Wikipedia:Reference desk/Archives/Science/2008 December 20

Science desk
< December 19 << Nov | December | Jan >> December 21 >
Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.


December 20

edit

Eye Layers

edit

The eye has 3 liquid to semiliquid layers over the sclera and cornea: tear, mucus, and what is the 3rd one?96.53.149.117 (talk) 01:48, 20 December 2008 (UTC)[reply]

You linked to all those articles and yet didn't bother to read them? Nil Einne (talk) 05:05, 20 December 2008 (UTC)[reply]
tldr. Why do 99% of people ask questions here?96.53.149.117 (talk) 05:15, 20 December 2008 (UTC)[reply]
Too long didn't read? I found the answer to your question within 1 minute. No one said you had to read them all from top to bottom, a bit of common sense and use of the find function that nearly any browser has goes a long way. As to what other people do, quite commonly a lot of questions are not answered so simply and clearly in one of our articles. While there's no doubt some people do ask a question without reading or searching, those questions are often ignored or given less then straight answers, as in this case. As the header says, many people can't be bothered helping those who show no sign of helping themselves and you will often get an answer faster if you make the minimum of effort to look it up. (As it stands, it's taken 3+ hours to get an answer you could have had in a minute). If you want, I could explain how I found the answer within a minute although I would recommend you seek further help since basic search skills are vital in this modern day and age Nil Einne (talk) 05:31, 20 December 2008 (UTC)[reply]
yes, that is what tldr means. What amazes me is that you even take the time to make an inflammatory comment at the start. And considering your net savy, you had to ask for clarification of the initialism?96.53.149.117 (talk) —Preceding undated comment was added at 05:58, 20 December 2008 (UTC).[reply]
This is all too OT for me. Replied to your talk page Nil Einne (talk) 08:05, 20 December 2008 (UTC)[reply]
For the benefit of future readers, the answer is in tears Nil Einne (talk) 08:38, 20 December 2008 (UTC)[reply]

Milk

edit

Why drink milk after ingesting SOME (certian) toxins?96.53.149.117 (talk) 02:07, 20 December 2008 (UTC)[reply]

Such as Draino; what I mean is that on some of the safety labels, they will say, "If ingested give a glass of milk and call a physician.".96.53.149.117 (talk) 03:56, 20 December 2008 (UTC)[reply]

Milk has a mild buffer effect, meaning that it will not only dilute the acid/and/or/base but it will also tend to neutralize it. Pure water will only dilute it, and for particularly strong acids or bases, dilution is not usually enough. The same volume of milk compared to water will have a slightly greater nutralizing effect. Mind you that a) its not likely to be much of a difference and b) if you just drank drano, you just turned your insides into soap. A glass of milk is unlikely to make a huge difference at this point... --Jayron32.talk.contribs 04:03, 20 December 2008 (UTC)[reply]
Also, milk contains fat (except skim milk), so helps to dilute fat-soluble poisons, while drinks without fat will not. StuRat (talk) 04:31, 20 December 2008 (UTC)[reply]
Actually, that may be a bad idea, since the fat in milk may actually help increase absorbtion of fat-soluble compounds into the bloodstream... --Jayron32.talk.contribs 06:18, 20 December 2008 (UTC)[reply]
(wild speculation warning) That was my train of thought, too. But maybe the idea is to absorb fat-soluble compounds and hold them in the milk suspension until the stomach could be pumped. But sodium hydroxide is water soluble (although it would be soluble in milk also). Some other possibilites - Milk might coat the stomach better. Also, with Drano (assuming the amount was fairly small), maybe the milk would provide a substance other than you to undergo saponification (just don't use skim milk). But I've seen "drink milk" on other types of chemicals as well. (Side question: is water is absorbed more quickly than milk? That could be a reason) --Bennybp (talk) 07:10, 20 December 2008 (UTC)[reply]
Water must definatly be absorbed faster than milk. Milk is a suspended solution.96.53.149.117 (talk) 01:47, 21 December 2008 (UTC)[reply]
Also, that IS my other question. What about other chemicals? I really didn't mean, and didn't want, this question just for draino.96.53.149.117 (talk) 01:47, 21 December 2008 (UTC)[reply]
Water passes through the stomach about five times faster than milk. Someguy1221 (talk) 10:40, 20 December 2008 (UTC)[reply]
I think the greater thing is the same reasoning behind having kids hide under their desks during a nuclear attack. If they just printed "If you drink this, there's not much you can do for yourself. Go ahead and call an ambulance, but you're likely a goner" on the label, it wouldn't look too good. The idea is that people feel psychologically better doing something, even if that something is entirely ineffectual at solving the problem, like taking off your shoes at the airport or hiding under your desk to somehow save you from a 10 megaton blast from a nuclear bomb... --Jayron32.talk.contribs 03:04, 21 December 2008 (UTC)[reply]
Taking off shoes at an airport? What does that do? But I disagree with you about putting something on a label for the sake of reassurance. They didn't choose milk for any reason.96.53.149.117 (talk) 04:51, 21 December 2008 (UTC)[reply]
Taking off shoes at an airport? What does that do? Make Dr. Scholl happy because of many added athlete's foot cases. The label designers would also have to consider the added complication of finding something that is considered readily available, even if less effective than some not so common aid. Everyone knows pretty fast where to find their milk. Baking soda or vinegar or whatever could lead to a time consuming search. So a little now is better that a lot, too late. 76.97.245.5 (talk) 12:29, 21 December 2008 (UTC)[reply]
But why at an airport?96.53.149.117 (talk) 03:25, 26 December 2008 (UTC)[reply]
Drinking milk is Security theater? -- JSBillings 15:12, 21 December 2008 (UTC)[reply]
When I was younger a chlorine tank (which my neighbor who ran a pool cleaning business was storing illegally) burst in my neighborhood. I started coughing uncontrollably, the poison control center told me to drink some milk and the coughing subsided. I never really gave any thought to why that worked before, but now I'm confused. -- Mad031683 (talk) 23:37, 22 December 2008 (UTC)[reply]

