Yeah, they don't show up with a zero, it just says not enough data to show so hard to draw any conclusions. I agree this is an imperfect proxy and it's not definitive but I'd still absolutely take you up on that bet (conditioned on there being some way of actually resolving it).
But maybe the real question is, why aren't people asking "Is Reese Witherspoon's daughter smart?"
> accusations against huge swaths of people that are totally unsupported.
The search data is meant to support the accusations. A main point is that other methods of supporting such statements (like surveys) are unreliable.
> So nobody googles if their daughter is smart?
The article doesn't say that, it says parents are about twice as likely to google if their son is smart vs daughter.
> Maybe its because parents are worried that looks and not intelligence are important for cohesing in school and not getting bullied, for girls.
This seems like exactly the problem the author wants to bring up and address. Society values appearance above intelligence in women. This is apparently also true in how parents see their children.
The author doesn't draw conclusions about this prejudice, he is merely remarking that it exists.
Going back to the magpie though, they don't have a cerebral cortex at all, yet exhibit remarkably intelligent behavior (tool usage, self-recognition). So, what's up with all the smart corvids?
They found two really old skeletons in Europe that are older than other skeletons. They are skeletons of a "modern"-ish human. They know because the teeth of the skeletons have a specific thing that the teeth of humans have but not the teeth of other primates.
The older skeleton is 1 million years older than old skeletons found in Africa. So now the oldest skeletons that we know about come from Europe not Africa, raising the possibility that humans evolved in Europe.
NIPS is more math/theory heavy as a conference while ICCV/CVPR/other vision conferences will tolerate less math and more pictures. You have probably been reading mostly vision papers but if you read more papers targeted at NIPS you will find more math. This looks to be an ICCV submission, whereas if you read the original GAN paper (a NIPS paper) there are theorems and proofs. Your observation is correct, you are just comparing different subsets of the field.
Mini-batches are not an ugly hack, batch gradient descent is too slow since you have to go through the whole data and stochastic gradient descent is too high variance (plus you can't do cool things like batch-norm). Mini-batches give you stability and speed, the best of both worlds.
If Intel is smart they will invest heavily into library development. Researchers use whatever is fast and right now cuda is fast, not just because nvidia has the best GPUs (it's about even) but because the primitives in cuda are so much faster. Matrix multiplication on the same hardware is like 3x faster in cuda than opencl or competing libraries and using the neutral networking primitives is even faster. Intel needs to invest in good, low level libraries so researchers can hack on their platform, build new things, etc. Ultimately I think researchers drive what platform gets widely used since training takes so much longer than inference.
Yes. You would have to have a large training set with these labels but it would be pretty straightforward to train. You would probably want a tagging model not a classifier because there could be multiple objects of interest in the same image. If you get me the training data I could train a model for you pretty quickly.
> Despite various publications of results where hand-washing reduced mortality to below 1%, Semmelweis's observations conflicted with the established scientific and medical opinions of the time and his ideas were rejected by the medical community. Some doctors were offended at the suggestion that they should wash their hands and Semmelweis could offer no acceptable scientific explanation for his findings.
He published a lot on the subject and was rejected because it didn't fit in with how doctors saw the world and he had no theory to back up his findings. It wasn't until Pasteur that germ theory gained any widespread acceptance so it's pretty irrelevant that other people thought of it first.
Plenty of things are evaluated in controlled studies and still rejected today. For example, the article that you presumably just read discusses one such thing.
Doctors rejected hand washing for a long time because there was no theory behind why it would improve patient care, even when there was overwhelming evidence showing it decreased mortality rates.
Do you discount all evidence that doesn't fit into your world view? Maybe theory hasn't caught up with evidence yet.
I love that people do this. Do you really thing someone is going to write a web scraper and NOT include obvious things like converting AT -> @ and DOT com -> .com?
Also, who even worries about spam any more? I haven't seen actual spam in my inbox in years.
Bittorrent Sync is an amazing product, I'd recommend switching even if you don't have political issues with dropbox. I use it to keep work files in sync between 3 different computers as well as to backup 100s of GBs of pictures, something that would cost a non-trivial amount to do with cloud providers.
That's interesting, I don't generally think to myself, "Hm, I could go drive around for an hour or I could go for an hour long walk", I usually think "Hm, I could drive to this place or bike over there (I don't usually walk)". Although I also don't really think about the risks of everyday activities anyway, so who knows.
It's really difficult to compare the risks of walking vs driving since usually they aren't really interchangeable and they have such different use cases. But the original comment was about running a marathon vs driving the same distance.
Well walking is way more dangerous per mile than driving, this was touched on by Steve Levitt in Super Freakonomics, it looks like in Britain the mortality rate is about 10 times higher per vehicle mile http://road.cc/content/news/68212-dft-casualty-statistics-ra....
Of course, that is general foot transportation. Marathons have different risk factors, you aren't going to get hit by a car (if it's a well organized marathon with a set path and stuff) but I bet there are a lot of medical issues that arise from marathons and overexertion. If I were a betting man, I'd think the marathon was probably more dangerous.
Real world currencies are valued solely on the desire to have them; GDP and credit worthiness only play a role because they affect people's perception of how stable or valuable a currency is. There is nothing tangible tying the dollar to the US GDP, if everyone decided the dollar was worthless it would be.
The main thing is that since hundreds of millions of people use the USD and you can use it to buy many goods and services, it's difficult to sway the opinions of a large chunk of the population that use it so any change in value will be slow. Bitcoin on the other hand has a much smaller user base so its volatility is much larger.