Pubs embrace ID cards

According to this BBC article it looks like pubs will be amongst the first to embrace use of ID cards. Notice that the responses of the public all assume that this is essentially a benevolent system working on their behalf. However, suppose that one of the bouncers or the pub landlord is a crook. They could steal a lot of ID information, which can then be sold to fraudsters on the black market. The information stored on an ID card or a drivers licence is sufficient to apply for a copy of your birth certificate, or to take out loans or open bank accounts in your name.

It's claimed that the ID scanners are "secure", but as we all know security is a process, not a technology. Once you've handed over your ID details where are they stored, how long are they stored for (potentially indefinitely), who has access to the data and what other systems are data mining it (for example, insurance companies or government agencies). If you're unlucky enough to find yourself out of work during the current recession officials at the social security office might be very interested indeed to know how much you're spending on booze, and reduce the benefits for claimants deemed to be "binge drinkers" (which in the UK is just about everybody who walks into a pub on a Friday/Saturday night).

Government groks open source

Traditionally the UK government was one of the worst offenders when it came to proprietary software, so it makes a refreshing change to see that they're now taking open source more seriously with the publication of an action plan. The plan looks pretty good, and for the most part makes all the right noises.

"The OGC Standard Contract Clauses will contain a clause to ensure that the government secures full rights to bespoke software code or customisations of commercial off the shelf products it procures, and that it is clear that these rights cover re-use anywhere else in the public sector and the ability to release the code on an open-source basis. Where appropriate, general purpose software developed by or for government will be released on an open source basis."

Purists will be indignant about the consideration of Microsoft's dubious OOXML format alongside the more standard ODF, but apart from that the action plan looks like a good one.

However, there are powerful vested interests who I'm sure will absolutely rail against this, and I fully expect lobbyists from Microsoft and a few other big software companies to do everything within their power to ensure that the proprietary grip on the public purse remains vice-like. The bleak economic situation and consequent need to reduce costs should provide adequate incentive to begin implementing the plan though. Still, I expect a new salvo of proprietary FUD to be launched, because even moderately clueless IT folk (aka the "working dead") are now beginning to see through the misinformation of the past ("Linux is a cancer", etc), and that there might just be the tiny chink of a realization that there is a life beyond Microsoft Windows.

Incidentally, greater use open source does not appear to be a party political issue, since the conservatives have also been making similar noises. Really this should be a no brainer, but as we all know affairs of state move very slowly and with titanic inertia.

Did a 1981 movie predict the current economic slump?

Some clips on YouTube of a 1981 movie called "Rollover" almost seem as if they could be contemporary or very near future news articles. If some of the economic masterminds in the government get their way and Gordon Brown begins fraudulently printing money like the proverbial bat out of hell (described in official language as "quantitative easing") something like the scenes from this movie could actually occur, as people's life savings rapidly become worthless.

Moving windmills

Formalizing the debate

DebateGraph is an interesting site which allows debates on different topics to be formalized and viewed. I like the concept of this, which does bear some relation to the Engelbart vision of how networked computers would be used, although I think the actual execution in terms of usability needs more work (the explorer view doesn't seem to work on Linux, although other views are ok). An xml import/export function is also needed, so that debates could be visualised, analysed or refactored using different software.

Cult of personality

Trailer for a film which examines the cult of personality surrounding the inventor and singularitarian Ray Kurzweil.



I've voiced this opinion many times before, but I think taking large numbers of pills per day in an effort to stave off the grim reaper demonstrates a degree of faith in the pharmaceutical business which industry insiders, when talking candidly off the record, absolutely don't share.

Strangely the trailer features the discredited politician Colin Powell and Bill Shatner, who I think is in some vague way connected to the transhumanist movement but is probably better known for his unique Beatles music covers.

Empire of stone

General Intelligence

A fascinating story about how children can learn things without any explicit tuition. I've read about this "hole in the wall" story before, but this gives more details about the experiments and what happened.

That people can learn things on their own is not news, but I think this shows particularly plainly the power of human general intelligence, to be able to learn new things without any prior knowledge. It's claimed in the video that the children were basically illiterate, with no knowledge of English language, and that their parents were also illiterate, and yet after a relatively short period of time they were able to do some non trivial amount of learning.

The other interesting aspect is the manner in which the children learned, which was fundamentally through social interaction. In psychological studies it's usually assumed that intelligence is a quality which exists vaguely somewhere inside the head of an individual knower, but the hole in the wall example shows how intelligence can be much more of a network phenomena, with accurate knowledge emerging in a self-organized way within a group of individuals all of whom possess varying degrees of imperfect knowledge and conflicting hypotheses.

