Explosions in Afghanistan

A visualisation of the Wikileaks Afghanistan data, showing explosions of improvised explosives between 2004 and 2009.



I was thinking of doing something like this myself. With minimal programming effort it's possible to generate all sorts of information about the movements of particular task forces, covariances between task force activities, trends in use of robots and drone aircraft, and so on. For military historians, this is an amazing resource giving an unprecedentedly detailed view of how a modern war was prosecuted. The comprehensive nature of the data suggests to me that it was leaked by someone at a high level in the chain of command, perhaps unhappy with the way things are going.

From a robotics perspective, there are some detailed eyewitness accounts of Pacbot and Talon robots being used to dispose of roadside bombs. Assuming that these are at least moderately truthful it provides some evidence that robots really are useful, and on quite a number of occasions have saved lives - sometimes getting destroyed themselves in the process. Also, there are a number of occasions where operators describe having tried unsuccessfully to use a robot to dig carefully around or under an IED, so for robot designers at iRobot and elsewhere I think it would be a good idea to build a small articulated digger attachment (perhaps hydraulic) which can scoop out small amounts of soil.



One thing which I did check out within the data is whether ground based robots are being used not just in the traditional capacity for bomb disposal but also as offensive weapons, as depicted above. I couldn't find any such incidents, so in spite of the promotional videos showing roaming machine gun equipped Talon robots it looks as if this isn't what's happening at present - although it's surely only a question of time and confidence before that changes.

As far as I'm aware the Afghan war isn't an illegal war, in the same sense that the Iraq war was (based upon a claim of self defense against WMD subsequently confirmed to be false). When it commenced in 2001 I was actually in favour of going after the perpetrators of 9/11 but the war seems to have been conducted in a poorly focused fashion, suffering from a large degree of scope creep to the extent that they seem to be fighting rather tragically against the local population in a manner very reminiscent to the occupation by the former Soviet Union in the 1980s. The above animation strongly suggests that the situation is not becoming any more peaceful over time, and that sooner or later a Soviet-style withdrawal is the logical conclusion. It's just a pity that so many lives and so much money had to be wasted in the exercise, whilst the principal reason for entering that country (going after Bin Laden) has still not been properly addressed. Imagine, if you will, the same amount of money which has been squandered on fruitless adventures in Afghanistan being spent on space exploration, medical research, artificial intelligence or renewable energy projects and how that would would have provided positive returns in terms of knowledge and economic activity.

General intelligence

A video showing a mother bear retrieving her offspring from getting trapped in a net by gnawing through it. This seems like fairly unremarkable stuff, but on further consideration it does require some flexibility of intelligence. The logic would be something like:

X is trapped by Y, therefore I need to attack Y to free X.

It requires some general notion of being entrapped, and dealing with the thing which is causing it rather than just naively tugging on X as a less intelligent creature might do.

As far as I know there is no robot currently which can solve this problem. Nets don't naturally occur in the wild, so the bear has no ancestral knowledge of how to deal with this specific situation distilled within its genome. Instead it's improvising in an adaptable way to a situation which it has not previously encountered.

If a robot gets stuck it may have some hard coded behavior which it uses to try to back out again, or flip itself over, but if it got trapped in a net and there was no net cutting software directly programmed into it (assuming that it has an arm capable of cutting) it's unlikely that it would be able to reason about the situation and find an adaptive response to an unknown hazard - at least with any software that I've seen used previously. This requires general intelligence, which includes being able to perceive the physical structure of the surroundings, formulate an abstract characterization of the overall situation ("I am trapped" or "my friend is trapped") which includes a goal state (of being freed) then select a behavior from a range of possible responses which is appropriate to move towards the goal.

Job losses

Robert Reich writes an article about the separation of company profits from job creation. I think this sort of phenomena will be part of a longer term trend over the next few decades, probably culminating in campaigns to divorce jobs from income.

