currymj

44 ms·

541.
by currymj 6y ago
I can't think of any of those from recent history. Medicine has gotten pretty good. Exceptions may be in mental health. Somebody else mentioned psychedelic drugs, which might be FDA-approved for some conditions soon, but all official m
542.
by currymj 6y ago
i mean, these kinds of simple CS201 algorithms are still at the core of a lot of things that definitely count as AI by any reasonable definition. In particular, the best superhuman poker-playing systems use counterfactual regret minimizatio
543.
by currymj 6y ago
Artemisinin is an antimalarial drug derived from a plant used in traditional Chinese medicine to treat fevers, including malaria. First discovered and isolated in 1972. St. John's Wort started out as alternative medicine but is now off
544.
by currymj 6y ago
no false positives (except for rare bugs); it will usually either return some partial solution (like a simplified version of the input), or else just keep running for a very long time until you give up.
545.
by currymj 6y ago
There are three issues I can think he might be talking about: 1) if you get a solution, can you be sure that it's right? in this case, the neural network might spit out a wrong solution to the integral. However, this solution is easy t
546.
by currymj 6y ago
This is how many CS conferences already work (even with public anonymous peer review, go to openreview.net to see it in action). It does seem to be better, although the frustration level for researchers is not really less. Edit: also, in CS
547.
by currymj 6y ago
i just wish they could move it onto a different domain that doesn’t end with “nih.gov”. People can go on PubMed, find some bad article about 5G or fluoridation from a dubious journal, and then send it around social media. The nih.gov TLD g
548.
by currymj 6y ago
sadly, CS is also about the reviewer lottery. however because it is conferences rather than journals, turnaround is much faster at least. and there is a culture of preprints, so having a paper rejected once doesn’t delay getting it out into
549.
by currymj 6y ago
they can't just arbitrarily jack up the rent during the term of the lease. but they might when you try to renew the lease, or when you move out. the effect is as if there is no rent stabilization (unless market rents go up above $4000
550.
by currymj 6y ago
there have been complaints by serious people about the work in this book, see https://statmodeling.stat.columbia.edu/2019/12/27/why-we-sle... now, it's obvious to me that if I don't get 8 hours of s
551.
by currymj 6y ago
i forget who, but I remember reading someone who both writes and does podcasts saying that one of the nice things about audio is that you can't have your words screenshotted and mocked on social media. which means you can express yours
552.
by currymj 6y ago
In Bayesian terms, ridge regression is equivalent to putting a Gaussian prior on the weights, which gets wider the weaker the penalty term. The lasso is the same, but with a Laplace distribution as the prior. The "elastic net", wh
553.
by currymj 6y ago
In general math departments in the US are less frightened of accidentally teaching the students something useful.
554.
by currymj 6y ago
i somewhat dislike this idea. it seems like it treats scientific papers as independent, isolated containers of reliable facts which can be reliably and authoritatively cited without any need for context. This is a really common way of readi
555.
by currymj 6y ago
oh, also -- the advice to skim early chapters is generally good. In general, don't be afraid to skip past things you don't understand, as long as you have some way of testing that you are learning everything you need in the end (v
556.
by currymj 6y ago
if it's something you're self-studying, don't be afraid to put something down and pick it up later. There are topics and materials that I completely failed to understand the first time I encountered them, but months or years
557.
by currymj 7y ago
they probably do, often the university negotiates royalties or equity for spinoffs of research done using university resources.
558.
by currymj 7y ago
It simulates a simple model of a quantum computer, so that you can implement and test out quantum algorithms, albeit at potentially huge computational cost. quantum.country is a good introduction to quantum computing for people with no back
559.
by currymj 7y ago
i think it’s just because they want to publish, and as you say it’s not easy to find a really good publicly available dataset or simulator. I think if these were publicly available and there were a Python package, people would rapidly get i
560.
by currymj 7y ago
totally possible to learn a lot of economics and statistics and never have formal coursework in numerical methods for optimization.
561.
by currymj 7y ago
i mean for many classic ML models you have this — literally big O notation for the number of samples required to learn, number of iterations to convergence, cost per iteration, etc. You even get the cases where the algorithm with the best b
562.
by currymj 7y ago
i'm really curious what state this was in. some states are quite lax both legally and in terms of public opinion (e.g. Wisconsin) but most are pretty harsh.
563.
by currymj 7y ago
in the actual book, Gibson's vision of the future is basically as you describe -- it's extremely weird and alien with little mapping to our current political system. meddling in past timelines is an esoteric hobby for the super-ri
564.
by currymj 7y ago
there are a number of video games by ZachTronics in which you program in assembly language for fictional computer architectures in a sci-fi settings. “Shenzhen I/O” and “Exapunks” are both fun.
565.
by currymj 7y ago
https://www.rageon.com/a/users/invisibilitycloak if you read the article you'd see that they only work sometimes though.
566.
by currymj 7y ago
when they're generated digitally, it's surprisingly hard to get good robustness against adversarial examples, and people have been working on it for a while. of course you can train to stop any specific input, but then new ones ca
567.
by currymj 7y ago
there is some truth to this but some areas of math do really benefit from just raw memorization. At a certain point, remembering the definitions and the statements of theorems may not be so trivial, and having them in your head makes it muc
568.
by currymj 7y ago
so the days of the week in Spanish and Japanese have the same roots... that's wild!
569.
by currymj 7y ago
there's a few points where people tend to give up on it. i myself stopped reading it twice at the giant cocaine adenoid, and again when the man chases the harmonica into the toilet. never finished it.
570.
by currymj 7y ago
you ever read "The Varieties of Religious Experience" by William James? It's a good book despite its late-19th-century understanding of the brain and nervous system. In talking about individuals who have had profound religiou

More ›