Archive for October, 2008

The terrible truth about economics

Friday, October 31st, 2008

“Compared to physics, it seems fair to say that the quantitative success of the economic sciences is disappointing,” begins Jean-Philippe Bouchaud,  an econophysicist at Capital Fund  Management in Paris. That’s something of an understatement given the current global financial crisis.

Economic sciences have a poor record of success, partly because they are hard (Newton once pointed out that modelling the madness of people is more difficult than the motion of planets). But also because economists have singularly failed to apply the basic process of science to their discipline.

By that I mean the careful collection and analysis of observable evidence which allows the development of hypotheses to explain how things work.

This is a process that has worked well for the physical sciences. Physicists go to great lengths to break hypotheses and replace them with better models.

Economics (and many other social sciences) works back to front. It is common to find economists collecting data to back up an hypothesis while ignoring data that contradicts it.

Bouchaud gives several examples.  The notion that a free market works with perfect efficiency is clearly untenable. He says: “Free markets are wild markets. It is foolish to believe that the market can impose its own self-discipline.” And yet economists do believe that.

The Black-Scholes model for pricing options assumes that price changes have a Gaussian distribution. In other words, the model and the economists who developed it, assume that the probability of extreme events is negligible.  We’re all now able to reconsider that assumption at our leisure.

Bouchaud could also have added the example of economists’ assumption that sustained and unlimited economic growth is possible on a planet with limited resources.   It’s hard to imagine greater folly.

So what is to be done? Bouchaud suggests building better models with more realistic assumptions;  stronger regulation; proper testing of financial products under extreme conditions; and “a complete change in the mindset of people working in economics”.

All these things seem like good ideas.

But Bouchaud seems blind also to the greatest folly, which would be to imply that the roller coaster ride that we have seen in recent weeks can somehow be avoided in these kinds of complex systems.

Various physicists have  shown that stock markets demonstrate the same kind of self-organised criticality as avalanches, earthquakes, population figures, fashions, forest fires…. The list is endless.

And of course, nobody expects to be able to prevent the spread of bell bottoms or earthquakes or avalanches. If you have forests, you’re going to have forest fires.

What people do expect, however, is to have mitigation procedures in place for when these disasters do happen. That’s where econophysicists need to focus next.

Ref: Economics Needs a Scientific Revolution

Why PAMELA may not have found dark matter

Thursday, October 30th, 2008


This is the one we’ve been waiting for. For months, the astrophysical world has been abuzz with rumors that the orbiting observatory PAMELA has found evidence of dark matter.

Various people have speculated on the nature of this dark matter but the PAMELA team has been cautious, refusing to release the data until they are happy with it. (Although that hasn’t stopped data being smuggled out of private presentations using digital cameras to capture slides).

Now the wait is over. The PAMELA team has put its data on the arXiv and the evidence looks interesting but far from conclusive
Here’s the deal: PAMELA has seen more positrons above a certain energy (10GeV) than can be explained by known physics. This excess seems to match what dark matter particles would produce if they were annihilating each other at the center of the galaxy. That’s what has got everybody excited

But there’s a fly in the ointment in the form of another explanation: positrons of this kind of energy can also be generated by nearby pulsars.

So PAMELA isn’t the smoking gun for dark matter that everybody hoped. At least not yet.

For that, we’ll need some way to distinguish between the positron signature of dark matter annihilation and the positron signature of pulsars.

That means a whole lot more data and some refreshing new ideas. You can be sure that more than a few  astrobods are onto the case.

Ref: Observation of an Anomalous Positron Abundance in the Cosmic Radiation

Why ant colonies don’t have traffic jams

Wednesday, October 29th, 2008


Traffic jams are the bane of modern life. But could it be possible that one of this planet’s more ancient life forms could show us how to better regulate road traffic?

That’s the claim of congestion expert Dirk Helbing at the Dresden University of Technology in Germany and pals using a remarkable insight gained from the study of ants.

It turns out that ants are able to regulate ant traffic with remarkable efficiency. Let’s face it, you never see ants backed up and idling along a pheromone scent trail. On the contrary, ant colonies are a constant blur of organized and directed motion. How do they do it?

To find out, Helbing and pals built an ant motorway with several carriageways between a nest and a source of sugar. The carriageways had several interchanges where the ants could switch between longer and shorter routes.

Some ants soon found the shortest route to the sugar and others followed the pheromone trail they left behind until this shortest route became saturated with ants going to and from the sugar.

Then something interesting happened at the interchanges between the carriageways. When the route was about to become clogged, the ants coming back to the nest physically prevented the ants travelling to the sugar getting on to the highway. It wasn’t a conscious action, there simply wasn’t room for two ants to pass at these congested spots. So these ants were forced to take a different route.