Why do young women giggle all the time?

edit

And I am not just talking about when I am around. Though they seem to do it even more when I am. —Preceding unsigned comment added by Trevor Loughlin (talkcontribs) 06:15, 20 December 2008 (UTC)[reply]

If you are talking about teenagers, I would suggest that the onset of puberty leaves them with with more highly charged emotions in general at times, and that therefore they experience joy, giddiness, whatever you want to call it, a lot more. It also *seems* that they also face more "crises" than boys, from what I've seen, so they cry more and such, too, at that age. However, it may only seem that way because females, at least in Western culture, are often more open about showing their emotions.Somebody or his brother (talk) 14:25, 20 December 2008 (UTC)[reply]
Indeed, for good or bad western society tends to demand stereotypically masculine behavior from young men. This generally does not include giggling, therefore boys tend to restrain their emotions more. Young woman may also be influenced by equally stereotypical depictions of females in the media, which often tend to fall along the lines of valley girls. --S.dedalus (talk) 04:51, 21 December 2008 (UTC)[reply]
I strongly disagree with that explanation. So far as I know, almost every culture's girls express emotions more strongly than do its boys. You can't explain that away using only stereotypes, which have no reason to remain constant across societies that developed nearly independently of each other. --99.237.96.81 (talk) 07:37, 21 December 2008 (UTC)[reply]
Perhaps it's related to the fact that the start of puberty often starts earlier in female humans than in male humans? Also, the changes in women occur much faster than in men, and the hormonal changes often lead to an imbalance of neurotransmitters (according to the puberty article) which would make one more emotional. -- JSBillings 15:10, 21 December 2008 (UTC)[reply]
While stereotypes may be the wrong word, I don't think you can really rule out cultural factors. For starters, I don't think its true that cultures developed largely independently. A lot of cultures have been quite interconnected for a while. Some like the indigenous Australians for example were isolated for a long while but I doubt we know very well whether they have traditionally shown the trends of females displaying emotions more strongly then males. They didn't have written records and I don't know whether their oral records will really tell us such minutae accurately. There may be records of how they behaved from earlier European encounters but those would have been through the eyes of thoese people and I'm doubtful comparing female and male displays of emotion would have been something they observed in depth unless for whatever reason it seemed highly unusual to them which I doubt it would have been even if their displays were similar. Written records do exist for less isolated cultures, e.g. East Asian ones but given they were less isolated it's hard to rule out cross-cultural influences. Besides has anyone ever looked into the records in detail to find out if displays of emotion by females was stronger then by males? Most significantly perhaps the cultural factors like men as hunters and providers and females as care givers developed a long time ago and are common in the vast majority of cultures. There are obvious genetic and biological reasons why these developed but the important point is that it's easy to see differences in displays of emotions arising out of this, without any real biological or genetic reason (presuming as I've stated these differences are really unanimous throughout tradiotional cultures which I'm not sure we we know). Nil Einne (talk) 11:54, 22 December 2008 (UTC)[reply]

