CC’s Reviews > 三体 > Status Update

Yes. I just asked in another comment, ignore it. I suppose you could use methods of Monte Carlo for that, but it is not its foremost use nor the most obvious way to numerical calculate an area.....

Hirondelle wrote: "Yes. I just asked in another comment, ignore it. I suppose you could use methods of Monte Carlo for that, but it is not its foremost use nor the most obvious way to numerical calculate an area....."

Not sure if you've gotten to this part yet, but no I wasn't saying that you can't use monte carlo to calculate area. It's an elementary use but totally valid. It's the way that the book is describing it that's totally off. It said you should cover the shape with little balls, make sure that none of them overlap (what?), and by the time you filled the entire shape with balls (what?), you count the number of balls and then you get the area.

Unless this is some really bizarre version of monte carlo that I've never heard of, that doesn't sound right... Monte Carlo is all about sampling. You're supposed to overlay that shape over a known area and uniformly scatter those balls, then use the ratio of how many fell within the shape to get its area. How does the process even work the way the book describes it...

In the translation he mentions the probability of random points being inside the area will converge (he does not use that word actually, me extrapolating) to the area, but kind of missing the point, if you can describe that well computationally the area, you can estimate the area with other methods. Monte Carlo is also really useful for stochastic processes like particle interactions and decays, and economics.

It kind of reminded me of a story a very practical, experimentalist by nature and trade, professor of mine had about his good old days where they had no computers (they didn't) about drawing curves stuff on paper with compasses, cutting it and weighting the paper, LOL, though not sure I remember it that well or if he was serious...

Hirondelle wrote: "if you can describe that well computationally the area, you can estimate the area with other methods."

It's a bit hard to imagine how you can know it well enough to be able to both reject every overlapping point AND ensure that the entire area is fully covered, lol. You might as well know the precise area already...

LOL on the paperweight. That sounds efficient though. And oh, did you make it halfway already? I supposed I slowed down too much. Will need to read more this weekend!

Paperweight sounds efficient indeed, though if it happened I guess they stole scales from the chemistry department or something.

The english edition is divided into 3 parts

Part 1 is called Silent Spring and is the cultural revolution stuff, chapters 1 to 3, page

Part 2 is called Three Body is chapters 4 to 20

Part 3 is called Sunset for Humanity, chapters 21 to 35. I am about halfway, I think, on chapter 17 about Newton. I am planning to read till the end of part 2, chapter 20 tonight, clearly big reveals are coming up, but I might not, I am finding the game chapters all hard work.

Oh and some westerner shock, the smoke, matteroffactedly and it is not supposed to make us hate the people smoking... I think in modern western fiction seeing characters smoke and others minding and smokers not caring or apologizing, it is supposed to make the reader judge the ones smoking or say something about their position, but here, I think it is just background it is not supposed to be highly symbolic they are the bad guys or something...

It's a bit hard to imagine how you can know it well enough to be able to both reject every overlapping point AND ensure that the entire area is fully covered, lol. You might as well know the precise area already...

Ah, but the trick is, you don't need to ensure that the area is fully covered.

It's not as implausible as it sounds. Sometimes you don't really have a way of describing the region you're trying to estimate -- all you have is a test you can perform on any point to determine whether the point meets whatever requirements you have.

I just re-read the description of Monte Carlo in the book. I don't love it, but I don't hate it. The only weird thing is the bit about "tiny balls". Typically it is easier and more accurate to choose mathematical points.

But the point he's making, "it shows how, mathematically, random brute force can overcome precise logic." strikes me as 100% accurate.

It said you should cover the shape with little balls, make sure that none of them overlap (what?), and by the time you filled the entire shape with balls (what?), you count the number of balls and then you get the area.

Obviously I don't know what the Chinese version says, but the English translation doesn't say any of that. Specifically, it doesn't say the balls are not allowed to overlap, and it doesn't say you keep it up until you've filled up the entire shape.

It kind of reminded me of a story a very practical, experimentalist by nature and trade, professor of mine had about his good old days where they had no computers (they didn't) about drawing curves stuff on paper with compasses, cutting it and weighting the paper, LOL, though not sure I remember it that well or if he was serious...

Yup, people used to do that. They also used to plot on paper ruled with squares, then count the squares.

L wrote: But the point he's making, "it shows how, mathematically, random brute force can overcome precise logic." strikes me as 100% accurate."

Not disagreeing with that, of course. My point was merely that the example he used should not be called monte carlo. There's so sampling involved, and the whole point of being random is undermined by the requirements he listed out.

Specifically, it doesn't say the balls are not allowed to overlap, and it doesn't say you keep it up until you've filled up the entire shape.

I'm waiting on my English version from the library to check against, but here's a google translation of the Chinese original:

The specific method is to use a large number of small balls to randomly hit the irregular shape in the software, and the hit areas will not be hit repeatedly. In this way, after reaching a certain number (of trials), all parts of the shape will be hit once. At this time, the number of small balls in the graphic area is counted, and the area of the graphic is obtained.