Of course this doesn't mean that formal teaching is worthless. Probably the children would have been able to learn more efficiently if guided systematically by a domain expert (i.e. a teacher). But guided learning also has its drawbacks, and is still fundamentally about social interaction between teacher and students (the teacher becomes the top of the social heirachy, rather than a small group of children closest to the computer screen). If the social dynamic doesn't operate smoothly, the children may still learn nothing, such that the teacher can become a barrier rather than a facilitator of learning.

Is the future a concrete ghetto?

Years ago I often visited a town in Scotland called Irvine. The Wikipedia entry for Irvine looks quite enthralling, but the parts I frequented were much less glamorous, consisting of large expanses of grey concrete buildings which were poorly maintained with cracks and big dark stains running down their sides and window corners. These were public housing projects which by the austere appearance I guessed must have been constructed in the 1960s. Driving smugly around in my company people carrier (yes, we drove big fuel guzzling cars back in those days!) this architecture - grey square concrete as far as the eye could see - seemed relentlessly depressing, and I really felt sorry for the locals who had to live there. The thought occured to me that "this is what happens when someone with zero imagination designs houses". I very much doubt that any private individual would intentionally design housing this unattractive. To do something like this takes governmental intervention.

So it's somewhat disappointing to see Tim Tyler predicting basically that big government is the future.

My expectation is that capitalism will be discarded as a stupid, inefficient and primitive system - and future metabolic activities, will be managed and run directly by the government...

My expectation is that the state will obliterate individual liberty - and that any autonomous individuals will become slaves of the state - much like worker ants are slaves to the hive mind."

Perhaps future governments will be radically smarter than all current or previous ones. Maybe the activities of government, which are largely about supervising the deployment of resources, will be entirely automated by "machines of loving grace". If this turns out to be the case then we may justifiably look forward to a benevolent and ultra efficient world government, but if future governments turn out to be anything like the ones which we've endured throughout all previous human history then this is indeed a dark prediction.

The basic problem is that human intelligence - as it currently exists - just doesn't scale well. Large organizations or large mobs tend to be far less smart than their smartest individual constituents, and you can visualise this like a piece of unannealed glass with its molecules pointing in random directions. Any organization consisting of multiple individuals who are not all of the same mind may work against each other, in extreme cases resulting in a disorganized muddle or political paralysis. Those organizations which are successful tend to contain greater homogeneity - a common purpose - but homogeneity itself comes at a staltifying cost, driving out innovation and risk taking, and limiting the capability for bottom up self organization in response to envronmental change. The ultimate homogeneity is of course straightforward tyranny, in which one super-powerful individual basically runs everything. As history shows, under such conditions the opportunities for abuse of power are truly frightening.

Fortunately, I disagree with Tyler that the future is a kind of dark grey or greyish neo-communism on a worldwide scale. If I genuinely believed that the future was a statist dystopia in which individual liberty is obliterated then I'd give up on technology and take up painting or gardening or some other artistic pursuit. After all, what's the point of building new technology if it's only going to be used to disempower people? By contrast I'm much more inclined to believe that the future will turn out to be, in the words of James Hughes, "radically democratic".

Blimping heck

A remotely controlled blimp, with a wireless blackfin camera onboard. The dog does not look impressed.

Trossen Robotics apparently fooled by spoof website

Read it and weep.

Still, I think this does demonstrate that the credibility gap between science fiction and contemporary reality is narrowing. Real research and commercial uses of robots, especially in the military domain are catching up with the cinematic fantasy of the 1980s, to the extent that a naive observer might really conclude that this was a genuine research website. For example, a recent news report talks about "drone attacks".

Even if the name isn't an immediate giveaway in one of the videos the roboticists appear to be wearing white lab coats. No robotics people that I've ever known wore white lab coats, except perhaps as a joke (i.e. pretending to be "Les" from Vic Reeves Big Night Out).

The Mr Bean school of banking

It looks like once again there is talk of the banks fraudulently printing money (i.e. money not backed by any sort of assets). The new euphemism for speeding up the printing presses is "quantitative easing". I've also heard it described in recent weeks as "expanding the money supply". Easing and expanding sound like nice words - the opposite of tightening and crunching.

However, if the currency becomes devalued and people's hard earned savings are rendered worthless through the prosecution of irresponsible financial policies I think there's going to be a quantitative demand for Gordon Brown to be eased out of office.

Isaac Asimov talks about internet search

Asimov doesn't directly talk about Google or Wikipedia, since it would be impossible to predict the future in such detail. But like the bow waves spreading out in front of a ship the future does have some detectable preceding shape before it really arrives.