I'm not as pessimistic as Reich and I think that in the near term many of the jobs will return, but with each economic cycle the number of jobs which can be automated increases, and most companies don't care whether a unit of labour is performed by a human or a machine, so long as it is done with the highest possible yield for owners and/or share holders. Currently it's possible to carry out some unit of labour more cheaply by outsourcing it to other areas of the world, where the cost of living is lower. Before long that's going to be supplemented with telerobotic working, which has similar economic effects to conventional outsourcing, with the final phase being job losses amongst the tele-workers as the telerobots are replaced by fully automated equivalents (replacing the human remote operator with some software).

Presently the level of automation even in the wealthiest nations is not sufficient for this to become an issue of popular debate, but I expect that by the mid part of this century it will be. Assuming no major disasters occur (like nuclear wars, and so on) which set the clock back, there will be a point at which human labour becomes essentially deprecated except for a few niche human-to-human (H2H) occupations.

So how will humans get along in a situation where their labour value is reduced to something close to zero? The up side is that in an automated economy there is no problem with supply of goods or services, so long as there is adequate recycling of materials and use of renewable energy for the system to be sustainable over the long term. The most obvious route is to have something similar to a citizens income, where each person is assigned an income which is some fraction of GDP per capita. This would be similar to having a minimum sharehold in the economy, but where it was either impossible or illegal to sell all of your shares.

Killed by code

The Software Freedom Law Center highlights the problem of proprietary software used in medical implants. I think in general any system which is controlled by software and where there exists the possibility that a bug could result either in death or in someone being sent to jail then the code should always be independently auditable. Being able to check the code for flaws means that you can have the highest confidence possible that the system will operate as intended, and reduces potential liability issues. If the code has been out there for a while, and many eyeballs have scrutinized it, then it would be much harder for accusations of negligence to be made in the event that a previously undiscovered bug is found.

The situation for robotics is not quite so drastic as for medical implants, but if you have a robot operating in someone's home which is bigger than a Roomba and has at least one manipulator then there exists the possibility of accidents or unauthorized usage due to software errors which could have damaging or fatal consequences, so a similar principle might apply. Large companies may be able to absorb the costs of legal issues arising from software related accidents, but smaller ones probably won't be able to. So if robotics is going to become a larger industry then at minimum at least adopting an "open core" approach may be advisable.

Moonwalk One

Building a better rocket

Here is a BBC report on amateur rocketry. Getting up to less than 3000 feet doesn't sound very impressive, and other types of engine would still work well at that altitude.

One way to go higher and stand a better chance of getting to space would be to attach the rocket to a balloon, which it would then be released from once the balloon had reached it's maximum altitude, but balloons are not easy to control and would presumably need to be quite large to lift a significant mass. A better way of doing things might be to mount the rocket at the centre of something like a large quadrotor helicopter. The multiple rotors could be spaced some distance apart to ensure stability during the ascent, then at the maximum altitude the rocket would fire and the helicopter could either descend back to earth in a controlled way, or merely just free fall and autorotate. The gliding descent would ensure that nearly all of the helicopter's fuel could be spent getting it as high as possible, with power only being used during the final landing phase.

Raindrops

An example of detecting raindrops on a vehicle windscreen.



This sort of detection could well be very useful, since for automatically driven vehicles or safety systems you don't want rain drops being detected as landmarks and being used for visual odometry. Rain is going to be difficult to remove from the image, but anything which can be done to improve the situation will help to reduce measurement errors.

Bottom up and top down

Richard Gregory talks about the directionality of information flow in visual perception, and I think this is now the prevailing view on how biological vision works. What he's really saying is that vision is a closed loop process requiring continuous synchronization, and from an informational point of view this makes a lot of sense if you assume that the incoming signal (rays of light) is noisy and ambiguous due to the lossy compression from three spatial dimensions of the observed scene down to two on the retina.

