Does Farting make you Weigh Less?

17 January 2010

Can sea water keep roads frost-free? Why does the LHC need to be so cold? How does antifreeze work and what's the freezing point of beer? This week we run the risk of frostbite to tackle the coolest science questions as well as warming up in the hot tub to hear what household appliances devour the most power! We also find out how researchers are growing stem cells from umbilical cord blood, the scientific reason why a needle is so hard to find in a haystack, and how the smell of a fertile woman boosts a man's testosterone. Plus, do people really look like their pets?

In this episode

How many days in a month can a woman become pregnant?

Well, the fertile window is around the time of ovulation, which is day 14. So you're really looking at 2 or 3 days either side of that because sperm can persist or survive inside a woman for 2, maybe 3 days. And so that's why it's really important to use safe sex practices ahead of ovulation. Also, the time of ovulation can vary by a couple of days in some women, so it's really important to avoid, if you don't want to get pregnant, unsafe sex in that week in particular.

Do Women attract men more when fertile by subconsciously acting differently?

Well that was what people said about the study done [by Geoffrey Miller and his colleagues] on the lap dancers. [They found that lapdancers' tips shoot up by over 200% when they are at their most fertile]. Some suggested tThat perhaps the women were making themselves more attractive to the men and it was a female behavioural thing that was making them more attractive. In the present study [in which men were asked to sniff T-shirts worn by women at different phases of their menstrual cycles], what the researchers sought to do was to iron that out by making sure that the women who wore the T-shirts that were sniffed by the men only bathed using a neutral smelling soap. They all used the same cosmetics. They didn't wear perfumes and other kinds of things. They also avoided eating foods that might make them stink, like garlic. The researchers actually say that these steps were taken in the paper. So, in other words, they attempted to control for the possibility that the women were making themselves more attractive. Instead, what we think is happening is that it's actually some other chemical which is in some way stimulating the men, but we don't actually know what it yet. That will come next, I guess.

Why are my windows so drafty?

You're entirely correct about thermodynamics. If you've got two stationary objects and you join them together thermally, heat will always flow from the hotter object to the colder object. It's actually pretty much how we define temperature.

So what's happening with your window is because cold air is definitely coming into your house, this situation's a bit different. You're not touching two objects together and cold is moving into your hot house. What's actually happening is that cold air is moving in to your house. The laws of thermodynamics say nothing about where air goes, they don't disallow this. So basically, large lumps of objects can move around as they like, not ignoring the laws of thermodynamics, but they aren't constrained by them in that way.

Were medieval people less smelly?

Diana - I think for the most part, the answer is probably no. I don't know if you were ever a fan of Blackadder or anything, the stereotype for Baldrick, the smelly peasant, is probably quite true. And actually, if you've ever been to any of those reconstruction sites like Jorvik that's about 200 or 300 years before the medieval period, it's not quite medieval, but I think it gives you a really good idea of what things might have smelled like and they include also some toilet smells and cooking smells, mouldy things. Smells are pretty awful. And from what we know from the archaeology, there wasn't a lot of bathing going on for the peasants at least. And there was certainly, in the houses of medieval people, all sorts of bits of evidence for maggots and flies and things just living in the house with them, but it may have been for the more upper class medieval people like Lords and things that they did actually bathe maybe once a week. Chris - They used to keep their clothing near the loo in their castle, didn't they? Because castles had guarderobes which is the bogs and correct me if I'm wrong, but wasn't that done under the intention that the things that rot your clothes, moths and things that would make your clothes go horrible, would avoid the stink for the same reason we would? Diana - Yeah and they were probably slightly better ventilated than the rest of the places. Well, because of course their toilets were just, holes. Chris - Indeed because your clothes would smell otherwise, wouldn't they? Diana - Exactly. So, yes, they probably did smell pretty awful. Also, actually, there's another thing I'd like to add is that if you're constantly being exposed to horrible smells, you're going to be desensitized to them, aren't you? So perhaps, in that sense... Chris - I think there are limits.

