With automation set to have a powerful impact on the future of business and society, PwC has appointed its first UK artificial intelligence leader. Euan Cameron has taken up the newly created role, in which he is responsible for driving growth around this area of emerging technology. He is responsible for AI across the UK firm, helping not only clients to identify opportunities to embrace the shift to automation but also PwC itself.
“To be successful, AI needs to be implemented as part of a broader business transformation strategy,” says Cameron. “This year we expect to see more and more artificial intelligence initiatives move from theory to practice. We are exploring a wide range of opportunities to deploy this technology internally, but also to help our clients with many of their critical business issues. Different industries are adopting robotic process automation (RPA) and AI at varying rates but with the field developing so rapidly, those who wait could miss out on the myriad positive opportunities it will bring.”
MARGINALIA spoke with Cameron to explore PwC’s latest research, UK Economic Outlook, and his views on the impact of artificial intelligence on the future of work. He shares the likely influence of automation across industry sectors and according to the characteristics of individual workers. He describes the constraints that automation might face, but also the opportunities it creates.
Gloria Lombardi: Emerging technologies are already disrupting how we live and work, and their potential impact can no longer be ignored. You have explored this in details in your latest study, UK Economic Outlook. What types of jobs will be most affected by AI over the next 15 years?
Euan Cameron: We found that around 30% of existing UK jobs are susceptible to automation from robotics and artificial intelligence by the early 2030s. This is lower than the US (38%) and Germany (35%), but more than Japan (21%).
The likelihood of AI impact appears highest in sectors such as transportation and storage (56%), manufacturing (46%), and wholesale and retail trade (44%), where manual and routine tasks are more susceptible to automation.
Among the industries that face the lowest risks of automation are education, health, and social work for example, which still require uniquely human skills and therefore will be less automatable. That said, we believe that no industry is entirely immune from future advances in robotics and AI.
According to our analysis, the potential impact of job automation can vary according to the characteristics of individual workers. For example, we found that jobs with a high proportion of male workers (35%), particularly those of men with lower levels of education, are at higher potential risk of automation than jobs with a high proportion of female workers (26%). Generally speaking, this reflects the fact that the most highly automatable sectors, such as transportation and manufacturing, tend to have more males working in them. Whereas, female workers are more concentrated in occupations requiring higher levels of social skills, such as health and education.
GL: Which sectors are embracing AI more rapidly?
EC: The sectors that have to react very quickly to real-time information, and where small differences in their operations can make large impacts on profitability, are ahead of the curve. The financial services, for example, is further ahead. There are very few CEOs in that industry telling us that they have not thought about implementing AI in their business.
Obviously, the technology industry is up to speed, having a clear business proposition and interest in understanding AI.
Consumer intensive industries, such as Telco and utilities companies with a huge base of customers, are further up the curve too.
But many other sectors, including facilities and manufacturing for example, are still holding back.
GL: What are the main concerns around applying AI in the business world?
EC: In many cases, there is a confidence gap. To start with, there is not enough personal understanding among the key decision makers within a company of what AI actually is and how it works. There is also a lack of a conceptual framework to help leaders get comfortable with the technology and where the use of AI – from design to implementation – can be done in a controlled, transparent, and effective way.
There are also concerns around control because of the nature of the technology and the delegation of decision making which comes with it. Ultimately, there is a general lack of trust.
Finally, there are important questions around talent: How do we acquire and retain the right skills? How do we up-skill the current business to accommodate the technology? Do we have the right talent that is required to understand the nature of AI?
GL: As for any important phenomenon affecting our society there are understandably many discussions and even arguments around AI and the future of work. One is the ‘jobs destruction versus jobs creation’ comparison. How do you see it?
EC: In many cases the nature of jobs will change rather than disappear because of AI. And automation will also enable some workers to focus on higher value, more rewarding, and creative work, removing the monotony from their day jobs.
We are of course expecting new types of roles too. The confidence that we can gain by looking back in history is quite significant after all. If I had said to my grandmother that I would be leading an AI function, she probably wouldn’t have any idea of what I was meaning. Indeed, many jobs of her time no longer exist and many others have appeared in our society today.