Ethics and roboticists

This talk by Illah Nourbakhsh on the topic of ethics in robotics is more articulate than similar ones which I've seen in the past, such as the robotics discussion at the 2008 Singularity Summit or the AGI conference. I agree with his point of view, in that engineers have to accept responsibility for the consequences of what they do. A certain proportion of my own dealings with industrial robotics has always been about ensuring that machinery doesn't end up causing harm to people, and that systems are "fail safe".



I've never been in a comparable situation where I've been explicitly asked to work on weapons targeting, but I have been in situations in the distant past where for example the management of a company believed that they could get away with far lower standards of safety simply because they were selling equipment to factories in poor countries. I raised concerns about this, and eventually the machinery was re-engineered to the same safety standards as are mandated in Europe. I've also been contacted in recent years by individuals who according to my point of view are on the dark side of robotics, but I've always flatly refused to get involved with applications which are fundamentally unethical and likely to bring the field into disrepute.

One thing which I've noticed consistently is that automaticity is often used as an excuse to evade personal responsibility or behave badly. "It was nobody's fault, it's just the system" is a phrase often used to deny people services or to treat them without respect. You can see these very same forces at work today with the decision that nobody is to blame in the Charles de Menezes case. I expect that the Menezes case - a very straightforward human rights violation - will rumble on for years until someone eventually accepts responsibility, because in the end systems are ultimately constructed from or at least supervised by people so there is always someone who takes the final decision and has to take the consequences if their actions have fatal consequences.

Some component of becoming an ethical roboticist - or an ethical anything - has to do with letting go of tribalism. It's difficult, because tribalist instincts seem to be a fundamental aspect of our psychology and have obvious adaptive value in an evolutionary sense. When you let your thinking be dominated by the desire to protect those with whom you are most familiar, to the exclusion of all else, then that's when things can begin to go wrong. Especially in hard times there's always a tendancy to surcome to this mode of thought, and you can imagine it as the psychological equivalent of the body's reaction to a cold climate, withdrawing blood flow from the extremities to maintain the inner core temperature.

Balancing act

Another balancing humanoid. This all depends upon use of accelerometers, and how quickly the sensed orientation can be translated into corrective movements. Experimenting with robots like this gives you a good appreciation of just how fast biological creatures can react.

As if we aren't being watched enough...

Here comes mobile CCTV, demonstrating - if further proof were needed - the ongoing British love affair with surveillance of the public. I think this obsession with surveillance is really all about maintaining the social order, and has its origins in the British social class strata, where people were supposed to "know their place" and the upper classes went to considerable lengths to avoid mingling with the plebiscites whose behavior was conspicuously more hedonistic. Vestiges of the class system still linger on the railways, in air travel and also in the postal system which all have explicit "first class" and "second class" services.

Weisenbaum

"The truly fascinating thing about a computer is its universality. The programmer has a kind of power over a stage incomparably larger than that of a theatre director.

You can construct a system in which the natural and physical laws are turned on their head. You're a director on a stage of unlimited possibilities, and that's enormously fascinating, and of course it leads to the dream I sometimes refer to as 'artificial intelligencia' - the dream of us being able to do everything, including being able to create an artificial human being"

A film about the life and opinions of Joseph Weisenbaum is now available to watch online. Weisenbaum was most famous for having written the Eliza chatbot in the mid 1960s - a program which tried to converse with the user in the style of a psychoanalist. One of the first computer programs which I ever ran, typed in from a listing in a book when I was about ten or eleven years old, was called "the clever computer" and was very similar to the original Eliza.

Weisenbaum later became known as a critic of AI, advocating the notion that the role of computers within society should remain limited. At first I had thought that he was probably a kind of luddite, but after reading Computer power and human reason this doesn't appear to be the case. In that book, written in the 1970s, some of his criticisms of the use of computer technology seem very dated because that was largely before the era of personal computers and the internet as we know it now. Weisenbaum's main worry seems to have arisen not from ignorance about computers, or even any elaborately constructed denial of their future capabilities (in the style of Penrose), but from a realization that computers could be used as fundamentally conservative force to stifle social change or drive society in directions incompatible with human dignity.

The Singularity University: where's the beef?

The splash of publicity about a new Singularity University sounded quite exciting. One thing that I have noticed over the years is that there appear to be key areas in AI and robotics which just don't get much research attention - presumably either because they have no obvious short term application in industry, or because they're just difficult and unlikely to fit neatly into a two or three year PhD thesis. Amongst these problems I would include general purpose visual object recognition, commonsense reasoning and detailed reverse engineering of brain structure. There is of course a certain amount of research going on into uncovering the detailed structure of the brain, but from what I've read and people I've chatted with all of this seems to come under the remit of medical funding, and apparently the organizations responsible for that are extremely averse to any talk of "mind uploading", because that's not strictly a medical application and might raise undesirable controversy.