The result was that just before the shortest route became clogged, the ants were diverted to another route. Traffic jams never formed.

That’s an impressive feat because the efficient distribution of limited resources by decentralized, individual decisions is still an open problem in many networked systems. As Helbing puts it: “This is one of the most challenging problems in road traffic and routing of data on the internet.”

But one that ants seem to have cracked and this gave Helbing an idea. Obviously, you can’t allow cars to collide with vehicles coming in the opposite direction as a form of traffic control; but you could do the next best thing and allow them to communicate. His plan is to force cars traveling in one direction to tell oncoming vehicles what traffic conditions they are about to encounter so that they can take evasive action if necessary.

And it’s not just road traffic that might benefit. Helbing speculates that all kinds of routing processes could benefit from a similar approach.

Simple really, if you’re an ant.

Ref: Analytical and Numerical Investigation of Ant Behavior Under Crowded Conditions

Martian methane and the smoking gun of life

Tuesday, October 28th, 2008


The presence of methane in the Martian atmosphere is a puzzle. Methane is broken down rapidly by sunlight and cannot last long in any atmosphere. A few simple calculations show that the lifetime of a CH4 molecule in martian climes is around 500 years. So methane ought not to exist in the Martian atmosphere at all unless it is being replaced on a regular basis.

So where could it have come from? Astrobiologists find this exciting because the methane in Earth’s atmosphere comes largely from cow farts (or more precisely the bacteria that live in their guts). The absence of  cowpats on Mars rules out the presence of ruminants on the red planet but leaves open the possibility that another primary source could be responsible, such as bacteria .

But before they can consider the possibility of life on Mars, astrobiologists must rule out every other possibility. One of these is that clathrates near the Martian surface are constantly releasing small amounts of methane as temperatures and pressure near the surface change.

Now Caroline Thomas et amis at the Universite de Franche-Comte in France have worked out how likely that is and say there are two possibilities.

First, they say that clathrates could only exist near the surface of Mars if the atmosphere had once been methane rich (otherwise the clathrates could never have formed). Perhaps the atmosphere was once temporarily enriched by a comet impact.

Second, there has to be some other source of methane, perhaps biological.

So how to distinguish between these scenarios. The discovery of gray crystalline hematite deposits on the surface could be a proof of an early methane-rich martian atmosphere.

Otherwise a biological source is a real option. Let’s get those rovers a-huntin’.

Ref: Variability of the Methane Trapping in Martian Subsurface Clathrate Hydrates

Slime Mould intelligence points to a new model of AI

Monday, October 27th, 2008


Earlier this year, a group of Japanese scientists reported that with appropriate training, the true slime mold Physarum polycephalum can anticipate the timing of periodic events.

That’s more than some politicians can manage and P polycephalum is only a single-celled amoeba, albeit a talented one. A few years ago a Hungarian team showed that slime mold was able to find the shortest way through a maze.

Clearly, primitive intelligence has cellular origins but how might this work?

Yuriy Pershin at UC San Diego and pals think they know how. They say that this kind of behaviour is identical to the way a simple electronic circuit reacts to train of voltage pulses. The circuit consists of an inductor, capacitor and a memory-resistor, or memristor.

Interestingly, this learning behavior comes from purely passive components. This can easily be reproduced in the lab and the San Diego team say it may turn out to be a useful way to build passive circuits that learn.

Link several of these passive learning circuits together and you might be able to knock up a simple neural net. Suddenly, you’ve got a new kind of AI on your hands and the origins of cellular intelligence don’t seem so obscure, after all.

Ref: Memristive Model of Amoeba’s Learning

Beacons ‘n’ torches

Saturday, October 25th, 2008

The best of the rest from the physics arXiv:

Rapid Sequencing of Individual DNA Molecules in Graphene Nanogaps

Proving Inflation: A Bootstrap Approach

The Straw Man of Quantum Physics

Quantum Big Bounce

Star Formation in the Multiverse

Why SETI will have missed any cost conscious ET civilizations

Friday, October 24th, 2008


If we want to contact any of those other civilizations out there, we’ll need a beacon to send messages with. But what to build?

Gregory Benford at the University of California Irvine and family (?) have done a cost/benefit analysis on the types of microwave generators out there that can produce the 10^17 W necessary to reach a significant proportion of the galactic habitable zone.

There are various ways that the cost can be optimised and the Benfords summarise them like this:

“Thrifty beacon systems would be large and costly, have narrow searchlight beams and short dwell times when the Beacon would be seen by an alien observer at target areas in the sky. They may revisit an area infrequently and will likely transmit at high microwave frequencies, ~10 GHz. The natural corridor to broadcast is along the galactic spiral’s radius or along the spiral galactic arm we are in.”