Basically this is saying that you're covering the shape once and exactly once with those balls, and then you count the number of balls to get its area. It's like filling a swimming pool with ping pong balls to gets its volume, which really had nothing to do with monte carlo. Randomness is not required to achieve that and is actually counterproductive, because 1) you have to make sure no area is hit repeatedly, which will happen plenty in random mode, and 2) you have to make sure all parts of the shape are eventually covered, which is quite difficult in random mode...

CC wrote: "Not disagreeing with that, of course. My point was merely that the example he used should not be called monte carlo. There's so sampling involved, and the whole point of being random is undermined by the requirements he listed out."

Here's the English translation

Specifically, the software puts the figure of interest in a figure of known area, such as a circle, and randomly strikes it with many tiny balls, never targeting the same spot twice. After a large number of balls, the proportion of balls that fall within the irregular shape compared to the total number of balls used to hit the circle will yield the area of the shape. Of course, the smaller the balls used, the more accurate the result.

It's possible that Ken Liu, who is something of a mathematician, cleaned up the description. In particular, the ambiguous phrase "never targeting the same spot twice" may be an attempt to translate the phrase "the hit areas will not be hit repeatedly" without making it an inaccurate description of Monte Carlo integration. (The difference is that "target" means "aim at", not "hit".) And Ken Liu apparently just left the phrase "all parts of the shape will be hit once" out entirely (which was the right thing to do if he was trying to produce an accurate description of Monte Carlo integration, but not, apparently, if his goal was to accurately translate Liu Cixin).

@ Hirondelle I haven't gotten to the smoking yet, just finished the chapter on Newton. I think the book actually just took a bad turn for me, unfortunately... I was liking the little cosmic science facts a lot until it's now getting too much into things I do understand, and now I don't like how it's done.

The computer thing makes no sense. Somehow building a computer out of people makes it faster to do complex calculations than by hand? Computers are fast because electrical signals work faster than real people, not because they use magic. Bit operations have certain advantages, but on a human reaction time scale, it still has to take centuries to do what they're trying to do.

I feel like this is a similar issue with the monte carlo thing. The idea that the author is trying to get across is very valid (yes to random brute force, yes to breaking down arithmetics into bit operations), but the examples are just off. Maybe that's how you've been feeling about the more physics oriented parts of the story... Oh well, I'll probably prefer it if the book gets back to those ideas that I don't actually understand instead :)

L wrote: "Here's the English translation..."

Ha! The translation fixed it! That's precisely what I was trying to say in my first comment above (You're supposed to overlay that shape over a known area and uniformly scatter those balls, then use the ratio of how many fell within the shape to get its area). That is how monte carlo works, and the original does none of this. It's literally filling the irregular area with little balls and then counting the number of balls (like measuring swimming pool volume with ping pong balls).

I'll give you the Chinese original so you can go play with the translation yourself, but yeah, kudos to Ken Liu. He only added a few words but completely changed the meaning there.

具体做法是在软件中用大量的小球随机击打那块不规则图形，被击中的地方不再重复打击，这样，达到一定的数量后，图形的所有部分就会都被击中一次，这时统计图形区域内小球的数量，就得到了图形的面积

More precisely, none of the bolded stuff below is in the original:

Specifically, the software puts the figure of interest in a figure of known area, such as a circle, and randomly strikes it with many tiny balls, never targeting the same spot twice. After a large number of balls, the proportion of balls that fall within the irregular shape compared to the total number of balls used to hit the circle will yield the area of the shape. Of course, the smaller the balls used, the more accurate the result.

Ha! Sneaky!

@Hirondelle Ah, never mind my previous comment... Apparently the first chapter in part 3 did a "gotcha" on the human computer problem I just babbled about :') And appears to have got you too on that environmental activist who did research on genetically engineered food...

Fascinating discussion, thanks!

I just want to whine about something

>randomly strikes it with many tiny balls, never targeting the same spot twice.

Well, it could be targeting the same spot twice (within definition of the size of the balls) if it were really random. Random numbers can strike the same number twice, even if it does not look natural or right to us, if we want something to look patternless and "random" it can not be really random, lol... (same teacher as paperweights, and he is right)

I also just started book 3, we will see, it is certainly speeding up.

The human computer game remind me a lot of Neal Stephenson's The Diamond Age: Or, a Young Lady's Illustrated Primer which I wonder if Liu Cixin read (almost surely yes?) or if he is a Neal Stephenson fan (yes? Or I guess all sf fans will sooner or later ending up reading NS?) But, so far, I prefer Stephenson's takes on big conspiracies and so on. Let us see how this goes, but so far it is most likely this will make me pick still unread Stephenson novels than the sequels to it...

Oh, also I forgot, about the smoking. I don't think I saw a scene where someone complained? Smoking isn't that uncommon in China (at least not as uncommon as it is in the US. maybe similar to Europe though, I've seen lots of people smoking everywhere in southern Europe...) but if someone complained and they did nothing about it, then yeah we'll still think they're assholes.

