cfcef

Forced-choice is deliberate because it maximizes the information of the question. If you put in intermediate options, that can mask real differences by letting peoples' differing preferences for certainty influence whether their responses are informative or not. (People have information or preferences they don't even realize they have, like in some perceptual tasks, and you can't detect this without something like a forced-choice because if you give subjects any other options, they'll just say they don't know.)

If you're reading files rather than writing them, then you must be decrypting them and in possession of the private key, which means you are permitted access to everything; so you could cache all the decryptions. Initial reads might be more expensive to locate, but then free. Also, how many cycles could it take to decrypt the few hundred or few thousand characters that could possibly make up a full filepath?

About time. Cryptsy's end has been more telegraphed and expected than Mtgox's was. The site had many bugs in basic functionality long before the hack.

Perhaps I'm missing something obvious, but why can't the filename be encrypted as well?

(So you have public key 0xDEAFBEEF; you want to write a file named 'secret.txt' with the contents 'We attack at dawn'. OweFS encrypts 'We attack at dawn' to 010101 and writes that to 'secret.txt'. But why couldn't it have encrypted the contents to 010101 and encrypted the filename 'secret.txt' to 111000, and then written a file named 111000.encrypted with the contents 010101? Then when the owner of 0xDEAFBEEF wanted to read it, he simply decrypts 111000.encrypted to 'secret.txt' and decrypts its content 010101 to 'We attack at dawn.')

Third sentence of OP: "Google has also been testing its autonomous vehicles in the snow."

They don't necessarily have to, that's true. But if there was any appetite to have gargantuan 500MB+ decompression libraries to save 20 or 30% on downloaded bytes while browsing, we would have seen much more uptake of existing simpler compression schemes like SDCH which takes a tiny step in that direction with a relatively large (but still tiny) pre-built dictionary for the WWW.

> the results here are extremely poor compared to modern techniques.

Could you link a demonstration of the much superior results?

Asymptotically, yes; prediction = compression (if you have a model for a bitstream which produces probabilities over the next bit, it can be fed into an arithmetic encoder and you now have a compressor). In this case, it's not practically helpful. A VGG is 528MB all on its own, so you need to compress a lot of images to make back that 0.5GB use plus runtime dependencies.

So the problem with this gambling site, which defends itself as being a game of skill, is that some players are far more skilled than others and so win regularly. I see.

You'll pick up a lot of post-WWII differences, but you can still try. There's an interesting paper exploiting the meandering of river-defined national borders to do a discontinuity design for long-run effects of the Austro-Hungarian Empire, for example, which is at least not obviously wrong.

Of course C relies on an OS. The point of an OS is securely mediating access to resources; a C program is going through the OS, using syscalls, in order to do stuff like 'write a file'. Which bears a certain resemblance to what this code is doing...

Typical epi-nonsense: ignore the huge and omnipresent genetic and other confounds which lead to long pre-existing differences, find congenial event to blame as origin, then leap to the most exotic methods of transmission possible; bonus points for each mention of Holland, rats, or mice.

The South as a region is very distinct from the rest of the country: settled by different ethnic groups (no, not all 'whites' are genetically identical twins; consider the different genetic BMIs by country in http://www.nature.com/ng/journal/v47/n11/full/ng.3401.html ), different cuisines, different industrial patterns, different weather and climate... It is eyebrow-raising, to say the least, to bring in the Civil War as at all relevant and speculate even more wildly that it has intergenerational effects.

If they want to talk about that, they should be trying some sort of regression-discontinuity or other natural-experiment design involving borders or Civil War activity, not, effectively, saying 'look at how different the South is! the Civil War?!' Uh, yeah, I think we all know the South differs. That's not very helpful.

Another question: is the old code even valid in any sense compared to the current code? Since the original was written before 1985, then 99.9% of the people running Emacs will be running it on an OS that did not exist in any form in 1985 (Linux, Mac OS X, Windows, Android, etc) and the rest will be running on OSes which have changed substantially or been replaced entirely since then (SunOS->Solaris).