Automatica 2010

The war on blogs

A recent case highlights the pitfalls of having your blog "terminated without any notification or explanation". It's not unknown for individual blogs to be occasionally taken down by hosting companies, because they contain spam or are used to distribute malware or unauthorised versions of copyrighted material. However, in this case there was a mass takedown of 70000 blogs, which I assume must be because an entire domain was blocked, or maybe they really did erase all the data from the servers. It seems implausible that this number of blogs were all actively involved in terrorist plots. A more likely scenario is that there was just one or two offending blogs, and that someone either at the law enforcement level or the server administration level behaved incompetently and just yanked out the plug.

This highlights the problems of hosting your content on other people's servers - now popularly known as "cloud computing". Most of the time such hosting doesn't present any problems at all, and is very convenient, but it does mean that content which you've produced can be subject to terms and conditions which can be changed on a whim without your agreement, and your data can be erased without warning.

I've recently been setting up various pieces of web content on a plug computer, and this is just an experiment to see what's involved since it's always fun to play around with embedded systems. Perhaps plug computing is the way forward though for hosting basic services such as blogs or wikis, since it allows for data to be easily backed up and even if you are accused of posting objectionable content you have some reasonable chance of dealing with the issue in a civilized manner rather than just having all your content deleted immediately. The software for plug computers still seems to be in its infancy, so I wouldn't recommend this to the average non-technical user at the moment, but given a few more years of development it might be that hosting all of your content on a small, low power, permanently plugged in device is completely feasible for everyone.

The Big Society

Back in 1998 or 1999 the then prime minister, Tony Blair, gave speeches about what he called "the third way". These ideas sounded ok - supporting families, empowering communities, etc - but they were also incredibly vague and wooly and just hard to pin down to anything remotely anchored in the real world. David Cameron's "Big Society" reminds me of the same sort of earnestly delivered but ultimately vacuous rhetoric. There are the same kind of bland statements made, which are in principle hard to find disagreeable, but lack elucidation.

Even Wikipedia doesn't shed very much more light onto the topic, with the responses suggesting that "The Big Society" is just a ploy to replace public services with bands of unpaid and poorly motivated volunteers.

"The trouble is that running a bus route is a professional job, not for a group of local enthusiasts. How many bets that five years down the line, the enthusiasm has run out and there is no more bus route"

But suppose for a moment that this isn't just a cynical exercise, and is instead a genuine experiment in anarchism. Could you apply a sort of open source methodology outside of the scope of technology within the wider society? Open source projects aren't without hierarchy, but they do usually only have a minimal organizational hierarchy, and they also involve a lot of volunteerism by people who are highly motivated. The more successful open source projects, such as the Linux kernel, also usually attract funding or sponsorship of one sort or another.

Probably the biggest issue in terms of applying an open source methodology outside of the business of software authorship is scale. Most projects are small, and the individuals working on them are scratching a personal itch - addressing some problem which they're passionate about, rather than because they've been told "it's your duty to work on this code". Whether this would also apply in the case of providing public services seems more doubtful. Most people are not highly motivated by the nitty gritty details of providing public services, and expect that delivering this should be something mostly handled by paid professionals.

Directed strivings

"All these observations fascinate us, and fill us with an intense desire to know more about the nature of these phenomena. Each day we go to our work in the hope of discovering,—in the hope that some one, no matter who, may find a solution of one of the pending great problems,—and each succeeding day we return to our task with renewed ardor; and even if we are unsuccessful, our work has not been in vain, for in these strivings, in these efforts, we have found hours of untold pleasure, and we have directed our energies to the benefit of mankind."

- Nikola Tesla

Towards primitive societies

Tom Barbalet gives a talk at the Stanford Research Institute which mentions some of my recent additions to the Noble Ape simulation. It would be interesting to try to create a simulation of a primate society which might be similar to that of early hominids. Some of the development needed to enable that is already in place, but there is obviously a lot more which could be done, especially in terms of the social graph and generating narratives for individual apes.

One amusing feature which I hadn't thought of was that after the introduction of tides apes can fall asleep on the beach and then get rudely awaken when the tide washes over them. Because I've been dealing with the command line version of the simulation this sort of scenario is difficult to visualize, but you probably will be able to see it happening in the graphical version.



