The march of the robots
The emergence of artificial intelligence (AI) and machine learning – effectively the use of technology to automate processes and make decisions on the back of experience and data – promises, or threatens, depending on your view, the biggest upheaval in the world of work since the arrival of the internet.
According to research by the Office for National Statistics, which analysed the jobs of 20 million people, some 7.4% of roles are deemed at high risk of being replaced by automation, while the majority of technology leaders responding to a survey by recruiter Harvey Nash expect at least 10% of roles in their organisation to be replaced in the next five years. For employers, this presents challenges; not least in making sure they have the skills required to meet the future needs of the business.
A survey by the Institute of Direct and Digital Marketing (IDM), for instance, found that while 45% of organisations already have skills in AI and machine learning, 87% believe that developing skills in these areas is vital. “Established roles that are in increasing demand up to 2022 are roles that are built around the use of technology, such as data analysts and scientists, software and application developers, ecommerce and social media specialists,” says Karlien Vanderheyden, professor of people management and leadership at Vlerick Business School.
“Also expected to be in demand are roles in which ‘human’ skills are key aspects, such as customer service employees, sales and marketing professionals, training and development specialists and innovation managers.” Other roles will have to adapt. According to the World Economic Forum, more than half (54%) of all employees will need significant reskilling and upskilling to help cope with a more AI-led world.
“Parts of most professional jobs will be automated, but few of these jobs will be fully assumed by machines,” says Alexandra Levit, futurist, consult - ant and author of Humanity Works. “Rather, humans will move into more strategic roles responsible for overseeing machine contributions and explaining these to human leaders. If they are smart, rather than haphazardly automating huge swathes of their work product and removing humans from the equation, they will proactively train existing employees to leverage technology to do the work better and more efficiently.” A good example of how jobs might evolve comes from the air industry, says Albert Ellis, CEO of Harvey Nash. “The best examples are pilots of modern commercial jet aircraft who are there to oversee the autopilot and take control should anything unexpected happen,” he says.
“The modern pilot’s job has changed completely, and is now focused on being able to operate a highly tech - nical and sophisticated aircraft. The number of flight officers required has reduced with modern computer-based integrated satellite navigation but the piloting of the craft remains a key task, albeit requiring modern technology skills as opposed to an ability to physically fly the aircraft.” The good news for employers is that employees appear willing to adapt. According to global research by software firm Blue Prism, 83% of knowledge workers are comfort - able with reskilling in order to work alongside the digital workforce, while 78% say they’re ready to take on a new job role.
A study by Ricoh Europe, meanwhile, finds 70% of workers expect to have to upskill throughout their career. “The expectation for years has been that introducing new innovations in the workplace will result in friction from employees who are set in a specific way of working,” says Edward Gower-Isaac, vice president and general manager at Ricoh Europe. “But we’re seeing a real shift in attitude in today’s workforce.”
Alongside adapting existing roles, entirely new jobs will also be created. Analysis from jobs site Indeed found the number of roles related to artificial intelligence rose by 485% in the three years prior to 2017, and suggests there are currently twice as many jobs requiring AI and machine learning skills as there are applicants, meaning firms need to think now about how they can go about hiring and developing talent to fill such posts. “Key new roles that will soon be in demand will be data science leaders, data translators who can provide the link between the data function and the business, and data professionals that specialise in particular domains, functions or analytics techniques,” predicts Laura Timms, product strategy manager at MHR Analytics. Some 80% of UK companies are planning to hire a data scientist or seek data consultancy in 2019, according to its recent Data Surge report.
Vanderheyden predicts an even more exciting line-up of new technology-related positions, including human-machine interaction designers, robotics engineers, drone instructors and operators, and virtual reality designers. “Several ‘human’ skills will become more important, like creativity and originality, critical thinking, complex decision-making, and persuasion and negotiation,” she says. “Emotional intelligence, social influence, leadership and service orientation will also become significant aspects of jobs in the future.”
Bruce Morton is a specialist in international workforce design and author of Redesigning the Way Work Works. He believes employers need to take advantage of the willingness of employees to engage with change, by having conversations with them about how the business operates. “Rely on them to suggest which tasks are redundant, less efficient or a pain in the neck,” he says. “Then huddle with project leaders and managers to redefine workflows. If you’ve made the right changes, productivity will improve. If not, you’ll see where a new approach is necessary.” Already, sectors are responding to the opportunities presented by AI and machine learning. “Organisations that are investing heavily, such as large insurance companies where processing claims and issuing policies form the majority of their activities, are preparing staff for an accelerated transition as the benefits are so material,” says Ellis.
“We’ve seen a number of announcements from that sector in relation to this.” The legal sector, too, is making use of AI. “We are seeing intelligent software solutions to assist the large-scale review of documents and contract reviews and negotiations,” says Jonathan Rennie, partner in the employment team at UK law firm TLT. “This is not about replacing individuals; it’s about lawyers developing new competencies and working with technology to enhance the client offering.” The accountancy sector is also a prime candidate to take advantage of new technologies, although it remains early days, says Kirstin Gillon, technical manager, Tech Faculty, at ICAEW. “AI and machine learning have particular resonance in this profession because they’re about intellectual work and knowledge which we haven’t been able to automate in the past,” she says.