Does too much calcium make your bones brittle?

The answer is, yes it does, surprisingly. Calcium links up with phosphates to make the chemical "apatite" (calcium phosphate), one of the hardest chemicals we know, which is what makes bones hard and stony, and tough. But if you have too much calcium, that can be as bad as having too little, as in condition osteoporosis where the bones actually begin to lose their calcification and the matrix of the bone, which makes them weak and more likely to break. There's also another disease called "osteomalacia" where you have too little just of the calcium and that also makes bones weak. But some people actually lay down too much calcium in bone, a condition called "osteopetrosis" from "petros" as in stony. This is where people can have say, five times the amount of calcium in their bones that they should have. It's a genetic condition. It's very rare and I think it also goes by the name "Albers-Schonberg disease" or something like that, but it's very rare. In these individuals, the cells that break down bone, called "osteoclasts", don't work properly. This is because the way bone is normally formed involves an equilibrium between laying down new bone and breaking down old bone, and that's how bones are continuously remodelled. So if you shift that equilibrium on one direction or the other, you either lose or gain bone. And what scientists have found is that in people who have osteopetrosis, the osteoclasts that normally breakdown bone cannot work properly. They have a deficiency of an enzyme called "carbonic anhydrase" and you need that to break down the calcium phosphate, the apatite, in order to remodel the bone. As a result, they just keep on making their bones get harder and harder, and harder. Eventually they go beyond the point of this being beneficial and the bones become less flexible and more likely to break, so they get very brittle.

31:35 - Artificially Intelligent Games

Meera Senthilingam becomes Sherlock Holmes to investigate the uses of Artificial Intelligence in the world of online gaming... and possibly in call centres!

Artificially Intelligent Games
with Chris Vallance, BBC Technology Correspondent & Rollo Carpenter, Icogno Ltd

Meera -   It's time for our first technology update of 2010 and I'm here with our resident techie, Chris Vallance.  Hello, Chris and happy new year!

Chris V -   Happy new year to you too, Meera.

Meera -   So, what's been happening in the world of technology over Christmas?  I hear that artificial intelligence has been playing a big part.

Chris V -   Well, you said 2010, so I'm going to take you to 221B. 221B Baker Street home of Sherlock Holmes.  Over Christmas, we had the release of the new Guy Ritchie film.  As Sherlock Holmespart of a trailer for that, if you like a digital interactive trailer, there's an online game called 221B.  Now what's interesting about that is the way in which it uses artificial intelligence.  As part of the game, you get to play as Holmes or Watson and you have to interrogate suspects.  You have to ask them questions and the answers they give help you complete the game.  The problem is, when you're asking a game character something, you could ask them it in any number of ways.  Just using ordinary English, the possibilities are endless.  So your options are: let's do it like one of those old, choose your own adventure game books - essentially, "if you want to do A, click here, if you want to B, click here."  That's a bit limiting.  So what the game designers decided to do is approach an expert in artificial intelligence to help them enable natural language type interactions with the game characters.

Meera -   So how does this actually work and what's different about it?

Chris V -   This is one of the big challenges in game design.  Natural linguistic interaction with non-player characters.  What this does is it uses chatbot technology to recognize what a question means.  Now it doesn't have any understanding of the meaning of words.  It doesn't have lots of long lists of complicated grammatical rules.  What it does is fuzzily pattern-matches your question.  So, it's very good at matching your question with the questions it knows how to answer, it does that and produces an answer according to the script.  So the real key here is its ability to recognize questions and know what the appropriate answer is.  In effect, it's creating a digital actor in the game, with a script, but one who can understand kind of what you're saying and respond appropriately.

Meera -   And now, you actually met up with the creator of this to find out a bit more.

Chris V -   Yes, I spoke to Rollo Carpenter who developed the artificial intelligence running the game characters and he explained how his system worked...