Probably the main feature which I added was grooming. Having seen apes in zoos and TV documentaries, they seem to spend a large proportion of their time grooming each other. Grooming is interesting because it has a dual purpose. On the one hand it has an obvious adaptive function to remove parasites and keep fur in good condition, but also it has a social function helping to form social status relationships or friendships. With the introduction of parasites and grooming a dominance hierarchy soon emerges, together with some primate politics of who likes, or doesn't like, who. Without any explicit coding, the social status distribution derived from grooming looks very much like the classic distribution of wealth graphs.

Grooming can be thought about in vaguely economic terms. The groomer increases in social status, because they are providing a valuable service, whereas the groomee decreases in social status - they are in effect "spending" some of their status to obtain a service which is likely to lead to increased health and longevity (being free from parasites).



Probably the closest thing to grooming in humans is hair dressing, or maybe massage or manicure. However, unlike in the rest of the primate kingdom I don't think that these activities have a particular association with establishing a dominance hierarchy. Maybe more exotic hair styles do boost your social status, but this seems unclear and I think in the human case dominance is established primarily through use of language (political speeches, manifestos, etc). So at some point in the past our hominid ancestors lost most of their fur and so had to switch the main method for establishing and maintaining a social organization to some other modality - language.

There is also the question of the narrative aspect which Tom brings up. A good test for any AGI system would be for it to be able to answer a question like "what did you do yesterday?". Being able to provide a coherent reply to this implies both a sense of agency - that the system is an agent independent in space and time - and also being able to distill a lot of sensory information into a very compact story which can then be communicated. This story creating aspect seems to be at the heart of human-like language based intelligence, and it's something which various AI luminaries have discussed over the years. We all create stories about ourselves and others, and construct histories in story format.

Tax anything that moves

There's some flim-flam going on on the Less Wrong forum in relation to the question of whether humans will be essentially deprecated in a future highly mechanised society. For a precursor of the sort of scenario under consideration it might be a good idea to examine the so-called "petro-states". These are countries such as Saudi Arabia, Venezuela, Russia and maybe Libya and Iraq, where a large proportion of the wealth is coming from the sale of oil to other countries.

Although oil production does require human labour, the number of people who need to be directly employed in its extraction and refinement is rather small. Curiously though this doesn't necessarily mean that the rest of the economy falls into a Fordian collapse. Without being an expert on the economies of those countries, probably the oil companies are significantly taxed and there is a very large sprawling civil service, with many people being either directly employed by the government, or indirectly providing goods and services to government employees.

If you replace the oil companies with fully automated industries then probably something quite similar would occur. Note that oil is something not directly consumed by the human population (not an "end product"), but rather is of use to machines which then provide utility value to humans, so this seems like a situation analogous to an automated factory producing parts for other automated factories or machinery. So long as there is taxation - one of the few certainties of life - an economic collapse caused by the automation of most jobs seems avoidable. Automated corporations would of course try to avoid being taxed, as far as is possible, but this is not a new phenomena since companies have tried to avoid taxes since the dawn of civilization.

Honey, I shrunk the client

Whilst watching a review of Chrome OS I'm developing a sensation of deja vu.



This looks just like the "thin client" concept from the mid to late 1990s. Under the thin client paradigm you only require a minimal amount of local computing power and storage - enough to run a kernel and a browser - and everything else descends with benevolent grace from heavenly servers on high. It's easy to see why this concept failed to gain any popularity over a decade ago, since although it may have been viable within corporate LANs internet bandwidth was not sufficient to deliver applications remotely to the unwashed masses of computer users in their homes.

So maybe this is merely a question of it taking 10-15 years or so to reach the end of the hype cycle. Certainly if you have a fast and reliable internet connection there's no reason why applications can't be remotely delivered, and the various Google programs - most notably Gmail - have been doing this for some time. It also makes a lot of sense on mobile devices, which may not have a lot of local computing power or where you may want to have the longest possible battery life.