Artificially recreating the earth's magnetic field

edit

At some point in the distant future, the earth will lose its magnetic field because its core cools to a point that the current of molten iron will stop flowing. Without its magnetic field, the earth's atmosphere will be stripped away by solar wind. It is impossible to tell if humans will still be around at that time, and if they are, what kinds of technology will be available. However, based on what we know about physics today, are there insurmountable obstacles to replacing the earth's natural magnetic field artificially? —Preceding unsigned comment added by 98.114.98.236 (talk) 08:32, 20 December 2008 (UTC)[reply]

Atmosphere will not be stripped away. The field has reversed many times, and during each reversal there was a time when the field was zero, but nothing like that happened. There will be many more cosmic rays reaching the Earth, increasing the frequency of spontaneous mutations in DNA, but not as much as to endanger the survival of the human species. -- Army1987 – Deeds, not words. 14:29, 20 December 2008 (UTC)[reply]
I read about the atmospheric loss as a consequence of magnetic field loss in a science article. The theory sounds plausible to me, but I'm not a geophysics expert. Maybe the loss occurs slowly enough that the temporary loss of magnetic field during a geomagnetic reversal only cause a little atmospheric loss. However, if the earth's core cools down, the loss will be permanent; even slow loss will accumulate. In another article that I've read (one about an upcoming global helium shortage), it is stated that when helium floats to the outer edge of the atmosphere, it escapes into space. That tends to lend credence to what solar wind can do to the atmosphere when there's no magnetic field to mitigate the effects. --98.114.98.236 (talk) 14:58, 20 December 2008 (UTC)[reply]
What is the basis of your claim that the atmosphere will not be stripped away? The questioner is not asking about a very short period of null magnetism. This question is about a permanent state of no magnetic field - for centuries and centuries - until the Sun expands and swallows the Earth. Looking around, planets without magnetic fields have all the upper atmosphere stripped away. Are you claiming Earth is special and the Sun will be extra gentle to make sure none of the upper atmosphere is harmed? -- kainaw 15:03, 20 December 2008 (UTC)[reply]
The upper atmosphere? Maybe. But all the atmosphere sounds implausible to me. Also, the other planets you refer to have weaker gravities (the Earth has the greatest acceleration due to gravity of all terrestrial planets in the solar system). -- Army1987 – Deeds, not words. 15:17, 20 December 2008 (UTC)[reply]
According to Solar wind, it only stripped away one third of Mars's atmosphere. In four billion years. And Mars is farther from the Sun than the Earth is. And its atmosphere was much less dense than that of Earth. -- Army1987 – Deeds, not words. 15:24, 20 December 2008 (UTC)[reply]
Exactly - the upper atmosphere is easy for solar winds to strip. That is what the questioner was asking about. Refusing to answer the question because some heavy gasses might survive is being a bit of an ass. The question is: Is it possible - with any technology that could ever be developed - to create an artificial magnetic field around the Earth. I personally do not believe it is. We create fields like that with electricity. The amount of current we'd need to pump through the Earth to get that big of a field would do more harm than good. If a new way to create magnetic fields was invented, it could be possible. -- kainaw 18:40, 20 December 2008 (UTC)[reply]

← You're right that I didn't answer the OP's question (I apologize), but I was pointing out that the scenario he described is implausible. (And it is the "some heavy gasses" that we most care about, as we breathe them, not the hydrogen and helium which tend to go up.) Also, I don't think that the Earth's core is going to freeze over within a reasonable amount of time (where by "reasonable" I mean "before mankind will be very likely to be extinct for some other reason"). I suspect he's watched The Core recently and took it too seriously. -- Army1987 – Deeds, not words. 20:38, 20 December 2008 (UTC)[reply]