Rollo -   My work has been all about creating artificial intelligence that talks, that communicates about anything and everything, and that attempts to look intelligent. To appear to be self-aware.  The same fuzzy technology, fuzzy interpretation of inputs is used in the game. So it is our role to predict what you as Holmes or Watson, might know at that point of the game and the conclusions you might draw, the questions you might ask.

Chris V -   But given the nature of language, there are a near infinite number of possible questions you might ask. Some grammatical, some not grammatical.  So how does the computer understand the question?

Rollo -   It understands by predicting what you will say and finding the nearest match amongst its trillions of predictions.

Chris V -   So essentially, it's a kind of pattern recognition.

Rollo -   Exactly.  It is a kind of pattern recognition, but where the patterns are all treated fuzzily.  If you mistype a few letters, it's not going to be upset.

Chris V -   You've used the technology you've developed to build artificial personalities, but there are clearly commercial applications for this kind of technology.  What do you see as its potential uses?

Rollo -   A machine that can genuinely understand what people say is valuable in almost any number of ways.  But some of the most obvious and immediate concern with customer services, support, marketing, and so forth.

Chris V -   So when I ring up the call centre at some point in the future, I might get a computer?

Rollo -   Indeed.  The people at the end of the phone all too often have to work to a very restricted script and a computer would be capable of handling scripts 10,000 times larger.  And so, it may actually do its job enormously better.

Chris V -   That was Rollo Carpenter whose chatbot technology drives the characters in the game.

Meera -   So Chris, have you actually had a go on this and if so, were you able to solve a mystery?

Chris V -   I didn't get as far as solving a mystery, but I did try out interrogating one of the characters and you can see the results of that up on a video on BBC News online.  It does work surprisingly well actually, asking questions in natural ways, even trying to catch it out a little bit. 
You can still have a go with the game online.

Meera -   So, I'm going to probably have a go on that as soon as possible, but is this something we're likely to see more of in other areas of gaming?

Chris V -   I think it's hard with the big budget games.  Game designers have told me the challenge there is essentially that you have very complicated pieces of animation for console games when you have these stunning visuals.  You get voice actors to do the dialogue.  So, spontaneously creating dialogue and the action to go alongside it is quite a big technical challenge, just from the sound and visual side, let alone making that dialogue meaningful.  So this is still an area where there's a lot of development to be done.  It's still a big technical challenge.

Who uses the most electricity?

Diana - I'm actually a bit of a games consoler myself. So, this was quite enlightening for me. I've got some numbers for you. I've made the assumption that you pay 15 pence per kilowatt-hour, assuming the Euro is on parity with the pound at the moment. So let's imagine that your husband is playing on a 42-inch LCD TV. That's pretty big and will draw about 200 watts. So that will cost about 3 pence per hour and if you're doing two 5-hour sessions over weekends, that's going to be about 30 pence per weekend. Now his Xbox on its own draws about 160 watts, so that's going to be 24 pence for the same amount of time. The total will be about 54p for a 10-hour weekend session. Now the hot tub is where it gets a bit serious, I'm afraid. From my research, most hot tubs seem to use about 2,000 watts which is a lot and that's going to cost about 30 pence per hour, so it's going to be 90 pence for a 3-hour session, Ã?,£2.40 for a full [8-hour] day and for a full week, Ã?,£16.80. Diana - And then if you want to compare it with your bulb usage, so if you've got energy saving bulbs, I'm guessing that you're good boys and girls over there. Amanda - I don't use them for all the lights, but we've got a lot of spots in our house. Diana - Well if you're using a nice 20-watt energy saving bulb and you've got three of them for 8 hours overnight, seven days a week, that's going to cost you about 50p. But if you've got the old fashioned incandescent ones, that's going to be Ã?,£2.52, but I'm afraid the whirlpool there or the hot tub, that's going to be the real killer.

Why does the LHC need to be so cold?