Something like Chrome OS is also the way I envisage telerobotics applications going. Some of the applications for robotics may be quite computationally demanding, and if that can be spread over a number of servers you can still have a lot of functionality delivered to very cheap and low power client systems which can be mass produced in large numbers. The thin client approach also effectively future-proofs the client end, so that continual hardware upgrades are not necessary, even if the software technology and its requirements change quite radically.

The price that you pay for this though is potential loss of local control, high dependence upon wireless networks and loss of security, but these are not necessarily insurmountable issues. There are also potential dangers of lock-in and ransomware from the point of view of the diss-empowered client. A thin client world would also be much more vulnerable to catastrophic failures, since if the server is unavailable for whatever reason (which could include, but is not limited to, "cyber warfare") potentially this effects a large number of users.

The mother of all Mars rovers

Curiosity looks much larger than any of the previous rovers, although the design is very similar with the now familiar six wheel rocker-bogey arrangement. Unlike previous rovers this has no solar panels, and is instead nuclear powered. How long the nuclear powered engine will last for I don't know, but in the case of other space vehicles such as Voyager 2 they can remain operational for decades. A rover receives far more wear and tear than a probe traveling through the nearly frictionless void of space though.

Comment moderation

Due to persistent but low level spamming on every new entry I've enabled comment moderation on this blog. It seems likely that the spambots are now able to both automatically create Blogger accounts and also solve the CAPTCHAs which Google uses, since the spam messages posted always have a distinctive Chinese character format. Possibly this could be human generated spam, via Mechanical Turk or equivalent black hat systems, but the spam comments are so formulaic that this suggests to me that they're machine generated. Whatever spam filters Google currently has on blogs, my guess is that the logic doesn't yet include Chinese text.

Obviously, enabling moderation is regrettable, and really a measure of last resort. In an ideal world anyone would be able to freely post comments without needing to jump over any hurdles, but it now looks as if I'm in a battle with Chinese spam AIs. Fortunately, I'm not expecting many comments anyway, so moderation is feasible even though it generates a little more manual labour.

I'm also running a new blog in parallel with this one, called The Computationalist, running on my own server. It's just a duplicate of this blog, and I'll keep the two synchronized. If after a period of six months or a year it looks as if there are no show-stopping issues with hosting my own blog then I'll switch over to using that full time. Wordpress provides more options for dealing with spam, and it's open source nature means that at least in principle if I was receiving persistent and easily identifiable spam, as I am currently, I could modify the software to deal with it.

Lack of accountability

Some pretty scathing evidence submitted to the Chilcot inquiry into the Iraq war.

"This process of exaggeration was gradual, and proceeded by accretion and editing from document to document, in a way that allowed those participating to convince themselves that they were not engaged in blatant dishonesty. But this process led to highly misleading statements about the UK assessment of the Iraqi threat that were, in their totality, lies."

But this is really the crux of the matter.

"Measures that should be considered include that individual officials should be held personally and legally accountable for their actions in government. There is little such accountability today; instead officials are protected by anonymity, the secrecy cloaking so much of their work, and the legal immunity largely accorded to civil servants, including in conducting actions of such enormous import as sanctions or wars. If officials like me or my colleagues know that one day they might personally be held legally accountable for these actions, it should, one hopes, instill in them a greater sense of responsibility and integrity than that which is sometimes evident in the Iraq case. It goes without saying that the same legal accountability should also be applied to the political masters who make such decisions. The UK accepted this principle in recent negotiations on the powers of the International Criminal Court whose parties, including the UK, have recently agreed that the crime of aggression should be a prosecutable offence, including for heads of state."

It seems unlikely that anyone who took the decisions in government in 2002/3 - principally Tony Blair, but also others - will really be properly held to account for what happened. As far as I'm aware the Chilcot inquiry, like the other Iraq inquiries which preceded it, has no powers to hold individuals accountable.

Another request for comments

Yet another government site has appeared, requesting ideas for reducing spending in order to reduce the magnitude of the gigantic hole in the public finances caused by buying or bailing out banks and quantitative easing. However, the way they're going about this currently doesn't seem very sensible.