Indeed. When we see these events in isolation, like "Someday, the moon will escape the earth's gravity and drift off into space" or "The earth is slowing down and some day it will stop turning on its axis" or "Someday the core will cool to the point when it is solid" ignores the fact that none of these events happens in isolation. Remember the most important event, that the sun will essentially swallow the earth in 5 billion years or so (see Stellar evolution); all of these scenarios require timespans SIGNIFICANTLY longer than that. So, no, the earth's magnetic field is not going away at any time, unless being vaporized by the expanding sun counts as destroying the magnetic field... --Jayron32.talk.contribs 02:55, 21 December 2008 (UTC)[reply]
By the way, the Moon won't drift off into space, it will gain angular momentum until Earth's rotational period equals the period of the Moon's revolution, and at this point it will be still within the Hill sphere (less than 600000 km from Earth). Icek (talk) 08:28, 21 December 2008 (UTC)[reply]

What about Venus? It has no internally-generated magnetic field but a dense atmosphere and lower surface gravity than Earth. Icek (talk) 08:28, 21 December 2008 (UTC)[reply]

The atmosphere of Venus is primarily CO2. There is a touch of nitrogen and not much of anything else. On Earth, the atmosphere is a good mix of nitrogen and oxygen. They are not comparable. Gravity alone can hold heavy CO2 to the planet. The question is: Did Venus used to have lighter elements in the atmosphere? If so, where did they go? If they went into space, was it lack of gravity, solar ray stripping, or a combination of both? -- kainaw 20:54, 21 December 2008 (UTC)[reply]

Flash point of fission in atmosphere

edit

What is the flash point of fission in the atmosphere or the hottest temperature in the atmosphere before there are adverse reactions? Tiailds (talk) 08:35, 20 December 2008 (UTC)[reply]

For the first, impossibly high (if such a concept even works). Spontaneous uranium fission almost never happens, and that only in ideal underground conditions. Sustained fissioning of atmospheric elements, which we don't produce even in labs, can be safely ignored as a threat to humanity. Certainly your latter question doesn't require such extreme conditions. Air temperatures over 100°C would be catastrophic. Temperatures just 10° greater than those at present (perhaps even less) would be sufficient to melt the bulk of the ice caps. But the whole planet turning into a giant nuclear bomb? That doesn't even meet the threshold for bad science fiction. — Lomn 14:41, 20 December 2008 (UTC)[reply]
(Spontaneous fission in general is not that uncommon. Self-sustaining chain reactions almost never happen. There's a big difference there. An atom or two fissioning—who cares. It's only a chain reaction where that becomes an issue.) --98.217.8.46 (talk) 15:56, 20 December 2008 (UTC)[reply]
(And before someone else mentions it, there were, of course, some concerns that nuclear reactions set off by the first atomic bomb would accidentally induce a fusion chain-reaction in the atmosphere, turning the planet into a giant hydrogen bomb. But the problem was studied carefully and it was determined that it was not at all likely to occur. It turns out self-propagating fusion reactions are pretty hard to induce even in ideal situations.) --98.217.8.46 (talk) 16:01, 20 December 2008 (UTC)[reply]
I think the last thing is what the original poster was asking about, rather than a "before someone else mentions it". Anyway, you've provided a link. --Anonymous, 18:50 UTC, December 20/08.

Why do some children have the concept of growing down after one grows up?

edit

I have heard a number of times that little children have the idea that one grows down after one grows up. Why do some get that idea? I don't think all do, my younger cousins never did, from what I noticed. It usually is with children around 2-3, I think; maybe 4, too.