So, if there is to be a Singularity University you would expect them to be doing this kind of research, boldly going where other universities can't or won't, right?

Wrong!

The description of what the Singularity University will be doing, as described in their FAQ, looks very disappointing and I tried in vain to locate information on any particular research which they'll be doing.

"The GSP and Executive Programs will not conduct traditional research."

Ok. So what about non-traditional research?

"the GSP program will organize detailed student Team Projects that will allow the student body and faculty to look at how to use exponentially growing technologies to solve some of the world’s grand challenges. One could imagine, for example, that issues such as global poverty, hunger, and climate crisis could be studied from an interdisciplinary standpoint where the power of artificial intelligence, nanotechnology, genomics, and other key technologies are brought to bear in a cooperative fashion to seek solutions."

Look at how to use...in a cooperative fashion? What does this mean? This just sounds wooly and vague, as if they have no real idea what they will be doing. The rest of the site seems to be populated by similarly non-specific and hand wavy evasiveness.

So, based upon the content of their FAQ I'll predict that the Singularity University will fail as an organization, due to lack of focus (i.e. the lack of any coherent supergoal). Simply making graduate students or business executives aware that the pace of technology is accelerating with a few graphs probably isn't going to impress many people, especially if they already have a background in technology. Also, if social networking with prominent singularitarians is your penchant there are probably better and cheaper ways to do this now than via attending a university.

Last but not least there appears to be a disproportional focus on Ray Kurzweil himself. As probably the most well known singularitarian this is to be expected, but beware of cults of personality based around one person. When things are skewed around the opinions of one man this could bias research in very unfavourable directions. People trying to bring about a technological change should focus on the science and the task at hand, and avoid any taint of cult worship.

Pay what you want

Interestingly it looks like hard times have driven a restaurant in London to basically adopt the same business model as Magnatune. The results appear to be similar too.

"Customers have already paid 20 percent more than the original price"

This depends upon the idea that most people aren't fools, and realize that if they pay less than the going rate the restaurant will soon go out of business and they'll have to find somewhere else to go for lunch - perhaps somewhere less convenient and more expensive. It will be interesting to see whether this catches on in other restaurants. In that industry there is already a precedent for paying what you want to some extent, in terms of tips.

They think it's all over...

Robotic soccer does seem to have progressed quite significantly over the last decade. I remember a time not so long ago when about the best which a robot soccer player could manage was to slowly move towards the ball and push it up the field vaguely towards the goal, with very little in the way of team work (you might think some human players don't do much more than this either!).

The speed and agility of these robots is quite impressive, and it's hard to deny that there appears to be some kind of team tactics going on here.

Warehoused

Another warehouse robot video showing the Kiva robots in operation. One possible disadvantage of this kind of strategy is that if one robot fails it could cause a big jam, or stop the system entirely. Consequently each robot has to be engineered to a very high level of reliability. That's probably not hard to do - you just use better quality components with a longer mean time to failure - but might have cost implications.

Still, this system seems to work if the video is to be taken at face value, and there is a substantial market for this kind of system if it can perform more cost effectively than traditional warehouse methods. Promotional videos should always be viewed skeptically though, since sales people will often try to sell inventions which don't exist or "sex up" their true capabilities.



Robotic forklifts are another "killer app" in industrial robotics which has yet to be fully achieved. There have been attempts to automate forklift driving for at least a decade, but so far I know of no system which can compete with a human driver in terms of speed and accuracy. A recent attempt at automating forklift driving is described here. The absurdly contrived military tag should be ignored - this is overwhelmingly a civilian application.

For you Tommy ze vor is over

Here's a talk at a conference which is some years old, but which discusses issues which are still relevant and likely to become more-so. In my view although I agree with many of the points made this is a rather one sided view of the relentless march towards what I've sometimes called in the past "digital fascism". What's not discussed is that assuming that the surveillance war is over and Honecker won (from beyond the grave) one way to push back against fascism is to try to move towards equiveillance - surveilling the surveillers and the aparatus of state itself, and exposing anything going on which could be considered unethical or against the public interest.

As a practical example, there seems to be no log of exactly where all the street surveillance cameras are for all cities in the UK. There also seems to be no log for the expanding installations of "shouting" CCTV. Such a map could easily be created through collective social networking effort using web 2.0 tools such as Google maps.

Electric vehicles in Bradford, 1904