Of course there are lots of suggestions, and some of them are good ones, but there is also a lot of noise too. If I were someone in charge of making policy, I'm not sure that I would find many of the suggestions made to be very helpful in practical terms.

I think a smarter way to go about this sort of consultation would be first to compile a detailed list of all departments and their current spending levels, then put this up on a web page and ask people to vote on which items should be reduced or removed entirely, with the goal of getting to a 25% reduction. It would almost be like turning the exercise into a kind of game or puzzle.

The overall aim should be to implement spending reductions which have the highest possible level of public support. In this way hopefully the worst of the unrest can be avoided, and there won't be any Greek-style rioting.

Still, all this consultation is an interesting exercise, and my views about the future of technologically enabled direct democracy may change as a result of it. It's certainly evident that there is a strong regressive element within popular opinion, which would have us revert to a medievalist world of public executions and slavery.

Lacklustre spies

There has been a lot of flapping in the media about Russian spies in America, but from what I can ascertain from these stories those accused don't appear to have done anything very exciting at all. Even if they are spies, if they just sat around in their apartments watching bad daytime TV shows and occasionally mowing their lawns every now and again they're hardly stealing crucial US national secrets. The whole episode seems like a storm in a vodka glass.

Spying as a phenomena seems so utterly pervasive in contemporary society that old-fashioned cold warrior style spying appears as being incredibly quaint. There's no need for attractive red-headed Russian master spies to seduce their targets and obtain information if they can simply log on to Twitter and read the latest geo-located status updates.

The only thing which was perhaps mildly interesting is that one story alleged that one of the spies had been caught using steganographic software. In the current age I don't suppose that ought to come as any massive shock, since it's relatively easy to hide information within images, but it at least adds some algorithmic variety to an otherwise unremarkable story.

Blocky bots

A kind of modular toy robot designed by Lars Marcus Vedeler and Ola Vågsholm. This could be the new Lego, and would be quite fun for kids to play with. Probably the modules could be made fairly cheaply, since they look like just a potentiometer, battery, servo and Arduino.

Olars from Lars Marcus Vedeler on Vimeo.

Fetching beer: the holy grail of robotics

Expensive web sites?

Why do government web sites cost so much? There will be a certain amount of cost in buying multiple servers, making sure that they're backed up, checking that the security is good and so on, but even so the prices do seem high.

One thing which the journalism, either from the BBC or others who have commented on this story, hasn't provided is some context so that the reader can make a reasonably informed judgement about cost effectiveness of these apparently wasteful government web sites. For example, what is the cost difference between HMRC running its services on a web site and doing things in the older way by sending out forms and processing the returned forms? My guess would be that even though the web site is expensive - perhaps excessively - it's probably an order of magnitude less so than doing the same thing via snail mail with all the stamps, storage issues and human labour that entails. When you look at the cost per visit, in most cases it's significantly less than the cost of a postage stamp.

So unless the government plans to discontinue the services of these departments entirely, getting rid of the web sites "because they're too expensive" could turn out to be quite a false economy.

Probably the government could get better value for money on its web sites, especially by using web servers running open source software and wherever possible using existing open source content management systems for wikis, sites, forums and blogs. It would also be a good idea to turn some of these web based government systems into open source projects, which serves the dual purpose both of providing transparency in the way that public information is stored and processed, and also helping to keep development costs under control. Via the open source methodology best practices in provision of online services could be spread around the world.

All they really need to do is knuckle down and implement the Open Source Action Plan, except that the information about it seems to have been removed - not a good sign.

Climategate scientists victorious

So the climategate scientists are vindicated on nearly all counts, with only minor criticisms. Despite the scandal, I think that the whole climategate affair had quite a positive impact. It made the public scrutinise the scientific process as currently practised, which is not always the same as the way that science is supposed to be done in principle, and ask probing questions. It also provided more appreciation of the uncertainties involved, since often the popular media portrays these things in very black and white terms. For example:

"On one occasion, when presenting a graph combining tree-ring and instrumental data to the World Meteorological Organization, it should have made clearer the way in which the data was combined."

