You sound like an average stupid brainwashed westerner and have no idea that most corporate censorship's goal is supporting western corporate colonialism and evil.
This is just the brainwashing that you've received growing up in the west. I know because I grew up in the US, but I've learned more about other countries and realized that things aren't really like this.
As predicted, the article doesn't contain any genome wide association study to back up its claim in the title that entrepreneurs "don't have a gene for risk".
Instead they present another causal factor for creating startups and just implicitly deny that other factors exist.
I think Quartz is basically the archetype of fake news at this point.
Why do hourly employees care how long your lunch break is? At least in the US it's usually not counted as work time. Assuming people clock out for the lunch break, I don't see why they'd care if it's 30m or 2hr, since people could just make up the hours by working later. Although I suppose it depends on the type of work being done.
> Greenland is a poor country. Perhaps Denmark simply wants to help out the many Inuit & Danes & descendants of both. This is ethically reprehensible. Greenland is poor, but compared to many African countries it is fabulously wealthy regardless of whether you take the $20k per capita at face value or discount subsidies etc to get a smaller number like $10k per capita
I strongly disagree with this line of reasoning. One's ability to have impact through altruism is closely linked to one's knowledge of where they're intervening. Denmark could easily have more impact in helping Greenland because they know more about what things need to be done in Greenland and can more easily get feedback on whether they've improved things. If you send money to a country that you have no connection to, how do you know that the money is being used effectively? How do you know that you're solving the right problem? How do you make improvements to your spending unless you're getting continual feedback?
There are limits to this of course and I think it can be okay to donate to international charity, but I think the impact of a well-targeted and knowledgable charity is much more certain.
It's kind of an inverse problem. Once you put the planets in position you can (reasonably easily?) simulate their oribts with some noise. The idea is that the neural network would try to quickly approximate which planet configurations could produce those stable orbits.
Is there any way to use inkscape and have it seamlessly integrate with sharelatex? I'd be kind of concerned about using sharelatex with inkscape PDFs, and then forget about or lose the folder with the inkscape source files.
"Is there any reason, specific to Japanese tradition/convention/bureaucracy to keep the name of the era secret until one month before the actual change?"
What do people mean by this exactly? I've heard some people make the case that the exact boundaries between ethnic groups can be socially constructed, which makes sense to me (like how one draws a hard line between two groups which have some admixture).
But the idea that ethnicities have nothing to do with biology and is a purely social phenomenon sounds like a lay-person's misunderstanding of the above claim. For example, it should be obvious that using a sperm/egg donor can lead to a child of a different ethnicity even if the socialization is kept the same (which you can't necessarily do perfectly in reality, but still).
At least in my experience, I don't see any pressure to not submit if you don't review (I'm in ML though, so could be different in other fields).
I think why people review is:
-You should be reading more papers anyway and we like doing it, so why not take it as an obligation?
-I assume that eventually it helps one to get selected to be an area chair or something along those lines.
I think some people buy multiple copies or order over and over again, so the actual probability is high relative to other items (because as others have said, the total number of items is huge).
I think maybe the real mistake is that someone who is buying a bunch of copies doesn't really need the item to be recommended to them. The recommendations are more helpful for things that you can't find in the history.
The breastfeeding research seems like peak "bad science", because there's so much obvious confounding between the causes of breastfeeding as a behavior and the properties of the child.
I probably would suspect minimal effect unless there were (1) a clear mechanism of action, (2) really strong associations even after adding controls, (3) a causal study, perhaps on non-humans.
Aren't you creeped out by what could happen if companies started applying this same reasoning to their customer's private communications, even discussions in their own homes - and using that as a justification for blocking access to essential services?
"If you don't know how genes affect intelligence you don't actually have any useful theory to go by. You also don't know how to judge the effect size of genetics against other causes. These are highly non-linear, complex systems we're talking about: brains and their societies. "
I made a video that actually analyzes the data behind this. Basically the entry-level of the field is about 50% Chinese and the more experienced levels are about 10-20% Chinese. On the other hand only perhaps 5% of work actually comes from Chinese labs.
To be frank - the problem is not poorly defined, you're just not aware of the definition.
In general in generative models, you have some "true data distribution" P and an estimator distribution Q.
The goal is to make P and Q the same, generally by minimizing some divergence between them.
The actual objective is defined as being between the actual distributions P and Q, but because we only have so many data points, we define an empirical loss that just uses the real observations from P. So if the model makes Q just memorize the samples from P, then it actually hasn't made P and Q similar, it's only minimized the empirical loss.
One practical way to get around this with GANs is to train a conditional GAN instead of an unconditional GAN, and then run the conditioned generation task on held-out samples from the validation set. Another good and perhaps more general solution is to train an inference network and to generate reconstructions on held out data points. If they look totally different, then the model is probably not very "representative".
1. If you train a conditional GAN to do image inpainting (for example, left to right), it should be quite apparent the degree to which the model is copying and pasting the training set - by running the model with "given" parts from the test set.
2. I disagree that an ideal GAN could just output the training set. I think the right conceptual framework is that any generative model is trying to produce a distribution similar to the data distribution, and we try to accomplish this by using samples from the data distribution. So if the model memorizes the training set, then it isn't actually that close to the true underlying data distribution. In likelihood-based models (for example the usual generative RNN) you can test this by evaluating likelihood on a validation set.