I don't think that's an answer but passing the buck. Why are they separate in the first place? Why aren't record labels, as part of standard modus operandi, been buying copyrights from bands? They are best placed to do this sort of thing, much better placed than random musicians who happen to get rich and then dabble in investing - record labels have (or had) the contacts up front, they had the money, the diverse portfolios to spread risk...

> In the Stanford lab, researchers have found even a single experience can affect people. In one test, researchers found that test subjects who had just cut down a virtual tree used fewer paper towels when cleaning up a spill after the experience than subjects who hadn’t cut down a tree.

Oh great. More priming studies.

What I'm wondering reading this is why song rights were so profitable in this time period. Why were McCartney and Jackson making so much money and why didn't the record labels keep the rights locked up tight if the rights were so lucrative long-term?

Very similar to patio11's writeup. Note the same caveats apply, more or less: this is working at a big Japanese company, which is far from all of them.

A better link might be http://flatassembler.net/docs.php?article=design which helps explain wtf this is and why anyone should care.

See the comments: lots of tweaking and unreliability and not guaranteed to work in emails, apparently.

They say it would be illegal and unethical for them: https://www.reddit.com/r/MachineLearning/comments/3yy0vp/usi...

Could be Instagram, judging by that recent security writeup...

As far as ignorance goes, check out this fascinating quote from the judge:

"In fact, as plaintiffs point out in their submissions, Deputy Burns testified that, at the time he tested the plant material he discovered in plaintiffs’ trash, he had no knowledge that anything other than marijuana could test positive on a marijuana field test kit and that he was “not aware” of the possible occurrence of false positive test results. Deputy Blake’s testimony on those issues is substantially the same. Moreover, Sheriff Denning testified that, since he came to the Johnson County Sheriff’s Office in 1978, the Office has conducted “thousands” of field tests and the only false positive results of which he is aware are the results at issue in this case."

Apparently the judge is under the impression that this justifies his ruling, rather than damns both him and the entire police department.

"Levine easily admits that 16 and Pregnant is unlikely to be the entire story. One data point his theory has trouble explaining is the steep decline in teen pregnancy rates between 2012 and 2013, long after the MTV shows' ratings had peaked.

"The continuing, rapid decline is the part I find a little confusing," he said. "The recession and the show get you a lot of the way through 2011 and 2012. But I was a bit surprised by the most recent, rapid decline in the 2013 data."

http://www.vox.com/2014/8/20/5987845/the-mystery-of-the-fall...

That output doesn't look good, especially compared with the Obama char-rnn output: https://medium.com/@samim/obama-rnn-machine-generated-politi...

Perhaps the Trump corpus could be fed into char-rnn for comparison and figure out why the output is so nonsensical?

(Noka isn't the Mast Brothers, in case anyone is confused, but an earlier boutique chocolate seller who was employing the same basic scam.)

That is an impressively thorough and readable set of posts by Scott. I don't care that much about chocolate, but it's still a fascinating takedown.

Borowitz is not as clever as he thinks he is, and the attempt to make a satirical analogy here in imaging Shkreli hoist on his own petard fails because lawyers lack the defining feature of drug patent law which made Shkreli's notoriety possible: if you need a particular patented drug, one company has a monopoly on it; in contrast, there are many lawyers you can hire to defend yourself. The analogy simply doesn't hold.

You might enjoy http://genekogan.com/works/a-book-from-the-sky.html https://news.ycombinator.com/item?id=10750252 as a perspective on automated font construction.

Looks like Lasagne (https://github.com/alrojo/lasagne_residual_network) and a stab at Keras/Theano (https://github.com/ndronen/modeling/blob/master/modeling/res...). At a guess, we'll see more implementations pop up in coming months as researchers and grad students recover from NIPS and begin pondering how they could use residual learning.

You could try looking at existing animal neural counts: https://en.wikipedia.org/wiki/List_of_animals_by_number_of_n... Doublings or 20xing get you a long way: a 1/20x cerebral cortex neural count decrease takes you from human to... horse. I like horses a lot, but there's clearly a very wide gulf in capabilities, and I don't like the thought of ever encountering someone who is to us as we are to horses.

I was thinking about Instapainting, but then I realized - way too late for Christmas.