andreasvc

That's quite the low effort response. You could've at least mentioned a point or two that applies to this situation. The list you linked to is not a check list, it only lists ways to shoot ideas down. Although I don't necessarily disagree with the skepticism in this case, consider for a second that a very similar list could have been written up until right before the Wright brothers took off with their plane, supposedly proving manned flight impossible.

"Can it really be true that violent crime was declining between 1973-91? There's no question that this is what the NCVS data tells us, but it sure doesn't match what everyone thought was going on at the time. By the early 90s, virtually everyone in America was convinced that (a) crime had been rising relentlessly for decades and (b) it would continue rising. (Remember "superpredators"?) At a purely gut level, a crime decline during this period sure doesn't feel right."

I think this can and may be true. See Pinker's arguments in The Better Angels of our Nature. It seems obvious to me that people's perceptions of crime are influenced to a huge extent by media reporting. People have irrational, unfounded beliefs all the time, so I don't see why they should be indicative of anything.

Lots of people consider both to be bad. NK also murders lots of innocents, I honestly fail to see your point.

> I'm not interested in natural language processing with machines and machine learning.

I'm confused, why not? "English grammar visualization" is inherently an NLP problem. Both handwritten and statistically learned grammars are very large and therefore hard to visualize. I think it would be better to focus on a manageable language fragment. Some kind of interactive second language learning environment--but that probably already exists. What is it exactly that you want to do?

BTW: I wonder for how many people "graphing sentences" is an effective approach to language learning.

Don't know if this solves your particular problem, but I liked this LDA visualizer that uses D3: https://github.com/agoldst/dfr-browser

No not necessarily. If the headline claims something is misunderstood, you set up the expectation that you'll set the record straight. Misunderstood implies that there is some correct point of view, otherwise you'd call it a mystery.

I didn't expect hard data, but a list of traits specific to female autism would have been useful.

> She is currently at 27% for her age.

That sounds precise... Is that based on the size of her vocabulary?

For an article addressing the issue of autism in women being misunderstood, I would have hoped to find a clear description of the actual difference with male autism. Instead we get a few human interest case studies and scientists working on this. The article simply doesn't seem to go to the heart of the matter.

No I haven't. I just found it very odd that that was the conclusion drawn from this (even if it's only called a hint), while not saying anything about how bad those psychiatrists were.

> It’s the long list of diagnoses Maya collected before she was 21, from borderline personality disorder to agoraphobia to obsessive-compulsive disorder, that begin to hint at how little we understand autism in women.

That doesn't follow. It could simply mean that she had really bad psychiatrists who just threw around diagnoses. Maybe it reflects that mental illness in general is poorly understood, because they didn't discover that the other diagnoses were false positives. I am afraid that unless a diagnosis is followed by a successful treatment, it's not worth much.

> deem some one being autistic without any actual developmental tests, brain imaging or even in depth psychoanalysis.

The latter two have nothing to do with psychiatric or psychological diagnoses. A diagnoses is made on the basis of symptoms, it doesn't get more exact than that. There are symptoms which frequently go together, and this is how disorders get a name. Although science attempts to find biological markers for mental illness, this has not been very successful. Psychoanalysis is just pseudoscience.

> Physics is an empirical science

Theoretical physics is also very much a thing.

> and should therefore be testable by definition.

That sounds nice but the definition of testable is hard to pin down and changes over time.

What I'm saying is that not only does Python not have superb performance, its performance is actually largely irrelevant. The advantage of Python is in ease of use, capabilities as glue language, readability, etc. So that's why I really don't see why you wouldn't remove the GIL at a reasonable slowdown factor (2x was a figure thrown around IIRC).

People always proclaim that you can just use multiple processes for parallelism. It's nice when it suits your (embarrassingly parallel) problem but when you need to share a large amount of data between processes it's a major hassle.

Yes it does mean that and I believe it is unavoidable, because you need to replace the coarse-grained GIL with a fine-grained locking instructions. However, I think the trade off is now worth it, since multi-core CPUs are ubiquitous. Besides, Python itself was never notable for its performance to begin with, so I've always found the argument very odd.

No need to get snarky, because the % formatting is still in Python 3. It is also my preferred formatting method.

