alphaBetaGamma

I tend to think that the average fiction book is less interesting that the average non-fiction book, but that the best fiction books are much, much better than the best non-fiction ones (and teach you more, and deeper things).

This one is up there with the very best. It is exquisitely well written (John Williams was a literature professor, directing a creative writing program), and despite the plainness of the plot(or maybe because of it?) it offers a remarkable view of what gives meaning to life (IMHO of course. Your millage may vary).

Stoner, by John Williams

The story of an average, not very successful literary professor, told with great humanity and tenderness. This does not seem much of a story line, yet I believe this book is so good it will still be read in one hundred years.

http://en.wikipedia.org/wiki/Stoner_%28novel%29

Einstein, Albert. "On the electrodynamics of moving bodies." Annalen der Physik 17.891 (1905): 50.

http://ganapathymani.com/On%20the%20electrodymics%20of%20mov...

This paper establishes special relativity, and is remarquable for how clear it is, revolutionizing physics while using only elementary math. The first "Kinematical" part in particular does not use anything more complex mathematically than Pythagorus theorem. It is so clear that the explanations and though experiments are reproduced in all textbooks to this day; the only change is that textbooks include diagrams.

Watson, James D., and Francis HC Crick. "Molecular structure of nucleic acids." Nature 171.4356 (1953): 737-738.

http://www.nature.com/physics/looking-back/crick/index.html

Partly because of the fundamental importance of the paper, elucidating the structure of DNA; partly for the wonderfully understated third to last paragraph: "It has not escaped our notice that the specific pairing we have postulated immediately suggests a possible copying mechanism for the genetic material."

The reason why they get good and enticing but unactionable predictions, is that they assume that they can trade instantaneously at the quoted price. They are neglecting the time their order would get to the matching engine, by which time the opportunity would be gone often enough to make their trading strategy unprofitable.

In fact the authors of the original paper seem completely clueless about this point: "For example, since the prediction time of AAPL, 0.0311ms, is less than 0.0612ms, which is the time difference between the upward spread crossing events from the Row k −1 to Row k + 4 in Table 1, the model could in principle perform fast enough to influence corresponding trading decisions"

As if their order could reach the matching engine in 30 us...

It seems to me that a consequence of the second principle of thermodynamics is that your simulations are always going to be (much) smaller than your resources (I'm using the word loosely).

So you can't go very far down the road of simulations in simulations in simulations before the "deepest" simulation is to small to have conscientious beings. So the argument that we are in all probability in a simulation seems to break down.

Then why is it better than just

n

which also doesn't give a prime for any n, but does generate them all.

Life on other planets will be extremely different from us. Consider the following thought experiment. Imagine a planet like earth, where evolution follows exactly the same path, with the same species, the only difference being that it is 2% older. The earth is 4.5 Billion years old, so 2% is 90 million years.

According to Wikipedia [1] 90 million years ago is about the time placental mammals appeared. So in that time frame evolution pulled mammals in directions as diverse as bats, whales, and elephants. So the human decedents on the other planet will not look at all like us...

[1] http://en.wikipedia.org/wiki/Timeline_of_evolutionary_histor...

I can't seem to be access the article, and so don't know if it focuses on waiting at the airport or the air speed of planes. Regarding the latter, for physics minded people, I recommend David McKay chapter on air travel[1] in his book "Without the hot air" -- actually I recommend reading the whole book, it's free.

You need to spend energy to counterbalance the drag on the airplane (which grows as speed^3) and to providing lift (which grows as 1/speed), so there is an optimal speed (if you are optimizing fuel efficiency). So you don't expect to see major increases in speed unless you are ready to burn much more fuel.

[1] http://www.withouthotair.com/cC/page_269.shtml

I think we don't understand anything about how to teach.

Formal education started at least 4000 years ago (earliest writing is at least 4,000 year old, and I think you need a formal education to have scribes who can read and write). I don't think we have done any substantial progress in how to teach in that time frame, at least compared to how much progress we have done in most other subjects: medicine, math, engineering, metallurgy, you name it. heck, we are still discussing what methodology we should use to teach reading & writing, and we have had 4000 years to figure what's the best way.

On the other hand, anyone who has had a good teacher can attests that teaching quality can vary by a huge amount. So my theory for good education is: 1) find good teachers (which probably includes "pay them well") 2) let them do their stuff

Anybody who tries to do something better than the above should be able to explain why their pet theory has not been discovered before.

You are saying that the rods' spectral curve response is collinear with the cones. That would do the trick, but I'm a bit surprised as I don't see any reason for evolution to push you in that direction.

Do you have a reference for this? Quick googling seems to show that the rods' response curve is not, it seems to be a most sensitive at a wavelengths between the peak of the blue and green cones.

I never understood why we have three trichromatic vision and not tetrachomatic vision.

We have rods and three types of cones, which look like four channels to me. In very low light the cones don't fire and we see in black and white. In reasonably bright settings I can imagine that the rods saturate and we only have three useful channels. But there must be a level of light where both cones and rods are within their dynamic range. What am I missing?

We have exact solution to spinning black holes (discovered after Einstein's death) which settles that question. http://en.wikipedia.org/wiki/Kerr_metric

To add to your counterpoints:

6. While it is about sure that in the future they will be technology that is for indistinguishable from magic, this does not imply that every magical thing we can think of will be true at some point in the future.