I've often seen similar effects in papers on computer vision, where there can be missing information or unstated assumptions, which are revealed when you try to reproduce the experiment. Sometimes this might be deliberate, in order to give the impression that the research was actually more successful in obtaining meaningful conclusions than it really was, but a lot of this is no doubt just down to good old-fashioned human fallibility or confirmation bias.

Perhaps most importantly, climategate made me take a look at the data myself - something which otherwise I would have had no interest in doing.

"We demonstrated that any independent researcher can download station data directly from primary sources and undertake their own temperature trend analysis".

Writing computer code to process the data "took less than two days and produced results similar to other independent analyses. No information from CRU was needed to do this".

This is actually quite a fair summary of what I did, which eventually resulted in the tempgraph utility. However, one thing I did find, which perhaps differs from the graphs which appeared on the Met Office web site is that there is no "hockey stick" temperature spike in the data, although the overall trend is upwards. There is also such a wide range of variation in the data that trying to draw meaningful conclusions over short time scales, such as one or two decades, is likely to be quite misleading. On short time scales it's easy to "see faces in clouds". Also, a century of temperature recordings exist prior to the late 19th century and these values are typically not shown in popular media depictions of global warming (perhaps with some justification due to the introduction of the Stevenson screen). The earlier data suggests that the warming is part of a long term trend which existed prior to the late 19th century.

The Guardian summarizes quite well, as follows.

"Climategate reveals the urgent demand by a new breed of citizen-scientist for access to the raw data scientists use to do their work. Simply accepting a scientist's assurance that data are accurate and reliable is no longer enough. Scientists will have to make their data available for independent audit."

Parenting

A mother partridge and her chicks, roaming in the vicinity. The area that I'm currently living in is close to the North Yorkshire moors, and it's almost like living in a wildlife reserve.



Here they are dust bathing in a sandy car park area. The mother bathes first, and then the chicks imitate her. Dust bathing is a behavior which helps to remove parasites.

Imagination

Some discussion about the nature of imagination.



As far as I know, how this occurs in the brain is still not well understood. Possibly the thalamus may be important in switching different components of a concept in or out, so that novel combinations can be generated. Also from the above thought experiment it's apparent that language is strongly implicated in the way that this switching occurs. It could be that language provides an efficient index on all the non-linguistic concepts.

One thing which does seem to be the case from the existing evidence is that the neurons involved in imagining a particular kind of object are the same as the neurons which would be active when you're observing that object for real.

Downsizing

With the public sector looking to downsize by 40%, and VAT going up to 20% (inflating the cost of nearly all goods and services) it's looking as if the honeymoon period enjoyed by any new administration could come to a fairly swift end, and there may be a winter of discontent ahead.

Reducing the size of the state will have some benefits, since the Labour administration undoubtedly spend money in a wasteful way on unnecessary wars, ID cards, the millenium dome and so on. However, it should also be borne in mind that the state is the UK's biggest employer, so cutting out 40% could result in a significant rise in unemployment. Many of those people made redundant will probably be archetypal middle class workers - both well educated and politically connected - and looking back at previous episodes in history it's usually when the middle classes are delved into poverty that the peasants really begin revolting. Traditionally, the rebels, revolutionaries and rabble-rousers usually come from wealthy middle class backgrounds where there is time and resources to cultivate abstract ideals. Another factor adding fuel to the fire is that probably many of the redundant bureaucrats will own too many assets or have too much savings to qualify for the means tested state benefits, so they will have everything to lose in terms of living standards.

Also it appears that the government is trying to make sure that when mass redundancies occur they're on the worst possible terms, from the perspective of the person being made redundant.

So a near term prediction is that within 1-2 years there will be civil unrest comparable to the miners strikes or poll tax riots of the 1980s. If the rebel alliance are from sufficiently impeccable civil service backgrounds and are able to coherently articulate the popular mood of the crowd then possibly they might bring down the coalition government, or cause enough internal divisions for it to disband.

