Posts

The Future of Work: Automation and Continuous Change?

By Prof Ken Eason

Predictions for the future of Work

Christmas saw the publication of another forecast of the number of jobs that are at risk because of the march of robotics and artificial intelligence. This time it was the IPPR (Institute of Public Policy Research) forecasting that up to 44% of UK jobs are at risk across wide sectors of the economy.

IPPR Report on Managing Automation

There are now many forecasts of massive job losses and attention is being focused on a world where a small proportion of people (the highly skilled ones) will be employed and the rest will be out of work and poor.

But there is also another common theme in the debate about the future of work. It is that we exist in a complex, ever changing, interconnected, global economy and that to survive organisations have to be flexible, resilient and adaptive. The cry is that:

The Only Constant is Change

 Who will manage the change?

How do we reconcile these two different perspectives on the future of work?  Our clever technology may be very good at doing the operational work but it cannot help us make sense of the messy world of international trade, market forces, competitiveness, social change, government action and technical innovation. And it cannot determine what we should do to take advantage of new opportunities and defend against threats. AI may be smart but it is a narrow intelligence with a clever understanding of a specific work domain. Indeed, such narrow AI is also known as weak AI because it cannot replace the breadth of capabilities of a human. However imperfect they may be, human beings are currently the only general purpose intelligent resource we have that can make sense of a confusing, changing world – sentience is known as strong AI.

How is a work organisation to manage in a changing world with a small labour force and a large and sophisticated technological base that may be difficult and expensive to change? The small labour force may have a big agenda: to manage the technology and make sure nothing goes wrong, to monitor the outside world and spot opportunities and threats and to design and implement new ways of working to meet changing requirements. And to keep doing all of these things all of the time. There are many reasons to predict that this model of future work organisations will be ineffective and could be dangerous. One of the reasons for this prediction is what we know about how work actually gets done.

People as the adaptive, coping agents in work systems

 Every study of how work actually gets done shows that it is rarely done strictly according to the formal processes specified that may be embedded in the technology. The people in the work system embellish the formal processes with their own knowledge, often tacit and undeclared, in order to give work delivery the flexibility to meet varied and emergent requirements. They are the ‘oil in the system’ that ‘keeps the show on the road’. They recognize what is new and different, learn how to adapt, and add new, often unspecified, procedures to the repertoire of the organisation. In doing so they often have to ‘work around’ inflexibilities in the formal system to get work done and meet customer requirements.

As a result in any well-established work system there are people who have a deep but often implicit understanding of how the system actually works and a learning capability that means there is a bottom-up process of adaptation and evolution in place that responds to local changes.

The danger of the current narrative about robotics and artificial intelligence is that it implies the replacement of this human resource with technologies that will produce the work on their own. If that is the case not only will work systems become less resilient and adaptive but all the collective tacit knowledge will be lost. And as the saying goes ‘you don’t know what you have lost ‘til it has gone’.

There is always ‘Organisational Choice’:  changing the balance of task 1 and task 2

 To their credit, the IPPR recognize that it is only some of the tasks that can be automated and there are many other parts of jobs that are best done by people. So instead of just assuming technology will replace people we have to ask how the new technological capabilities and the very different capabilities of human resources can be harnessed together for the long-term resilience and adaptability of work organisations. The solution has to be sociotechnical change not just technical change. There will be significant organisational choices to be made to find the right solutions and we need some principles to guide this process. Here are a few to consider:

  1. Immediate cost-effectiveness may be a dangerous objective. The key argument for automation may be economic – you get greater and more reliable productivity from robots and they are cheaper than human resources. That may be so, but you also have to consider what you might lose….
  2. Knowledgeable and skillful human resources provide a sense making resource that can cope with the unforeseen. We need to keep a general sense making capability at all levels within the organisation; to keep a watchful eye on our technology and to provide flexibility and adaptability wherever it is needed. But to be effective people need to keep their knowledge and skills up-to-date and that means actually doing the operational tasks some of the time. So, enabling them to ‘keep their hand in’ is an important design criteria for future systems design.
  3. Having people who understand the task domain means there is a double-task resource to add significant knowledge to planning future developments. Task 1– getting today’s work done – has dominated.

People also have Task 2 abilities – to step back and reflect, to review their performance, to see what can be improved etc. The more they can do this, the better chance the organisation has of coping with the need for continuous change.

Helping people and organisations develop their Task 2 capabilities is an important part of the Bayswater Institute mission. It could be that one of the consequences of robotics and AI will be that people need to spend less time on Task 1 and they can spend more time on Task 2 – in particular thinking about how the work system may be changed to meet new challenges and opportunities. Exploring the potential impacts up-front would seem a good investment in that this is a global challenge and will generate new requirements of the work force that could benefit from planning rather than reacting.