There has always been disruption whenever new technology has ticked. We saw it with the industrial mass production, the internet, personal computing, and enterprise technology. In all those instances, the technology created periods of rapid disruption in terms of jobs. New technology has always made some tasks and roles disappear more quickly, but it also has created new occupations. This is how our society has evolved. The same is true with artificial intelligence.
But the AI revolution changes the balance of the scale – the demand for humans versus machines. Machines are becoming very good at performing certain types of tasks such as repetitive chores, knowledge identification and retrieval, and pattern recognition. They never tire, and remain vigilant at all times. This leads to the replacement of some jobs. But in most cases the technology only replaces bits of roles, freeing up the time for humans to focus on the tasks where they remain uniquely good at – emotional intelligence, common sense, resolving dilemmas, compassion, wisdom, creativity, and innovation, for example. So we see a sort of rebalancing in the AI economy where there is disruption, yes, but not ‘disruption of jobs’.
Having said that, AI does place a burden on both individuals and society to enable workers to refocus and retrain. It might even require a new type of education for a reinvention of skills. AI should boost productivity and will not necessarily reduce total employment in the long run. But it could also widen income inequality if workers do not develop the necessary knowledge and skills to thrive in the digital economy.
GL: What type of skills should we develop to stay relevant, valuable and productive at work?
EC: We need to develop an attitude towards continuous reskilling. It’s about understanding and recognising that existing relevant skills, that we may have spent so much time on to develop, might not be sufficient or applicable throughout our career. It requires the ability to adapt over time and retrain throughout our working life in the face of the accelerating pace of technological change.
We need to develop a mindset of continuous evolution. We must accept that no business function is immune from the developments in AI and automation. This technology is not just about humanoid robots, but intelligent machines that can interact with our shared environment to enable better decisions. Also, and importantly, a mindset that recognises the value of data, which is an asset to any business, and it’s going to be the fuel that makes AI work.
Finally, we need to encourage our society to focus on the importance of the human skills we talked about earlier, which will always make humans unique and valuable.
GL: Could you share some use cases where AI is already making a substantial impact inside the business?
EC: AI can potentially penetrate every business function within an organisation. There’s currently a lot of impact around HR administration, payroll, and account processing. Automation takes up administrative tasks, with virtual agents replicating the work of a human being – so the technology copies, emulates, and learns from a worker. Then it keeps learning from its own experiences.
Ultimately, it performs repetitious, slightly lower value tasks, allowing the workers to focus on where they can bring in more creativity and innovative thinking. We see this happening quite a lot inside financial businesses among others.
Other important transformations are around the application of predictive analytics to production lines and process industries. Here the machinery and systems are constantly monitoring their own behaviour, comparing it with data from historical performance. The machine is therefore able to predict when it’s going to need maintenance, when it is going to fail, and when human intervention is required. This is happening right now in environments such as the chemical, petrochemical, and manufacturing industries. The impact is quite remarkable – self-monitoring machines, perhaps on the production line or consumer facing (like vending machines), learn the high and low demand periods, call for maintenance, and order their own stock. The time and effort saved on these routine tasks is not insignificant.
GL: Could you share with MARGINALIA readers your top tips on how best to start approaching AI inside their organisations?
EC: Take small steps first to get comfortable with the use of AI. Make sure that you have a defined strategy in place (exactly in the same way you have a strategy around any other business matter). The strategy will allow you to know what you are trying to gain, what costs you are trying to save, or revenue you are trying to accrue. Have a clear data acquisition and data management programme in place too.
Be realistic about the time it takes to design and train the AI system. There are already incredible AI technologies out there, but to get them to run effectively inside your environment you must not underestimate the focus and effort required, and even the distraction that the project might cause to your human experts.
Finally, very importantly, incorporate reassurance and trust inside your business at every step – from the strategy through to the design, implementation, and operation. If your people do not understand why and how the decisions have been made, and what is going on with the systems, then any AI efforts and investments will be ineffective, decisions undermined, and results jeopardised.