currymj

32 ms·

421.
by currymj 6y ago
should the university system really just be an organized academic competition for undergraduates? I think the solution is as you say -- make assignments that are hard to cheat on -- but also reduce the incentive to cheat. Just make it less
422.
by currymj 6y ago
this seems very nice -- very polished and gives good results. I would like it if the bibtex entries had meaningful cite keys as opposed to long numbers. as is, it would be pretty difficult to actually write a paper using these bibtex files.
423.
by currymj 6y ago
you borrow a share of stock from someone (like literally they loan it to you), then go sell it and keep the money. eventually, you go to the market, buy another share of stock, and give it back. if you sold high and bought low, you make mon
424.
by currymj 6y ago
Elizabeth Holmes founded a startup a while back to solve this problem... It seems like it’s just tough because you generally need a fair amount of blood for each test. We could lose some of the bureaucracy but it seems hard to get away from
425.
by currymj 6y ago
it’s a great sign that someone at Apple cares about this; hopefully they eventually support PyTorch as well. there are a lot of use cases (smaller networks, non-deep-learning numerical algorithms) where having acceleration on a laptop is re
426.
by currymj 6y ago
The Ivies and comparable schools actually are now incredibly generous with financial aid. For instance Harvard automatically gives a full scholarship to any accepted student whose family makes less than $65,000 per year. For Stanford, the c
427.
by currymj 6y ago
one of Nassim Taleb's better ideas is that you should pick the worst-looking doctor you can find -- weird-looking, fat, lower-class accent, whatever -- because if they're successfully practicing medicine, they must be very skilled
428.
by currymj 6y ago
think of it like shell scripting for statistics, although not nearly as limited as bash is compared to other programming languages. it works best if it's used semi-interactively, as a glue language between statistical packages which ma
429.
by currymj 6y ago
i like gather.town. a totally frivolous suggestion i would make is that it should be possible to pick up and eat the coffee and cookies. you could have a meaningless status effect of some kind.
430.
by currymj 6y ago
"Universal consistency" is longstanding terminology for "error goes to zero under every possible distribution". So it seems reasonable to use the word universal in this way, referring to a "for every distribution&qu
431.
by currymj 6y ago
it is absolutely true that “Democrat party” as opposed to “Democratic party” is a Republican shibboleth for some reason. Democrats themselves don’t seem to think about it much but there is a portion of conservative media that is very carefu
432.
by currymj 6y ago
they'd let you get away with it, you wouldn't go to prison. i'm pretty sure individuals used to take way more weird writeoffs. but mostly these have been closed over the years for individuals. the laws are generally enforced
433.
by currymj 6y ago
whatever other non-caffeine effects coffee might have, they are probably neutral or positive (unlike cigarettes).
434.
by currymj 6y ago
this is some great science reporting — clear and accurate.
435.
by currymj 6y ago
note, for people who were briefly confused like me, that to be considered polynomial time, it has to be polynomial in the number of bits, i.e. log(N).
436.
by currymj 6y ago
US government does sponsor a lot of anti-censorship tools of that type: Tor was funded by the Navy and DARPA, Signal received initial funding from the same organization that runs Voice of America.
437.
by currymj 6y ago
there's gotta be something weird going on here. either they are rolling "tuition" waivers into this, or they are quoting salaries for 40 hours but the actual positions are paid at 20 hours (on paper). as a more typical exampl
438.
by currymj 6y ago
This doesn't seem quite right to me. $35k seems like a big overestimate, outside of very high CoL areas or the richer private institutions. Top-ranked public institutions will offer more like $20-25k because they can get away with it d
439.
by currymj 6y ago
i think it might just be a misunderstanding then — i assumed no state funding of the humanities would mean completely eliminating those departments at public universities.
440.
by currymj 6y ago
i think you are failing to see the value — much of what the academic community produces is a public good. i agree there are huge problems with higher education, but I think if all humanities departments were disbanded as you suggest, you wo
441.
by currymj 6y ago
The science departments produce even more nonsense that nobody reads or cares about. It's hard to tell in advance. Plus you want a strong research culture even if not all of it is directly productive. I also think people tend to undere
442.
by currymj 6y ago
that’s not how things work. lots of well-regarded papers are initially rejected in peer review but are commonly discussed and cited while in the submission process at other venues.
443.
by currymj 6y ago
this is not really the norm in computer science. it’s generally expected to cite things on arXiv although it is more of a grey area.
444.
by currymj 6y ago
mostly it seems that it tells you something sort of weird about how neural networks train. it’s not obvious that this should work, and that it can be made to work is interesting.
445.
by currymj 6y ago
It used to be more exciting. Overtaking is particularly exciting but there is less of it now, because being behind another car means turbulence and modern cars are too aerodynamically sensitive. Also, nowadays a few big teams dominate and t
446.
by currymj 6y ago
this article makes many valid points but I would have loved to read a version that was written without use of the terms "AI" or "artificial intelligence", as I think it would have been clearer and more interesting.
447.
by currymj 6y ago
auction theory is regularly tested by people running auctions with billions of dollars at stake. of course the assumptions of the theory are never perfectly met in the real world, and things can get quite messy. but for instance, the theory
448.
by currymj 6y ago
it’s very cool work; i wish we could ban the use of “AI” as a noun. it only ever confuses people.
449.
by currymj 6y ago
the real point of the store greeters is to limit shoplifting -- they hang around the entrance watching people and may check people's receipts against what they are taking out of the store. making them say hi and act friendly to everyon
450.
by currymj 6y ago
i have to imagine it would piss Nate Silver off to experience 4 years of continual mockery for "getting the election completely wrong" when he in fact got it right. anyway, I agree the cartoon fox seemed very condescending and I w

More ›