Dave - There are temperatures in the LHC of about 1.9 Kelvin. That's 1.9 degrees above absolute zero, so about minus 271 (and a bit) degrees centigrade. There's a couple of main reasons for doing this. One of the big ones is that inside the LHC there are huge magnets which are used to bend the particles around corners. They have incredibly strong magnets and you can't do that with a normal permanent magnet, you just can't get them strong enough. So you have to use an electromagnet. If you made the electromagnet out of normal copper wire, it would just melt in no time because of the huge current you need to get that strong a field. So, what they have to use is superconducting magnets.

A superconductor is a material which, when you cool it down enough, its resistivity goes to zero, so it will pass a current with no resistance at all. This is great for a magnet because you just start a current flowing in it and keeps on flowing forever.

The problem with superconductors is the bigger magnetic field you apply to them, the lower the temperature they start working. So, if you want to make incredibly strong magnets, you have to use very, very low temperatures and they're using about 1.9 degrees above absolute zero. This is actually below the temperature of liquid helium at normal atmospheric pressure so what they have to do is they take liquid helium which they have made essentially with a very, very big fridge by compressing gases and letting them expand. And as they expand, they get colder. You then take liquid helium and pump on it, so you reduce the pressure above it. So then it boils and boils, until it gets colder and colder, down to about 1.8 Kelvin. Then they pump that around the system and cool down the magnets. Chris - I would come off there by saying, "Cool!" but that would seem like an awful pun. But the bill for doing that must be absolutely huge or once you've got it to that liquid state, is it so well-insulated that it just stays that way? How does it work? Dave - It is going to need a huge amount of energy because although they're not moving that much heat out of the system, they're probably moving a couple of hundred kilowatts of energy, which by the standards of huge industrial processes isn't very much. The problem is the colder the object you are taking energy from, the harder work it gets. So, if you move energy from somewhere at room temperature to somewhere a couple of degrees hotter, then that uses very little electrical energy if you move it from something at 1.2 Kelvin to a room temperature that - with quite an immense amount of energy. Chris - And will you get a magnetic field once you've fired up the magnet that would just sustain itself because there's no resistance and will you only be losing the magnetic field because the particles will sap energy from it? So, it will therefore need to be topped up for that reason.the magnetic field that will be generated at those low temperatures,

Dave - The superconducting magnets, once they start going, they essentially work like a permanent magnet. People have done experiments with superconducting magnets and as far as they know, some of them will keep on going for billions of years. The real problem is if part of the superconductor warms up, at which point, you get a huge amount of electrical energy running around it. All of which is dumped into a very small area of the superconductor, so it gets very, very hot. All this liquid helium which you've got inside the magnet suddenly boils. And that produces a huge amount of gas which expands very, very rapidly. This was the problem they had last year when it went horribly wrong. This huge amount of gas, then pushed the magnets around. Billions of pounds worth of stuff just shuffled around this system and it caused an immense amount of damage.

How do mosquitoes lay their eggs?

Mosquitoes depend on water for their life cycle. Different mosquitoes live in different environments and as a result they depend on different types of water - stagnant water, big ponds of water, dumped car tyres with a bit of water and so on. You can actually tell which species of mosquito you're dealing with depending upon where they're laying eggs.

The bottom line is the mosquito goes down to the water once it's mated and different species of mosquito lay eggs in different ways. But they all lay their eggs into water.

Certain species lay them as individual eggs which drift off; others lay big rafts of eggs. The eggs mature and hatch into little larvae, which then have an aquatic phase; they grow in the water and eat algae and things, growing into large mosquito larvae. These then mature into full blown mosquito flies that then take off and come and bite you. You just have to keep an eye on a patch of water and you'll see the mosquitoes coming down to lay their eggs!

Can we create artificial nerve signals?

Chris - Yes we can. We actually know quite a lot about how nerve information travels. If you can imagine a nerve cell as a bit like a long straw, with sides and a space in the middle - what nerves do is to move positively charged ions, in this case sodium ions, from the inside of the nerve, to the outside world. So the inside of the nerve is a bit negative compared to the surroundings. When a nerve impulse travels along a nerve, what happens is that some "positive" (some sodium) goes back inside the nerve through tiny pores, which are on the surface of the nerve. This is called de-polarising the nerve, and it makes that section of the nerve become, transiently, a bit positive.