Bizarre citizenship ceremony

To become a British citizen it would seem that you have to swear allegiance to:

a) A fictitious deity
b) An outdated feudal system (the local warlord)



I suppose this highlights the absurd arbitrariness of tribalism/nationalism, and to some extent the role of ceremony in delineating the operational parameters of the tribe.

Electoral reform

So provided that the coalition government doesn't collapse from internal divisions within a year, as some expect it to, there will be a vote on a new kind of electoral system. The proposed alternative voting system however does not appear to be very alternative. This seems like a tiny revision to the existing system, which makes it marginally more complicated to calculate who won.

So if the new administration really was seeking "radical reform", what would an alternative voting system look like? Probably for the present the existing minimalistic paper and pencil method of voting should be retained, until such time as a secure electronic equivalent can be devised which minimises the possibility of fraud. Whatever the future mechanics of voting, the system needs to acknowledged to be a fair one, not just by experts but also by the general public too, and this probably means that whatever electronic system is eventually decided upon needs to be comprised of open source software and hardware, which anyone can independently inspect for possible flaws - negating the rise of any conspiracy theories.

In order to maintain some quality control it might still be a good idea to vote for a candidate - that is, an individual administrator appointed to oversee the implementation of policy. For the present this would remain as a person, but in the future could be replaced by some administrative system, which may or may not contain a human component. In addition to a list of candidates there would also be a list of policies which could be voted for by the electorate. The policies would be generated by the candidates, which hopefully avoids anything too silly from making its way into the list. The fundamental idea here would be that you choose what sort of policies you would like to be implemented, and which administrator you trust to implement them. For obvious reasons the list of policies would need to be short, so that it doesn't become too unmanageable, and candidates would still be free to devise additional policies as they saw appropriate over the course of their term of office.

In a way this proposed alternative voting system is quite similar to the way things happen now, in that candidates do produce manifestos. However, currently there is no direct link between voting and policy, and candidates are free to completely ignore their earlier manifestos subsequent to being safely installed in a position of authority. Including policies within the voting system would also make the political debate more focussed upon ideas and plans, as it should be, rather than on personalities and parties, as is currently the case.

Your Freedom

A new web site, called "Your Freedom", asks which laws or regulations should be ditched. The Digital Economy Act is an obvious candidate, as is the Dangerous Dogs Act - a piece of legislation often cited as an example of a bad law passed in haste during a media frenzy. Section 44 of the Terrorism Act 2000 also is too vague and has been used to harass or detain both amateur and professional photographers (even train spotters) legally taking pictures in public places.

In principle I'm in favour of this kind of use of web technology to facilitate greater participation of people in collective decision making. The current state of the art seems quite poor though, and there is plenty of scope for improvement. For instance, on the Your Freedom site there are numerous duplicates of the same request, phrased slightly differently, and ideally there should be some way in which these can be flagged as duplicates by the users of the site, which could then be automatically aggregated.

Also, the Your Freedom site perhaps provides an example of why technologically enabled direct democracy might not be such a good idea. There are plenty of silly entries, such as removing all speed limits on roads and entries like:

"Force all extra-terrestrials to carry valid passports from their homeworld and make it law that they have to stop at the International Space Station to gain access to earth."

and...

"Ban Simon Cowell"

and this one made me laugh

"the NHS should provide free sex for over 50's. Its great exercise and will help many people. Why should the footballers and politicians have all the fun"

Also there are entries which although they perhaps don't fall into the silly category are still rather disquieting, such as numerous calls to re-introduce capital punishment and calls to abandon human rights laws. If you disregard human rights for prisoners - no matter how despicable their crimes may have been - then recent cases like Abu Ghraib and Guantanamo Bay show what terrible abuses can ensue.

The news that at least some of the laws passed in recent years might be repealed is an encouraging sign that things may improve in future, and that the tide may be turning against the creeping fascism which characterised the Blair and Brown era.