SBA

Luddites Rool OK

In some circles Luddite has become a term of disparagement, someone unable to cope with new technology. The reality of course is that the Luddites were skilled workers who saw that new technology had the potential to de-skill and devalue their jobs, potentially making them redundant - a phenomenon we now call Technological Unemployment. It is undoubtedly true that technological innovation steals jobs, and it is a moral problem I have wrestled with most of my working life - my early years in computing were largely focused on developing technologies to make people redundant, and the theme has continued as an integral if intermittent part of my work ever since. Far from disparaging the Luddites I respect them, they were shrewd enough to see the future.

From early on in the economic debate about Technological Unemployment there have been those who have argued that the effects of Technological Unemployment are temporary and overcome by a range of compensatory factors which create better more rewarding work to offset the lost output - Karl Marx described these collectively as Compensation Theory and did his best to debunk it. Nevertheless in general Compensation Theory has held true; as we have created new technologies to replace labour the overall demand for labour in the market has risen in new industries and types of work, although the process has not always been smooth and people have suffered in the gaps between job destruction and job creation. Something I saw recently brought all this to mind again, along with the question of whether Compensation Theory can continue.

A robot stock controller, “walking” the aisle of a supermarket checking the stock levels against the store planogram, is being trialled in the USA. Until now the effects of robotics have been, whilst dramatic, limited. They have reduced the amount of manual labour required to deliver manufactured product, but rarely eliminated it. Most importantly robotics has not previously totally eliminated the human component in a process. The robot stock controller changes this, in that it removes the human component entirely. The robot stock controller can identify exactly what stock is needed where and transmit that back to the central stock control system, it creates the knowledge of what needs to be done. The next step will be robots re-filling the shelves, robots telling you where to find things in the supermarket, and when you check out at the automated self-service till you will have shopped in a service provider operated entirely without human intervention. It is reasonable to forecast that within twenty years the majority of shop floor jobs in the self-service retail sector will have evaporated.

Robots have been widely adopted to perform repetitive manual static tasks, but robots walking (or rolling) around performing their work is a game-changer. There has been extensive development in autonomous military robotics over the last five years to carry supplies for troops, scout ahead etc., but mobile robots roaming around performing repetitive tasks and reporting back to a central computer is a significant advance - the robot parking warden, street sweeper and delivery driver are only a few lines of code away. Similarly the self-driving cars we keep hearing about are making great strides, and the first driverless taxis are probably less than a decade away.

All these innovations depend upon three developments, battery capacity, miniaturisation and AI - “Artificial Intelligence”. Big batteries are required for robots to move autonomously, but we can make them now. AI is similarly making great strides and the best computer “brains” are awesomely capable, but at the moment the computers required to implement them are big and power hungry. Miniaturisation is steadily reducing their size and power demands enabling both the rapid advance in robotics and the availability of AI to companies without nine-figure IT budgets - it is not only the march of the robots performing manual tasks that we need to fear, higher occupations such as investment managers aren’t looking too safe either as computing power continues to shrink in price. Another new technology, 3D printing, really just another form of robotics, looks set to devastate construction employment. Basically the majority of rule or process-based jobs are under threat, the future belongs to people who solve problems or provide services  for which a rule-based approach is too complex, and much of this transition will be in our lifetimes.

Which brings me back to Compensation Theory. To date we have created new jobs, new services, new industries which have mopped up the talents of those who previously did something else and created even more jobs, but how far can that continue? It’s all very well creating new products and services but there is only so much more that we can consume. Personally I’ve run out of consumption, I can’t eat much more, I don’t need more gadgets, I replace my clothes when they’re tired etc., my consumption is not going to increase dramatically to create employment for the massive numbers of people who will be displaced by the next wave of automation.

To date the victims of automation have largely been limited by cost, their work must be so costly that it is worth spending a lot of money to eliminate them, or so simple that we can cheaply create a machine to perform it. The raw underlying cost of developing complex technology has protected many jobs. The recent advancement in robotics show us that this is no longer true, computing, battery power and electromechanics have become both capable and cheap enough to replace the average worker. In this new reality we need to ask where the new work is coming from, and what we will need to be in order to continue to have work.

Recent research by Oxford University academics Michael Osborne and Carl Frey has categorised today’s jobs by their likelihood of computerisation. Unsurprisingly the medical and care occupations come out on top as least likely to be replaced by computers, closely followed by creative occupations and scientists and engineers. At the other end of the scale things are looking very bleak for telesales people, insurance underwriters, all sorts of clerks and tellers, machinists and most factory workers. In conjunction with Deloitte the academics have determined that around a third of current jobs in the UK are highly susceptible to computerisation over the next twenty years, so we had better start thinking about how to cope.

Harking back to the Luddites, there’s no way out of this cycle as the Luddite movement clearly demonstrated. Automation is a business arms race, if you don’t do it your competitor will, and they will gain advantage so you can’t afford to not do it. It’s been going on since the industrial revolution, but again computing and IT are changing the pace. Whilst an IT Director in the UK I helped to lead my business through tenfold growth in shareholder value whilst tripling the size of the workforce - I didn’t destroy jobs, I created them, but at a decreasing rate per pound of corporate income as I progressively made each worker in the business more efficient. That can’t go on, there will come a point soon where people like me are destroying jobs again because we are increasing corporate efficiency faster than the growth in available work to occupy our more efficient workers.

The academic research that has analysed the susceptibility of jobs to computerisation has also identified the primary characteristics which determine susceptibility; these appear to be Social Intelligence, Creativity, and Perception and Manipulation; and those jobs which have lowest scores in these characteristics are the the ones which will go first, the high scoring jobs are the safest. Hence the apparent dichotomy that a Care Assistant is as safe from computerisation as a Surgeon, because the former has to be high on the Social Intelligence scale whilst the latter tops out on the Perception and Manipulation scale. Unsurprisingly the Computing disciplines are also generally pretty safe.

Twenty years is not long, today’s school-leavers will probably be working for around fifty years, so we need to plan ahead. Our future prosperity depends upon constructing an economy based upon social interaction, creativity and unique problem solving - so perhaps this is where government and business leaders should be focusing their attention. Some strategic thinking now will avoid a lot of misery as the age of robotics and AI rolls into town.