I'm thinking it has to do with the idea that their minds pick up on patterns like the bedtime routine, days of the week, etc., but it could just be the attempt to make sense of the world. (i.e.: "Mommy has to be a baby sometime, I never saw her as a baby, therefore she must be a baby sometime in the future.") It is definitely from the preoperational stage, as I' think 4 is the oldest I ever heard of tha child thinking this.Somebody or his brother (talk) 14:06, 20 December 2008 (UTC)[reply]

"What grows up must grow down" ? StuRat (talk) 17:23, 20 December 2008 (UTC)[reply]
"... Last scene of all,
That ends this strange eventful history,
Is second childishness and mere oblivion;
Sans teeth, sans eyes, sans taste, sans everything."
Well, humans do shrink slightly if they make it beyond middle age - the skeleton starts to degenerate, the posture becomes more stooped etc. Exxolon (talk) 02:29, 21 December 2008 (UTC)[reply]
Additionally, due mostly to better nutrition, each of the past few generations has been slightly taller, on average, than its predecessor. Your parents are probably taller than your grandparents. A young child may assume that all adults top out at the same height, so the older generation must be now be shrinking. B00P (talk) 08:10, 21 December 2008 (UTC)[reply]
I've read, incidentally, that that trend mysteriously stopped in the US at about the end of the Baby Boom (i.e., with me). —Tamfang (talk) 08:23, 28 December 2008 (UTC)[reply]
It resumed about thirty years later; I started noticing very tall young people in 2010. —Tamfang (talk) 04:05, 3 September 2023 (UTC)[reply]

Why does chronic stress sometimes lead to stroke

edit

I can understand why stress can lead to a heart attack - that's pretty logical, with the strain on the heart. I'm a little confused, though, on why stroke can be a result. Is stroke only a result because of the impact on the heart itself, which thus leds to blockages? Or, does persistent, extreme stress actually do something to the brain itself, which causes clots to form there, just as it can lead to a headache, etc.?

Or, perhaps it can be either, depending ont he type of stroke, which is quite plausible.209.244.30.221 (talk) 16:04, 20 December 2008 (UTC)[reply]

Strokes that are caused by chronic stress are due to a disturbance of a blood vessel in the brain. Chronic stress increases your blood pressure, which increases the risk of rupture of a blood vessel. If this blood vessel is in your brain, and it ruptures, it can cause a stroke. —Cyclonenim (talk · contribs · email) 16:21, 20 December 2008 (UTC)[reply]
Stress may also drive a general inflammatory state, leading to hypercoagulability (this study is an example: [1]), endothelial dysfunction (e.g. in the carotid artery or other location), and then embolization resulting in ischemic stroke that can sometimes become a hemorrhagic stroke. Stress has many negative effects, and exercise is a positive balancer for almost all of them (whereas watching TV is not). --Scray (talk) 19:12, 20 December 2008 (UTC)[reply]
Rare, but in people with Sticky platelet syndrome [2] stress can also set off a stroke. http://www.cli-online.com/products/haemostasis-coagulation/haemostasiscoagulation/clinical-laboratory/testing-the-sticky-platelet-syndrome/index.html?tx_ttproducts_pi1[backPID]=1031&cHash=24a96e1d04 (Don't know the latin name, so we may have a page)76.97.245.5 (talk) 12:07, 21 December 2008 (UTC) Sorry the long link won't work otherwise because it contains angled brackets. If s.o. knows how to fix it, plse. do.[reply]
Chronic stress can lead to hypertension. From "Stroke", "Hypertension accounts for 35-50% of stroke risk.... Lowering blood pressure has been conclusively shown to prevent both ischemic and hemorrhagic strokes." Axl ¤ [Talk] 23:32, 25 December 2008 (UTC)[reply]

is it possible to produce a novel algorithm USING THE POWER OF THE MIND ALONE? (no other tools of any kind)

edit

Is it possible to produce a novel algorithm USING THE POWER OF THE MIND ALONE? (no other tools of any kind) —Preceding unsigned comment added by 79.122.90.7 (talk) 18:31, 20 December 2008 (UTC)[reply]

