Tuesday, 21 May 2013

Asleep at the Wheel

As the reality of capital-biased technological change (more robots, fewer jobs) begins to bite, the supporters of social democracy become ever more desperate to find a silver lining. Will Hutton in The Observer tells us that "Suddenly a robotised, automated economic reality is moving off the science fiction pages and into daily life". Of course, there's nothing sudden about this. His article is illustrated by a contemporary photo of robots constructing Tesla sport cars (below), which immediately recalled for me the famous Fiat Strada "Hand built by Robots" advert from 1979.

There are three features of this picture that I think are worth noting. The first is that the robots are painted blood red, as if in an attempt to indicate some residual corporeality: once were workers. The Fiat opus celebrated the absence of the human in the manufacturing process, which was satirised by the Not The Nine O'clock News's "Hand built by Roberts" sketch. The second thing that strikes me is that you cannot immediately tell what is being manufactured here. It could be anything. The Strada advert understandably foregrounded the car, but in so doing suggested that robots were a heavy industry application and no threat to whitecollar jobs. The third feature is the clean room environment, enhanced by the use of white paint and grey metal, which contrasts with the Fiat advert's classically dark and grimy industrial setting. There is a practical aspect to this, inasmuch as keeping dust and dirt out is good for equipment maintenance, but there is also an ideological resonance in the pretension to a sterile lab.

The one thing the Strada advert couldn't do was excise the test drivers of the newly built cars. The measure of the distance we've come since 1979 is that the Google driverless car is already a proven technology, though not yet commercialised (unless they're charging a fee for media appearances). You wonder why? Hutton talks up the benefits of this technology (though he omits drunk-driving) but fails to put it into a social context. Automation costs money, and there is no offsetting saving on you, the erstwhile driver, so most people will initially forgo the benefits for a cheaper, non-automated model. Many of the collective benefits, such as increased road capacity and lower (possibly non-existent) driver insurance, will only arrive when automated cars are mandatory and the non-automated outlawed. It should be obvious that automated cars can only succeed as a result of government diktat, which means they'll probably happen sooner in Europe and Japan than the US (anti-gun regulation will be augmented by anti-car regulation as a leading libertarian identifier - just think of the bumper stickers: "Defend your right to crash!").

Hutton's social naivety is also clear in this statement: "Some argue that a dystopian world is emerging in which good jobs and full-time employment will become the preserve of an educated, computer-literate elite". Doh! Despite the consistent trumpeting of the benefits of education (skill-biased technological change), there will be no correlation between "good jobs" and computer literacy. While there will obviously be a class of highly-skilled and well-paid techies (atop a pyramid of digital peons), most "good jobs" in the future will be rent-extracting and their occupants (increasingly hereditary owners) will continue to pride themselves on their lack of IT skills (a corporate lawyer using a Blackberry does not constitute "skill"), just as the bourgeois of the first industrial revolution prided themselves on their uncalloused hands. As a social democrat, Hutton believes there is still hope that the future economy may produce large numbers of well-paid, skilled jobs. He identifies four areas.

"The first is in micro-production. There is going to be a huge growth in micro-brewers, micro-bakers, micro-film-makers, micro-energy producers, micro-tailors, micro-software houses and so on who will deploy the internet and micro-production techniques to produce goods at prices as if they were mass-produced, but customised for individual tastes". If the small number of goods we produce are at mass-production prices, then our turnover and profit will be tiny. We could make a living through customisation, but that means charging a premium price. These are middle-class dilettante businesses that assume a lot of well-off buyers - i.e. our fellow artisans supporting their own lifestyles through a trust fund, share portfolio or property rents. The future looks like Mumford & Sons.

"The second is in human wellbeing. There will be vast growth in advising, coaching, caring, mentoring, doctoring, nursing, teaching and generally enhancing capabilities". While there is certainly a market for life-coaching, assisted navel-gazing and other indulgences, mass-market "human wellbeing" services are already being automated, where they aren't being de-professionalised, such as in teaching and nursing. The reality of Hutton's vision for most people will be an increase in low-paid jobs providing care for the elderly. The future for many smells of wee and boiled cabbage.

"The third is in addressing the globe's 'wicked issues'. There will be new forms of nutrition and carbon-efficient energy, along with economising with water, to meet the demands of a world population of 9 billion in 2050". These aren't new industries, merely the evolution of existing ones. For "nutrition" read agriculture and food processing. If you think this is about to produce a huge increase in employment, then you haven't been paying attention. At this point Hutton starts to spin giddily out control: "Space exploration will become crucial to find new minerals and energy sources. New forms of mining will allow exploration of the Earth's crust. The oceans will be farmed". Sweet, suffering asteroids - we're going to exploit space! I presume by farming the oceans he means harvesting those creatures, you know, fish.