That's not the proper way. You should rarely need encode/decode if you make sure everything within Python is unicode. You ensure this by being specifying the encoding when getting data from external sources such as files, the network, or external libraries.

Oh, my mistake. I didn't realize that was what the 75% figure was based on.

No you cannot. If being a Facebook user is highly correlated with being depressed, the figure could be considerably lower. See Simpson's paradox. If you take a subset of a dataset, you can find the opposite trend or a completely different mean compared to the whole population. This is because the subset can have particular features causing it to be not representative of the whole population.

It doesn't make sense to sort a configuration file, order and context matters. Easier to just use grep '^#' (or even better, ag).

I think this is a really strange thing to say. Why do you care? Why would some famous hacker need to live up to your expectations? What you're saying comes down to a teenager complaining "why are you still on Facebook and not on <social network du jour>!?".

zgrep simply decompresses all the data and feeds it into regular grep. If the data is indexed in some way, it is possible to do better by not having to look at all the data exhaustively.

It's possible to compress and index a file at the same time, gaining both a size and speed advantage over the original. For example: https://en.wikipedia.org/wiki/FM-index

> It is one model of the human brain

No it isn't. It's loosely inspired, but the goal is to make something that works, not to simulate what we know of the human brain. There's a different field, computational neuroscience, which does try to make faithful models of the human brain.

Absolutely not. Somebody who claims to have "a couple of doctorates" is definitely a phony. Obtaining a doctorate takes 3-4 years, and it is highly unusual to get more than one. Moreover, you would not forget about which ones you obtained.

That's just one example, construction grammar (not limited to US west coast). The innateness of "universal grammar" can be argued against on evolutionary grounds (how could one universal grammar evolve and end up in the DNA of all humans?), neuroscientifically (how can you encode a fixed grammar when everyone's brain is different and continually changing?), from comparative linguistics (there are many languages and they can be radically different), etc.

If it's not about innateness but about the claims of "internal grammar" from the press release then it because less clear that there would be many people who would reject such a position. I think this is a misleading on the part of the press release. Whether something was innate or learned, I think most people would agree that knowledge of language is internalized. So then the question remains whether you call it grammar or something else, and whether statistics plays a part should have no bearing on that if you ask me.

> The notion of innate grammers has been pretty much replaced with evidence that you get much better and more human like performance (and the ability to actually learn the bloody things) out of RNNs and the like.

No not really. Barring concrete physiological evidence, innateness is simply an unfalsifiable position. I do think you can say that consensus among scientists might have shifted against innateness, but that is definitely not because of some computational model such as RNN with some passing resemblance to observed behavior of human neurons. The way these networks are configured and trained has nothing to do with real brains, nor does the abstraction in neural nets leave room for any electrochemical effects.

> The interesting question is why RNNs are able to produce things that resemble formal languages at first pass.

Is it? A neural network can achieve that by simulating a Turing machine or simpler automaton. I don't see what that brings to the table? That's just doing something we could already do, but less efficiently. I think the interesting thing is the opposite: if the RNN can do something which couldn't be done before with other models.

For one both our auditory and visual senses have limited bandwidth. Reading is done in saccades, so already at the source the information comes in sentence chunks at a time.

The opportunity for massive-parallelness is in disambiguating a huge number of possible interpretations. Before you can construct the possible candidates, you first need to have processed enough of the sentence. But since you can process new sensory input at the same time as disambiguating previous input, you can get the garden path effect. What I find interesting about the garden path effect though is that it's so artificial, so the way language is used normally seems to purposely avoid ambiguities that will trip up listeners/readers.

No point in asking for a citation on a press release statement, go directly to the paper (which seems to be much more careful in what it claims).

> How are they distinguishing between grammar that's encoded in the brain due to "nature" vs grammar that's encoded in the brain via "nurture"?

They do not.

> I think we all knew that the brain has some mechanism to detect grammar.

The press release is spinning the paper to make it sound much more controversial than it is.

> "massively parallel" is probably drastically overstating it, though.

Think of a single-core processor executing instructions one at a time. Now think of ten billion neurons in your brain out of which a large proportion is active at any time. That is definitely massively parallel.