Since I know of no physical effects associated with a mind as opposed to workings of the brain I would say no. If by mind you mean just thinking about things, power of the mind as th ability to think and find patterns, and algorithm as a way of doing something then I would say yes, where else would the novel bit come from, even noticing something novel is something that has to be thought. Dmcq (talk) 18:52, 20 December 2008 (UTC)[reply]
It's hard to imagine any algorithms that were not produced using the power of the mind alone! So, for example, Quicksort is a well-known sorting algorithm developed by C. A. R. Hoare - I imagine he used his mind alone to develop it. My son and I are writing a computer game (to give to my wife for Xmas!) - it has a couple of new algorithms in it (well, they are new to us) - and I just sat down and thought them up. I didn't need a pencil and paper or a computer to do that.
This seems (to me at least) like such a ridiculously simple answer - that I have to imagine that the OP is actually asking something more complicated. Perhaps a clarification is needed? SteveBaker (talk) 20:27, 20 December 2008 (UTC)[reply]
One way to look at it... You can imagine an algorithm and suspect that it will work as well as you think it will. Until you physically test the algorithm, you will not know if you made a mistake in your thinking. So, the mind will create the algorithm, but you need to test it afterwards. I know, there are thousands of mathematics students who will swear that you can prove an algorithm without testing - but that assumes you didn't make a mistake in your proof. You still need to test it to make sure there were no mistakes in the creation of the algorithm. Want a silly little anecdote? Many years ago (when I got my brand new TI-99 4/A), I thought of a cool algorithm for using an evolutionary approach to solving simple problems. I thought it through, then wrote it down, and finally programmed it in. My test problem was very simple. The bot (I called it) got two single digit numbers, 1 to 9. It was supposed to produce the sum of the numbers. The further away it was from the sum, the worse it was punished. After a few hundred generations, the bot kept answering 10. Why 10? My algorithm was bad. The bot didn't learn to add two numbers. It learned that if it answers 10, it will limit the punishment it receives over the long term. -- kainaw 00:58, 21 December 2008 (UTC)[reply]
Well - that's not a really fair example - you didn't come up with an algorithm to solve your problem - you essentially said - I don't want to be bothered with figuring out the algorithm, so I'll come up with an algorithm that'll allow the computer to learn - and I'll hope that IT can figure it out. Your algorithm (the learning thing) worked just fine - it learned...but because it's fairly stupid, it learned the wrong thing.
Obviously when we create algorithms using our minds alone, we sometimes make mistakes. Testing to see if a mistake has been made can certainly be done by implementing the algorithm in a computer and testing it - but our OP asks whether we can create algorithms without a computer - and we CERTAINLY can. There are plenty of simple algorithms we can test in our heads. It's the essence of a good computer programmer to be pretty certain that somethings' going to work BEFORE sitting down to write the code. You can't always be 100% certain - but sometimes you are - and the skill of computer programming is in spotting problems as early as possible. At any rate - we CAN and DO produce algorithms 100% in our heads. SteveBaker (talk) 01:30, 21 December 2008 (UTC)[reply]
I'm sorry, but you appear to have misread my comment. I came up with an algorithm to have a computer figure out an algorithm. My algorithm was how it would learn. I didn't have the computer figure that out. However, I explained that my algorithm (which seemed fine to me when I thought it through) didn't work. -- kainaw 04:16, 21 December 2008 (UTC)[reply]

I think the following question is more interesting than the original, "Is there any algorithm that has not originated in a human mind ?"
I realize that the phrasing is fuzzy and the answer will depend upon how we define the terms "algorithm", "originated", "mind" etc - so in order to avoid a semantic or philosophical discussion, let me specify that I am interested in a computer generated algorithm that is analogous to (say) the computer generated proof of the four color theorem. Any examples ? Abecedare (talk) 03:17, 21 December 2008 (UTC)[reply]