This has implications for the search for ET civilisations (as opposed sending messages for them). If ET civilisations are as cost conscious as we are,  then they may well have built beacons in this way.
And if so, say the Benfords, nearly all SETI searches to date would have missed them.

Ref: Cost Optimized Interstellar Beacons: METI Cost Optimized Interstellar Beacons: SETI

Frustration with fluid dynamics

Thursday, October 23rd, 2008

There is no shortage of fascinating videos for the Gallery of Fluid Motion at the upcoming meeting of the American Physical Society Fluid Dynamics division.

At least they sound interesting. We’ll never know because they’re practically impossible to download from eCommons library at Cornell University. That’s not good enough.

Surely YouTube (or one of its cousins) would be a better way to display these videos quickly, easily and above all reliably.

By all means place a hi-res version on the eCommons library for whoever has the patience to download it, but spare a thought for the rest of us. Spread the lurv, is all I’m saying. There’s plenty to go round.

Here’s a list of just a few of the interesting-sounding videos on the arXiv. If you’re thinking of downloading any, good luck.

Dynamics of Water Entry

The Bounce-Splash of a Viscoelastic Drop

Tornadoes in a Microchannel

Liquid acrobatics

The Clapping Book

Atoms in the Surf

Update: Turns out the videos are already on YouTube (see comments). Doh! (Might have been useful to mention this in the papers.)

Thanks Sheila!

Here come the quantum robots

Wednesday, October 22nd, 2008


Quantum robots were first investigated in the late 1990s by Paul Benioff, a remarkably original thinker at Argonne National Laboratory in Illinois.

Benioff is currently occupied in holding a candle for a theory of everything based on quantum numbers (more or less single handedly).

So a team of Chinese physicists led by Daoyi Dong at the University of Science and Technology of China in Hefei , China, has taken up the challenge to develop our ideas about quantum robotics a little further.

Benioff’s work explored the way in which a quantum robot might explore a 2D or 3D space using the laws of quantum mechanics to speed up the search. If memory serves, there is a decent speed up in two dimensions but not in three (which has interesting implications for molecular building machines). But he gave no thought to the internal structure of his robots or how they might be constructed.

The Chinese team have  now given form to this structure. Quantum robots, they say, will consist of three parts:

i. an information processor consisting of one or several quantum computers

ii. some kind of quantum actuator that interacts with the environment to carry out a task

iii. a quantum sensor which monitors the environment, such as a SQUID  (superconducting  quantum interference device)  which detects magnetic fields.

The team has mysteriously omitted a quantum communication module to send and receive data from their classical masters.

So what can a quantum robot do that, say, a classical robot attached to quantum sensors cannot?

That’s not entirely clear. Daoyi and co  say that most planning and control problems in robotics can be posed as search problems. So Grover’s search algorithm gives a significant speed up in the time it takes to solve these problems. But presumably the same would be true of a classical robot controlled by a quantum computer.

Where they might prosper is in size. Presumably quantum robots will operate at a scale that is not accessible to classical robots. And this raises the prospect of a world beneath our own populated by quantum machines operating on entirely different principles to ours. All this needs some fleshing out.

The China team’s vision is far more sanguine, having dreamt up the following predictable  applications. They say:
“Quantum robots have many potentially important applications in military affairs, national defense, aviation and spaceflight, biomedicine, scientific research, safety engineering and other daily life tasks.”

(It must have taken them weeks to think up that list.)

Now all they have to do is build one of these guys.

Ref: Quantum robot: structure, algorithms and applications

How leaves curl up in strong winds

Tuesday, October 21st, 2008


Various types of plants, fungi and even animals are known to change their shape in strong winds to reduce drag.

Leaves, in particular are known to curl up in strong winds. How they do this is not well understood, because of the dynamic nature of the problem and the difficulty taking good data.

So Laura Miller and pals at the University of North Carolina at Chapel Hill have had a stab at working out what’s going on by videoing leaves in wind.

They’ve also videoed the way some leaves aquaplane in water, that is rise to the surface to reduce drag. Curiously, leaves from herbaceous plants seem to display the aquaplaning behaviour whereas tree leaves do not, presumably because herbaceous leaves, being generally closer to the ground, are more likely to be caught in flood waters.

There are obvious selection advantages for organisms that can survive strong winds and floods so it’s no surprise that some kind of protection mechanism has evolved.  That makes it all the more fascinating to see it in action.

The video is here and is worth a look if you have a few minutes.

Ref: Leaf Roll-Up and Aquaplaning in Strong Winds and Floods