Pangram (low false positive AI detector) tags this article as AI-generated: https://www.pangram.com/history/a520d0fd-335f-4cf2-8522-c608...
HN user
orborde
[ my public key: https://keybase.io/orborde; my proof: https://keybase.io/orborde/sigs/k_7Bf7nTi5BH6bDVUhNCAyIjs9eA5Ozdm6OdYpVEL3g ]
The final versions used for the tables/figures in the book are here, I think: http://www.bcaplan.com/returns.htm
Why do you still work with the useless partners?
This uses domain fronting, which both Google and Amazon forced Signal to stop using in 2018: https://signal.org/blog/looking-back-on-the-front/
Did cloud providers get more permissive since then?
EDIT: Tor also got hit by some shutdowns in 2018 due to its use of domain fronting:
https://blog.torproject.org/domain-fronting-critical-open-we...
https://gitlab.torproject.org/tpo/applications/tor-launcher/...
From the second link, looks like the plan is that Snowflake will annoy cloud providers less by only using the domain-fronting channel to propagate routing info:
sending Tor traffic directly through domain fronting (rather than using it only to distribute bridges and snowflakes) enables these platforms to claim that this technique is used by malware and therefore harmful to users, justifying shutting it down.
Snowflake is a more sustainable way for us to use the expensive but high censorship-resistance features of domain fronting as a low bandwidth bootstrapping channel.
The US adult obesity rate is 42.4% [1]. To get to 75% of deaths being obese people, you'd need about a 4x death rate of obese people compared to non-obese people. Most obese people find it extremely hard to stop being obese and often fail despite immense effort.
Meanwhile, getting a COVID vaccine reduces the chance of death by >10x at a cost of <$40 per person. My impression is that vaccines are unusually cost-effective medicine and that the low-impact medical spending is elsewhere in the system, but it is nonetheless thought-provoking to consider this specific example.
(OP here)
This essay set may superficially appear to be another repetitive salvo in the interminable US healthcare political conflict, but I recommend reading a bit deeper. I think the perspective these essays offer (that, at the margin, medical spending doesn't affect people's wellspan much, at least not in the US) is both quite important and underrepresented in most discussions of healthcare.
What specifically has changed since this essay set was published that would render it moot? If "the ACA", what specific ACA policies?
Isn't this strongly explained by the likelihood that Valley companies will start expecting everyone to come in to the office again this year?
What is a PWA?
As a company, why would I want to pay the hourly rate at all? Why not contract with a reputable bounty hunter, give them the level of access I'd give the hourly consultant, and pay the hunter bounties for what they find?
Seems like that captures the "higher bugs per hour" advantage of the consultant while retaining the "you only get paid for directly producing value" advantage of bounties.
Okay, so that's like 5/40 of your workforce that's in training at any particular moment, right? That's not a trivial cost.
There is effectively no reward for the rich to make this technology broadly available.
What about reduced training costs because you can keep your existing expert staff instead of needing to constantly hire fresh but clueless youngsters?
More concretely: Amazon's drivers (arguendo) don't care what they're getting paid $15/hour to do. If Amazon removes the incentive for antics, then it can pay 14/15 the per-delivery fee, and each driver can do 15/14 the deliveries and still make $15/hour.
Efficiency.
If a driver has to spend 4 minutes of every hour doing stupid antics to get an order, and can earn $15/hour doing antics+delivering, then effectively $1/hour is going to pay for stupid antics, from Amazon's perspective.
Amazon would like to switch those 4 minutes from "doing stupid antics" to "doing deliveries" so that they can pay the same $15/hour but get a full hour of real work instead of only 56 minutes.
I don't think so.
Consider that many governments with professional public health departments, including those of the richest countries in the world, downplayed the unfolding disaster, completely blew it on containment as a result, and are now reduced to using _extremely_ expensive lockdown measures to try to buy time until someone can come up with a better idea.
Meanwhile, plenty of randos on the internet did the straightforward epidemiological math in January/February and correctly deduced what was coming.
I think a clearer lesson is that, sometimes, the designated professional experts will fail to outsmart a fairly simple and very well known system of differential equations.
If you find this conclusion terrifying, well, I agree.
Would you prefer that the companies involved did nothing instead of taking this costless action?
You can deduct it, but there are a couple of distinct minimums in the tax code for how much you have to spend before it starts reducing your taxable income. ( https://www.insurance.com/health-insurance/health-insurance-... )
Also, for simplicity, I left out payroll taxes, which are invisibly paid by your employer directly to the government ( https://squareup.com/us/en/townsquare/payroll-taxes-defined ).
So the choice from earlier is actually more correctly stated as:
- Your employer directly pays $220/mo for your insurance, or
- Your employer pays you $200/mo as cash and pays $20/mo directly to the government in payroll taxes, totaling $220/mo. Then the government takes $40 (adjust as appropriate for tax bracket) out of your paycheck as income taxes, and you have $160/mo left over to pay for health insurance.
The payroll tax issue would apply even if you could fully deduct the $200/mo from your personal taxable income.
seems easy to fix. Why isn't that part of the health care discussion?
I am as mystified as you are.
One major factor in the US is that, from the perspective of an employee, employer plans have a massive tax advantage compared to anything the employee could buy outside their employer.
If your employer pays your insurance premiums, this doesn't count as taxable income for you. So you effectively have a choice that looks like:
- Your employer directly pays $200/mo for your insurance, or
- Your employer pays you $200/mo as cash, the government takes $40 (adjust as appropriate for tax bracket) out as income taxes, and you have $160/mo left over to pay for health insurance.
I can't really speak to the political forces that keep this subsidy in place, though.
https://www.taxpolicycenter.org/briefing-book/how-does-tax-e...
The characters of this article are so carefully constructed to evade this question that I have to assume malice.
Julia Wise and Jeff Kaufman are real people who are, so far as I know, pretty much as the article describes:
I have little experience with this matter, but it seems like the correct answer here is "get a job at another company that doesn't have broken procedures that render them unable to authorize paying for the people they need."
Internet access is more valuable than you surmise, I think. Having good telecommunications allows people to coordinate much better, and that efficiency bonus could easily create enough extra real wealth to help with the food, water, and jobs problems.
Heck, having better communications directly helps with food and jobs; you have a much easier time finding the guy who needs a worker or has a spare bag of rice if you're both online. Might even help with "clean water"; you can find information on how to make water potable and sanitation in general.
I couldn't tell you for sure whether getting smartphones to a village is a higher priority than installing a good well, but the smartphones are worth enough that I don't think that either choice is obviously superior.
Citation?
The story of healthcare.gov is that one set of federal contractors totally messed it up, and that another group of folks managed to get it back on track. It's not reasonable to claim that the latter group is somehow profiting unethically from the failure of the former.
Meanwhile, I have friends who work for companies who write the software that control train scheduling throughout the country. After what they've seen, they refuse to ride on trains, and from the stories they've told I don't blame them.
Can you tell us more about this?
Not exactly. I read "I know this because of all the money they give me" as somewhat tongue-in-cheek, like he's saying something he wants to believe, that the world tells him to believe, but cannot help questioning the truth of.
This could equally well have been done to an SHA1 implementation. MD5 is not especially susceptible to brute-force preimage attacks.
The security audit linked in the article [1] mentions that the researchers submitted several documents via StrongBox, but never got a response.
[1] http://homes.cs.washington.edu/~aczeskis/research/pubs/UW-CS...
Link?
Hat tip to simulate, who brought this up in a comment thread: http://news.ycombinator.com/item?id=4429021
I don't quite understand the purpose here. Is the idea that you can keep rolling your credit card debt forward interest-free indefinitely?