You have to define "algorithm". It is normally considered something like "a sequence of steps to solve a problem." There are many algorithms that were not created in the human mind. For example, a hedgehog knows that if it feels threatened, the sequence of steps is: 1) Try to run away. 2) If that fails, roll into a ball. It isn't a great algorithm, but it is a series of steps to solve the problem. DNA replication is another (more complicated) algorithm that was not created in the human mind. But, if you want to claim that all biological, social, and natural processes are omitted from the things you want to call "algorithms", then you have decided to limit algorithms to humans and humans alone (assuming humans are not biological, social, or natural). -- kainaw 04:21, 21 December 2008 (UTC)[reply]
I think the example Kainaw gave of a learning machine is the kind of thing where computers do (in a sense) come up with an algorithm. There is an old story (and it may be apocryphal) of a Japanese subway operator who had problems with the platforms of his stations filling up with so many people that it became dangerous. Doors were added at the entrance to the platform that could be closed when the platform became too crowded. They decided to use a camera and a neural network software package to look at the images of the platform and close the doors when it started to become too crowded. They showed the software lots of pictures of crowded platforms and uncrowded ones - and trained it (as one does with neural nets) to recognise one set of images from the other. When they put the resulting (trained) network into service, it worked just great - reliably shutting the gates when the platform started to fill up - and opening them again as it emptied. The system was in service for almost a year. However, on Xmas day, it failed. Shutting the gates when the platform was completely empty. The story goes that the engineers work hard to find out why the neural network failed and they discover that only the neurons that are looking at the very top of the image from the camera are actually doing anything. The others have shut themselves off. It turns out that the camera was looking at the clock on the wall of the station and shutting the doors at the exact times of day that the station tended to get overly full.
I don't know whether that story is true or not - but (as Kainaw and numerous others have discovered) computer learning software doesn't always learn the things you want or expect it to. The story persists precisely because it is an excellent description of exactly the kinds of things that happen with learning software.
If the Japanese subway story is true - then the neural network "discovered" an algorithm for shutting the gates that worked pretty well. It had not been taught about clocks or anything like that - it had (in essence) 'deduced' an entirely different way of achieving the stated goals. Just as C. A. R. Hoare discovered a better way to sort things when he figured out 'QuickSort'.
SteveBaker (talk) 06:52, 21 December 2008 (UTC)[reply]
Are Japanese subways busy on Christmas day? There was a similar story about neural networks being trained to identify tanks in recon photos. Long story short the neural network was just checking for cloudy days. Again, I have no idea of the origins of this story, but I have a nagging suspicion I heard it on "The Machine That Changed the World" or a similar documentary. APL (talk) 17:26, 21 December 2008 (UTC)[reply]
I don't know about how busy Japanese subways ACTUALLY are on Xmas day - but the implication from the story is that the station was deserted because of the holiday - and the computer went ahead and shut the gates, not because the platforms were full but because the clock was showing the time of day at which the platform is USUALLY full. I'd heard the ones about tank recognition too. Somewhere in the boxes of books in my garage I have a book on Neural Networks that tells both stories (and several others also). The field of computer learning (be it neural networks or evolutionary algorithms or whatever) is simply littered with these kinds of stories. Some are certainly true.
One that I like (and I know is true) is the guy who was trying to evolve 'virtual creatures' using computer graphics and such. He was interested in having them evolve to be able to move around in their virtual world. He had simulated muscles and bones and such - with neural network intelligence to drive them. He set the simulation up and ran races between creatures with different 'genes' to see which would be the first to cross a line some distance away. The evolutionary part favored the genes that generated creatures who crossed the line the fastest - or in the even that none of them got there within a reasonable amount of time - rewarded the creature that got closest. The simulation ran overnight and the researcher was bitterly disappointed to discover that all of the creatures that survived the process were simply tall vertical columns - no intelligence, no muscles, nothing. The trick was that the creatures simply fell over. They'd grown taller so that when they landed, they'd be closer to the finish line - and the tallest creature 'won' the race every time. In essence, the creatures had found an 'algorithm' (falling over) that solved the given problem (crossing the finish line as quickly as possible) in a manner that was vastly simpler and more efficient than the human designer imagined. SteveBaker (talk) 21:29, 21 December 2008 (UTC)[reply]

What does the "LIO" prefix mean in the scientific name of "Pocket Mice"?

edit

Many scientific names of species have prefixes that define a group of animals within a larger group. Thus, among the various mice ("MYS" in classical Greek) the earth-based mice called in English "pocket gophers" are named in the scientific taxonomy Geomys (literally "earth-mice"). In Greek the prefix "Lio" denotes smooth, but in Latin the same prefix means lion-like or leontine. Since Greek and Latin are used indiscriminately in nomenclature, it is impossible to guess what an animal's name with that prefix means. My question is therefore twofold: 1) In the particular case of "LIOMYS", the Mexican Pocket Mouse, does the "Lio" prefix mean smooth or leontine? 2) In general terms, is there a way to tell which is correct, or does one need specific knoledge for each and every case this prefix is used? Thank you! --Bergeronz (talk) 20:11, 20 December 2008 (UTC)[reply]

You also asked this on the Language reference desk and it has been answered there. Please do not multi-post. If you really aren't sure which is more appropriate, you could post a query on one desk and a link to it on another. --Anonymous, 03:20 UTC, December 21, 2008.