But I think the point in the smoking here is not the fact that they smoke but rather what they were smoking. The first guy was smoking cigar, which has somewhat of a "pretentious western country influence" connotation. The second guy was smoking a pipe, which is very old schooled, so I think that's what it's trying to imply about their characters (though honestly, I'd also think that similar connotations work in other cultures too? I'd think most Americans would consider cigars pretentious and pipes old fashions too ... though maybe not always?)

About the smoking, there is this scene at the beginning, chapter 4 where while Shi does not smoke in Wang's house after he requests "“Please don’t smoke in my home,” Wang said, " but afterwards"” He finally blew out the lungful of smoke that he had sucked in earlier—right in Wang’s face." and throws buts on the ground. So that as a bit of characterization which made me not like Shi much right from the start.

now I think of it, maybe the person was not complaining, maybe I just extrapolated?

The other thing is chapter 16

"Wang arrived at Da Shi’s chaotic office and saw that it was already filled with a dense cloud of cigarette smoke. A young woman police officer who shared the office fanned the smoke away from her nose with a notebook. Da Shi introduced her as Xu Bingbing, a computer specialist from the Information Security Division."

which admittedly is not a request to stop smoking explicitly nor is he smoking, but still, made me feel very sympathetic for the poor woman and wishing somebody opened a window. Incidentally is Bingbing a kind of childish name? I am going by the nicknames Wenjie uses with the children in another chapter, is it like a name ending in -ie in english Katie rather than Kate or Katherine?

Thanks for the info about pipes and cigars and cigarettes, yes, authors would use that similarly I guess in a bunch of societies, I think.

Ah, I forgot about the cigarettes earlier. That does give him kind of an image that feels rude and rough.

Hmm, the question about the name is interesting... I've never thought of it that way. Repeating a single character is indeed a way to make nicknames in Chinese, so I suppose in that way it does sound childish, but it normally doesn't strike me as such. Maybe it's because it's common for girl names in Chinese to carry meanings that are cute and innocent, so being "childish" or unsophisticated doesn't stand out? Just my guess and don't take my word for it though... And also, maybe it's just because it's common. Like a lot of people go by Katie so I stop thinking about what it sounds like (which is actually how I feel about Katie in real life...)

That makes sense thanks!

I read a bit more and you are totally right that character depth and background is given as needed when convenient - more about wenjie’ s mother and the red guards who killed her father bit all of a sudden, compartmentalized. Ye Wenjie is an interesting character though I think none of the others gets the same kind of effort or focus in being given depth!

Science atitudes and plots and even context are making a lot more sense. I really had no idea, little of all those social and ecological convulsions of the cultural revolution. I like the way the chore of scientific discovery is shown actually that process did feel true.

I think Ye Wenjie is really the main character of the book, if there is such a thing as main character here...

I finished it. Would be very interested in talking about physics with you regarding that proton, whenever you get there :)

Oh yes. I will try to finish it tonight which seems very feasible!

I finished it, I got to the protons and I could spit, froth and lecture about it all. It just just did not work out at all as a concept for me, and it is totally missing the point, many points actually. I am rating it all lower because of the ending.

and omg (view spoiler)[ the sliced ship, was some wtf-ery. The tech working, working at the same time, the civil engineering and engineering in general all done that fast, the goryness, the recovery. The economics were also horrible, say the USA government for example being all ready to sacrifice the panama canal for months and maybe years afterwards over some vague threat of alien invasion in 400 years...

the coincidence also, that precisely closest star will have unlikely civilization... the impassioned speech of the trisolarian radio operator complaining of the lack of things they never had...

(hide spoiler)]

I will try to write a review now, and it was fantastic reading this with you, and I wanted to read it for ages, this was certainly a lot "trashier" than I thought it was going to be.

There was some other novel that I read recently (meaning, in the last ten years) in which subatomic particles (neutrons, in that case) had huge information-rich internal structures. I don't remember which novel. It is associated in my mind with Charles Stross, but I don't think it was a novel by Stross himself, rather one by an author he recommended on his blog, perhaps Greg Egan or Hannu Rajaniemi.

It is, of course, complete nonsense according to current physics, but it didn't bother me terribly. I just read it to imply that there is some deep physics beyond current understanding, which is probably true, but I am pretty sure if it is it will not be THAT convenient. Compared to the sins of most SF authors, this seemed minor to me.

An acquaintance of mine objected to the ending of the movie Interstellar, in which there are structures in the interior of a black hole. Of course she was right -- such things could not exist according to current physics. But I understood it to mean that in the fictional world of the movie humans had learned to engineer black holes. I'm pretty sure that's the way it was meant to be understood since, since physicist Kip Thorne|23177378] helped in making the film, and he's long been interested in the possibility of engineered gravitational structures.

>which subatomic particles (neutrons, in that case) had huge information-rich internal structures.

arguably Tchaikovsky is implying it, at using some really dense information, in some of his books. It was handwavy, and not more handwavy than the uplift virus working across all kinds of species and it was for a good purpose.

I did mind the "seeing" literally a single proton, though, all those photons interacting with it (and so on...)This book is literally propaganda pro particle physics research (I do not mind that part, just think it is a bit naive). But it is weird where we can draw the line at "no f*cking way". I had another one with the Panama Canal for example, including it being on the panama canal...