes. Quite ironically, this is highly related to the virus itself. The virus isn't a problem on its own. It is a problem because it infects hundreds of millions of people very quickly, and we have poor means for slowing or stopping it at all. Very much like misinformation.