Now this does two things; it starts an electrical signal rather like a Newton's Cradle running inside the nerve, but it also activates other little channels and pores on the surface of the nerve a bit further downstream. They open and let in some more "plus" to sustain and maintain the propagation of the signal. In a big nerve, this signal is actually travelling along at something like 100m/s. The impulse (or action potential) only goes in one direction though because, in the opposite direction - where it's just come from, the nerve pumps the "plus" back out again, so it goes back to being net minus and the nerve resets itself.

This process happens in milliseconds, so you can literally conduct hundreds if not thousands of these impulses down a nerve in less than a second. The information can travel very, very fast.

You can also make this happen by artificially stimulating the nerve. If you apply a little bit of electricity to the wall of the nerve fibre itself you can actually make that process trigger off, and then it self-sustains. The signal will propagate along the nerve to wherever it goes - in both directions.

Scientists use this for a number of reasons; one is that you can artificially activate muscles that way - so if you've been paralysed, for example, you can use techniques like this to restore movement to certain muscle groups by electrically stimulating certain nerves that supply those muscles. Another reason is to use brain stimulation - this has been done quite effectively in Parkinson's disease. Scientists implant a little electrode in a part of the brain which makes movements and is involved in the same circuit as is affected by Parkinson's disease. If you stimulate those bits of the brain electrically, you can trigger off impulses in the right way and the right rate to help people who have Parkinson's symptoms to overcome their symptoms and move a bit more easily.

So, it is possible to re-create nerve signals. At the moment it's fairly low resolution: you're not stimulating individual nerve cells, you're stimulating clusters of nerve cells. But at the same time, you can do that.

Also, if you listened to last week's programme, you'll know that we talked about cochlear implants, which are things that basically stimulate the nerve that conveys hearing information into the brain stem. They're doing effectively the same thing - stimulating nerve cells directly to send sound information into the brain.

Diana - I remember about a year ago we had a story on Breaking Science with the researcher Andrew Schwartz who monitored the signals being sent from Chimp's brains, and then translated them using a computer into muscle movements.

Chris - I guess that's going the other way - rather than putting input into a cluster of nerve cells, he was recording the activity, understanding how nerves are encoding information and using that (once it's been decoded by a computer) to then drive muscle movements in an arm.

I think they were making the Chimpanzees reach out and grab things just by the power of thought.

Diana - Yeah, that's right.

Melting Ice with Salt

We were asked why councils don't use seawater to melt ice on the roads, so why not find out the answer with an experiment?

Does beer have a lower freezing temperature than water?

Dave - Yes. Beer's got lots of things dissolved in it - alcohol, sugar etc. Anything dissolved in water will reduce its freezing point. So beer does freeze below zero, maybe at -1 or -2.

Chris - And CO2, I suppose, if it's fizzy - that will have an effect too?

Dave - That will also have an effect, definitely.

Chris - So if you were to take some beer and put that in the freezer, could you make it much stronger? Is this a way that you could illicitly do a bit of distillation, because you would freeze the water into ice first, and then some slightly more alcoholic liquid will be left?

Dave - That's right. When you freeze water it can't take in all these solutes, the dissolved things, into it's structure, because ice remains really quite pure. What's left gets stronger and stronger. I don't know if you've ever drunk frozen squash or frozen fruit juice? The stuff you drink first, which is the liquid which melts first, is incredibly strong and very sugary, and then slowly as it melts it gets weaker and weaker, until what's left is basically pure water.

Does internal gas affect your weight?