7. The whole premise rests on recursively improving AI. The human race has been making tool to help make them better tools for at least 50,000 years (lower bound to be uncontroversial). We do have amazing tools now, and tools that automatically do other tools, but we have not witnessed an accelerated auto-recursive thingamagig.

Of course we are in the midst of a mass extinction. This is not controversial.

http://en.wikipedia.org/wiki/Holocene_extinction http://en.wikipedia.org/wiki/Quaternary_extinction_event

Have they never heard of the "Significant figures"? I doubt they estimate to 0.3%.

In fact, I would not be surprised if the estimation error is bigger than 5%, so we don't know if they are selling at a loss or not.

I believe that as a few years ago, Bloomberg had zero terminals in china. I may be wrong, and thing may have changed, but I don't think that china is a big source of revenue for Bloomberg. Of course, I'm not saying that it is not a big potential source of revenue.

You should raise this issue to the guy doing this kickstarter project, and not only comment about it in HN. In fact, I will email him if you don't do it. I would rather you (or an other chemist) do it as you probably know much more about this than me, and can maybe suggest an alternative.

Actual, a lot of economic activity comes from small businesses: about half the US GDP is generated from companies with less than 500 employees.

http://en.wikipedia.org/wiki/Small_business#Contribution_to_...

Robert Brout would almost surely be sharing the prize if it could be awarded posthumously.

I often saw him and Francois Anglert when doing my Ph.D. Robert Brout in particular struck me as the kindest man on earth, and was perhaps someone who though even deeper and more profoundly than Francois Anglert.

If they have protein in their samples and find amino acids of both chirality, then get exited. If they only find the chirality naturally occurring in earth based life: forgetaboutit

My bet is on the latter.

As a counterpoint, I can distinguish a very bad chess player from a good one without seeing a single game he/she has played. Just looking at his/her behavior and countenance in front of the board is enough.

As an extreme example, you probably have seen a (very bad) player put his finger on the piece he wants to move, than keep it there and look around on the board, verifying that he will not immediately lose his piece when he complete his intended move. A better player will have made a final decision on his play before his arm starts to move.

A much more subtle example: better players place their pieces more accurately in the center of their square, particularly in blitz games.

Now I don't claim to distinguish a 2400 elo player from a 2500 elo. But a 1500 from a 1800 elo? Sure.

What your suggesting is going to be tricky to implement. First you can only add delay, not subtract delay, so your specific example will not work: you can not "rebroadcast at the drone's antenna" before the original signal reach the drone.

To overcome this, could record all the gps signals, and rebroadcast them with carefully timed delays. But than gps time as determined by the drone will be different from what a clock in the drone gives, so it could be detected. If you are quick enough it might work, but it's non trivial.

And if the drone uses carrier-phase gps measurements you have a whole bunch of other problems.

Not saying it can't be done, but it's not easy.

> It's basically just random chance that some planes every year will crash because of pilot error.

When you design a system where errors can cause a large number of deaths (airplanes, trains, nuclear reactors,...), your objective is to have very few accidents (nobody can guarantee zero). Having someone to blame when an accident happens -- the pilot did not follow procedure -- is not enough.

Your system design must be human error tolerant, because you know these will happen. To take an extreme example, if a drunk pilot error causes an accident it's not only his fault, but there probably also is a systemic error which caused this criminal behavior (which is a possibility you expect) to cause an accident: maybe the cockpit culture was such that the co-pilot was frightened to stand up to the drunk pilot.

Could you suggest a good board game that is playable turn by turn? We are a group of friends in very different time zone, and playing a turn based game, about one turn a day works nicely.

We have played Tigris & Euphrates, and it works great, but would be looking for something different.

Thanks.

I was speaking of a model of how python does things, based on a C style implementation. Not an explanation of how C works.

I have a C based mental model of what is happening. Can anyone comment on whether it is correct?

In C terms, I think of all objects (including primitives such as numbers) as a struct with a void* pointer that holds the actual data, and another field that describes how to interpret the data: struct Obj { void* data; ObjDesc descr; }

Whenever you do objA=objB you create an new struct Obj with objA.data == objB.data. The same things happens when you pass an object to a function: you create a copy of the Obj.

When you modify an Obj, two things can happen: if the Obj is immutable you change where void* data points to: n1 = 4 n2 = b1 n1 = 5 ==> n1.data != n2.data

On the other hand, if Obj is mutable, than you don't change void* data, you modify the memory pointed to by data: n1=[] n2=n1 n1.append(1) ==> n1.data == n2.data

Is this correct or is mental model going to bite me sometime?

Moreover, "three body problem" is generally meant to mean the three bodies with comparable mass, so they influence each others trajectory. But in this case, Mass_sun >> Mass_planet >> Mass_probe, which greatly simplifies the problem.

Nonetheless, kudos to the guy to understand that a gravitational slingshot is possible, and proving it.

I rather like this one. Though to be fair, it was probably correct to award it: it's definitely inventive, and I doubt there is prior art.

http://www.google.com/patents/US3216423

Something seems wrong: 1e5 /(deaths per 1e5) should be about equal to the life expectancy. But the drop in deaths per 1e5 is much bigger than the increase in life expectancy.

Am I missing something, or is the data SNAFU?