“AI can help us gain more insight from data, freeing time to spend with clients to give other types of strategic advice.” Accountants in future will at least need to be able to ask the right questions to get the most out of the potential insight from such systems, she adds, and some will need to be more embroiled in the models and processes that sit behind packages. For now, it’s robotic process automation (RPA) that is having the biggest impact, says Andrew Moyser, a partner at MHA MacIntyre Hudson. “I’ve seen examples of ‘BOTs’ being able to deal with new starters in checking applications, entering them onto the systems and contacting each of the teams, as well as client set-up procedures and the preparation of basic personal tax returns,” he says.
“But the most significant benefit comes from consistency in procedure and compliance across multiple employees and offices. Accountancy firms spend vast resources ensuring procedures are completed correctly and consistency; something which computers and RPA do as standard.” Accountants can also help clients prepare for the introduction of AI and machine learning in their businesses, believes Grant Anthony, partner in the business solutions team at accountancy firm Crowe UK. “The first step is therefore to be prepared and open for this change by keeping themselves knowledgeable and informed on this topic so they are in a good place to inform clients and gradually improve the status quo,” he says. In time, the use of such technology is likely to change the way in which the profession recruits and develops staff, believes Moyser.
“The standard training programme adopted over the last 30 years is unlikely to be fit for purpose,” he says. “Arguably there will be a greater requirement for advisers or consultants, and less of a requirement for those doing the more manual or administrative tasks. “Traditional trainees are unlikely to be needed at the same levels whereas the requirement for qualified staff is likely to increase,” he adds. “The challenge is how you train and develop staff to get to the required level.” He believes this could lead to an accelerated college-based training programme in the early years with more emphasis on practical experience towards the end.
Yet not every role is heading for automation, and there are some elements of many jobs which simply cannot be replaced by robots. “Areas such as critical thinking, advising, customisations, arts, design, dedicated customer service, empathy and sympathy are not likely to be fully integrated into robotics any time soon,” says Antonio Espingardeiro, a member of the Institute of Electrical and Electronics Engineers and independent expert in the field of robotics and automation.
“Robots can’t take over where there is creativity, emotions, social intelligence and human contact involved.” It’s a point that Alexey Utkin, principal solution consultant in the finance practice of technology consultancy DataArt, also makes. “It is sometimes said that barmen and barwomen will be the last people to be automated,” he says. “While it is possible to automate the role, it is not necessarily desirable. A bar server is to many a psychiatrist, a person, a friend. Who wants that to be replaced?” He also raises concerns about the potential for technology to go wrong. “Recently there was a case of recognition algorithms being broken by someone holding up a picture in front of his face,” he says.
“One doesn’t have to think too hard to work out opportunities for terrorists, thieves, industrial spies and other wrong-doers.” In reality, much of the use case and application of AI and machine learning has yet to be worked out, but the potential benefits and implications for staff mean this is an issue that no firm can afford to ignore. “The impact of technologies such as automation are being felt from the boardroom to the shop floor,” points out Pat Geary, chief evangelist at Blue Prism. “Our global research reports that 92% of business decision-makers plan to extend use cases of automation across their business. The opportunities for growth are clear.”
Since starting up three years ago, online mortgage broker Mojo Mortgages has made use of AI to process application data and submit forms to lenders. “It’s a perfect example of something that a human being would traditionally do but doesn’t any more,” says Richard Hayes, CEO and co-founder. “The driving factor towards deploying robotic automation was to improve the colleague experience, because no one likes keying information into an application form, and to give us the ability to scale quicker.”
The business has grown from 30 to 77 people in the last seven months; something he attributes to the ability of AI to help it process applications effectively. Mojo also makes use of AI to suggest suitable products for would-be customers based on their individual circumstances. “We can more accurately match people with the most appropriate mortgage in a significantly faster timeframe than the traditional model,” says Hayes.
“It means we don’t have to wait for a human being to process an application before we can go back to customers.” Employees, meanwhile, have more time to speak to customers where help is required, he adds. “We try to make sure they spend as much time as possible doing the work that only human beings can do,” says Hayes. The firm doesn’t yet have sufficient volumes of data to make use of machine learning, says Hayes, but believes this will form part of its operations in the future. “With customer consent, we could get insight into streams of data to help support them to become mortgage-ready,” he says.
“We want to help more people get on the property ladder by giving them information in a more timely and sophisticated manner, without having to speak to every customer.” For technology-based businesses such as Mojo, Hayes believes making use of such new tools such as AI and machine learning is vital, especially for those seeking investment. “There’s a level of expectation from institutional investors that you have some of those boxes ticked,” he says.
Originally published in Economia on 18 July 2019.