"And fourthly, digital and big data management will foster whole new industries – personalised journalism, social media, cyber-security, information selection, software, computer science and digital clutter removal". This is techno-nonsense that could do with some serious "clutter removal". Personalised journalism is just a euphemism for targeted advertising. It doesn't mean you get Jon Pilger or Peter Hitchens to give you their undivided attention for a day. It will not produce any more jobs, beyond the unpaid artisan-microblogger. While it is true that techno-bollix like big data will be used to create supernumerary, rent-extracting roles, these will just displace similar roles based on older bollix, like PR. Parkinson's law operates at the aggregate as well as the individual level.

I can't be too harsh on Will Hutton. Though he is trying to keep his spirits up with the power of positive thinking, he does at least recognise the reality of the transformation we're going through, the current absence of any "double movement" to protect society from its effects, and the ideological role of neoliberalism in ruling out the possibility of resistance.

He concludes: "Britain will need the open innovation structures, financing mechanisms and social support institutions to capitalise on the opportunities quickly, rather than be overwhelmed by the risks". The problem is that these very mechanisms and institutions are currently being reengineered to preserve the interests of the few in the face of the incoming tide. Translated into plain English, education ("open innovation structures") is being forced to readopt its traditional role as a class filter rather than a talent amplifier; The City ("financing mechanisms") is determined to restore the ancien regime under which the interests of society are subservient to the interests of Money Capital; and the welfare state ("social support institutions") is being picked apart and the profitable bits handed over to Big Capital.

The historical irrelevance of social democracy stems from its belief in the necessity and desirability of full employment, and the associated use of universal income tax to redistribute wealth via a welfare state. It can cope with temporary unemployment, and may even benefit from a small, persistent underclass as a goad to keep the mass of workers "respectable", but it has no answer to a society in which a growing minority are simply surplus to requirements. Inventing fantastic jobs, from micro-tailor to ocean-farmer, simply avoids confronting this.


  1. It's good to hear that one of the four pillars of our economic future is effectively the arts and crafts movement.

    I can excuse John Ruskin for not being able to see that artisan produced goods will always be for an economic elite, but it seems absurd for an economist to not see it!

  2. It's a pretty gloomy picture you paint Dave

    But "The future looks like Mumford & Sons." Good god

  3. Hi,
    I think you seriously overestimate the cost of making a modern car driverless. >90% of the hardware is already in midrange cars already on the road and in many cases has been for years. Diesel Transits lost any mechanical connection between the throttle pedal and the engine in the early 90's IIRC and petrol engines have all been electronic throttle for the last 5 years or so at least –(probably with the shift to Euro 5 emissions but I’d need to check). Similarly, ABS is now a legal requirement in Europe and Electric power steering is also very common now (for efficiency reasons). All this means that the hardware is already in nearly all the vehicles currently being sold to allow a computer complete control over acceleration, deceleration and direction of travel.

    The sensors required for driverless cars aren’t quite as universal, but aren’t far behind. Parking sensors are common and even a Ford Focus can now come with a road sign recognition camera. Radar controlled cruise control has been around for best part of a decade as well and is steadily making it into cheaper cars.

    So from a hardware point of view the cost delta of a driverless car versus a current production one is negligible.

    What is missing is the software to do it (a huge amount of effort goes into making sure the systems mentioned above don’t take on a mind of their own, so an even greater amount of work will need to be done to make sure the systems are still safe when they do) and the legal framework to allow it to happen. As an example the self parking systems now on the market (even in B-class cars) were held up for a couple of years while the car companies and the lawyers worked out who was liable if they went wrong, although this may set a precedent for future autonomous driving.

    Despite fully driverless driving being perfectly possible in the fairly near future, the manufacturers are still not planning on allowing you to fall asleep at the wheel or drive drunk anytime soon due to legal problems, (see above) – you will still be responsible for the driving, even if the car is doing all the work.

  4. Driverless cars require a lot more than "control over acceleration, deceleration and direction of travel". The ROI for autonomous vehicles is not fewer road deaths but increased carrying capacity. This means that its about central control of traffic - i.e. more smart roads rather than smart cars. This isn't cheap, which is why I can't agree that we're over 90% of the way there and that the last step will be negligible in cost. The last big hurdle isn't software so much as telemetry.

  5. This comment has been removed by a blog administrator.