Diana - Well we have had a bit of a discussion about this and I think what it comes down to is what your farts are composed of. Of course, there is some hydrogen sulphide in there and occasionally some methane. I think it depends what kind of bacteria you have in your gut but I think it was about five out of seven people will have methane produced in their farts. And of course, methane is lighter than general gas (air) and therefore, younwould think if you have that in your gut, then it would make you a bit lighter. But of course, there's a question of pressure. So, if you have a lot of gas in your body that's under pressure, it would actually be denser than air and actually make you heavier. What do you think Dave? Dave - Yeah. The very things which could be in the gas, one which could really make the constituent is hydrogen which is one of the lightest gases we know of and that can be up to 50% of your flatulence and so if it's made mostly of hydrogen then it's going to be much less dense and the pressure change is going to be very, very small. So when you expel the gas, it could even make you heavier because the hydrogen could've been acting a bit like a helium balloon in floating you ever so slightly. Diana - I don't know if you could do a test by seeing if that gives a squeaky pop with a lighted splint? Chris - Maybe Kitchen Science for the future.

Does tapping a drink can to stop it exploding work?

It does certainly have an effect. Or it can have an effect. I haven't got it to work 100% but it's supposed to help. The idea is that if you have shaken up a can, the reason why it explodes is there's lots of little bubbles of carbon dioxide in the liquid already. That gives the dissolved gas lots of nucleation sites - places to start bubbles, and they can get bigger very, very quickly. Because these bubbles start under the liquid, it all explodes and forms a foam. If you tap it, the idea is you dislodge the ones which are stuck around the bottom and along the sides to get fewer bubbles in there at the bottom, so it makes the foam much more slowly and it doesn't make a mess.

Do all satellites orbit in the same direction?

Well the answer is, in most cases, yes. This is because they condense out of what's called a proplid, or proto-planetary disc. This is a big spinning disc of material which initially starts off as an envelope around a star that forms the centre of the solar system or in a particular galaxy area you're looking at. This disc of matter then condenses or accretes into planetesimals, little objects that accrete more matter to become much bigger objects. And since angular momentum is conserved, this means that they will all be spinning. Then, once everything has collided with everything else that it's going to, and when you look at the equations, you end up with a cluster of bodies that are usually all turning in one net direction.

But that's not to say that big collisions can't come in and then turn things around a bit. Uranus, for example, is turning on it's side, and we think that this is possibly due to a big collision with something else way back in history.

So the answer is yes. We think things do go around in one direction and in the same plane, but there are exceptions at the level of individual bodies, which can spin in funny directions.

61:16 - Do people look like their pets?

It’s often said that people resemble their pets or vice versa; is this a phenomenon that's been studied? Do people choose pets that resemble them, or is this a matter of us noticing...

Do people look like their pets?

We put this question to Dr Lance Workman, psychologist at Bath Spa University:

I think there are two reasons why pets look like their owners. A first reason is that we tend to buy pets, and dogs in particular, that fit in with our lifestyle. So if you are an active robust person with lots of energy, who likes to go outdoors a lot, then you may well buy a dog that is likewise full of energy. It's a bit like selecting a spouse. It'll work really well if you find someone who fits in with your lifestyle. That's the first reason. I think the second reason is a more subtle one and that is that I think we're drawn to others that look a bit like ourselves. It's a bit of an old Freudian notion, but I think it does stand up to scrutiny. So, at some subconscious level when we chose a pet, we're looking for something that in some way reflects ourselves. I think there's both of these reasons going on there. I think you begin to look for similarities rather than differences. So I think if we say, "Oh, here's Fred and his Jack Russel - don't they look similar?" You're already led along that path. There's something that some people call confirmation bias. For example, if you go to a fortune teller and they say six things out of 20 that are correct, you go away and you remember those six things and you forget the other 14. My own study suggests that people get this right about 2/3 of the time, maybe 60% or something like that. So there's a good 30 to 40% of the time they are wrong in matching pets up to their owners.

I think, again, we remember the ones we get right. But 2/3 is well above chance, so it's a mixture of some real truth in the idea that we look like our pets and the fact that we look for evidence that supports that case. Diana - So, although we do tend to remember better those pets which resembled their owners, it looks like many pet buyers will tend to choose animals which look a little bit like they do. And in Lance's research, he tested this by showing 70 volunteers photos of dogs and dog owners and asked them to match one to the other, and 60% of the time, they got it right.

Add a comment