feral

12 ms·

61.
by feral 6y ago
They are making the argument that total perceived time would still increase logarithmically as absolute time increases, even under the model discussed. (And doesn't, say, approach some fixed sum.) Hence there's still a point to tr
62.
by feral 6y ago
I wasn't advocating being naive about conflicts of interest. But the solution to a conflict of interest is not to internalize your advisor's point of view! So don't think of them as your manager, in some traditional sense. Th
63.
by feral 6y ago
Nooo :-) If you find yourself in a PhD program with an advisor who thinks like the above, it's really important to not internalize their viewpoint. It's not a job. A job would pay much better! It's not your advisors money, an
64.
by feral 6y ago
There's a big difficulty difference between Engineering, and Reverse Engineering. If you want to build a supersonic fighter, it's a lot easier if you can look at how someone else has done it. This focuses your search to certain pa
65.
by feral 6y ago
I work on such a bot @Intercom. A lot of support requests aren't actually intricate and special: there's almost always loads of simple requests that come in again and again. When you ask a request like that and get the answer ins
66.
by feral 6y ago
There's roughly about 1.3M downloads out of an adult population of 3.7M people, so about 35% of the eligible population.
67.
by feral 6y ago
Probably. I would say that depends on the overall strategy. If pre-symptomatic people infecting others is a big issue, then it's probably worth the cost of quarantining folks who display no symptoms until you are sure they aren't
68.
by feral 6y ago
You may be misunderstanding the purpose of contact tracing? It's not to guess whether a patient you have is infected. It is to find the people that patient has infected before those infect more people. If things are being run properly
69.
by feral 6y ago
This is a weak argument from Schneier. Most of his concerns (false positives, false negatives) also apply to contact tracing done by humans, which he advocates at the end of his article. If a medical professional interviews you about your c
70.
by feral 6y ago
That's a correct understanding of specificity. That would imply the correct rate is closer to 22% than 30%, as of the 64 positives out of 200, we would expect roughly 20 of those to be false positives. (So the remaining 44/200=&gt
71.
by feral 6y ago
Yeah - actually worse than wartime level, for a lot of wars. If you look at economic indictors for ww2 for the UK or USA they are actually less bad, less sudden, than this lockdown. Also if you think about the restrictions placed and the nu
72.
by feral 6y ago
The simple math comment was about estimating IFR, given the sampling in the Netherlands, not about the whole domain. Maybe you are saying the implied model is too crude. Perhaps, but there's a value in it's simplicity - folks can
73.
by feral 6y ago
Humanity's understanding of this disease is much more sophisticated than the two comments above this would imply. Researchers are not stupidly looking at Diamond Princess and getting confused because the population is old. Rather it&#x
74.
by feral 6y ago
The Netherlands has 3.3k deaths vs 29k confirmed cases, which would give a CFR of ~10% I.e. the Netherlands aren't testing much. A 10x reduction there brings it back to the CFR range we see in countries with bigger testing programs. I&
75.
by feral 7y ago
On the 'essential workers' bit: Worth being clear that it is not necessary to completely eliminate every case of the virus. If the natural r0 is 3, and you eliminate roughly 3/4 of each person's contacts, we've won.
76.
by feral 7y ago
Otoh, the epidemic in the US is likely growing exponentially, and that encourages the current CFR to underestimate the actual CFR as new cases haven't had time to die. In terms of estimating the IFR, it's hard; estimates are about
77.
by feral 7y ago
Someone always says this. And yet here we are with a virus that has an r0 much higher than flu, and is ~10 times more lethal. That the two things are correlated in certain modeling scenarios does not mean you can't end up something tha
78.
by feral 7y ago
I'm guessing that the UK has decided this can't be practically contained, that even total shutdown will only push it back a couple of months with huge cost. As such they are just going to accept mass casualties. See figure 4 at th
79.
Predicting the Impact of Coronavirus on Ireland(medium.com)
3 points by feral 7y ago|0 comments
80.
by feral 7y ago
I work as Principal ML Engineer, and did a little epidemic modeling in my ML PhD. I wrote an analysis of how this is going to hit Ireland, including the best calculations I could find on fatality rate: https://medium.com/@fe
81.
by feral 7y ago
Cruise ship analysis is naive. People from the cruise ship are still in the process of dieing. 2 more deaths occured today. The number of recovered cases from the cruise ship is low, so naive inference on death rate so far is misleading.
82.
by feral 7y ago
Facebook is an outlier monopoly, having successfully privatized a massive commons, so I'm not sure their ability to spend 40B a year proves much of anything. (If those figures are even the right lens.) As a counterpoint, Wikipedia&#x27
83.
by feral 7y ago
Contarian opinion: We are great at software engineering, and our field is the best at engineering generally. Engineering isn't about how reliable or high quality the system is. Engineering is about balancing the costs and the benefits.
84.
by feral 7y ago
That says 7 percent of deaths are from influenza, not that 7 percent of people with influenza die.
85.
by feral 7y ago
I'm sure he has, he's a very famous economist. And you didn't need to copy-paste all that, I had read the article. I stand by my post, the claim in the article is exaggerated. No intention of taking it up with him, I'm s
86.
by feral 7y ago
>first comprehensive Supply/Demand curve in the wild That's a huge claim. Anyone doing dynamic price optimization is doing this, long before 2016.
87.
by feral 7y ago
The progress that has been made so far is already good enough to deliver tons of real business value. The tech is way ahead of application, as the tech has jumped forward so much in the last 5 years, and progress continues to be rapid. I&#x
88.
by feral 7y ago
You are thinking of 'Neverwhere', by Neil Gaiman!
89.
by feral 7y ago
As you appeal to Programmers: do you believe quicksort is a general sorting algorithm? Would it still be a general sorting algorithm if someone had just developed it and used it solely to sort numbers between 1 and 100? It would be. Why do
90.
by feral 7y ago
We really need another name, which refers to people uninformedly quoting 'the AI effect'. How about "the AI effect effect"? This happens all the time: - X suggests that success at a particular task is a good proxy to gen

More ›