Staying ahead of the AI game

In January, the Reskilling Revolution Platform was launched by the World Economic Forum. The scheme aims to future-proof workers from technological change and prepare them for the Fourth Industrial Revolution The evolution of AI from being the butt of Terminator jokes to a serious policy challenge is now complete. The announcement included a pledge for “better skills, better jobs, better education for a billion people by 2030” with six countries as founding members, including the US. The platform also set expectations with large businesses such as Salesforce, LinkedIn & PwC to bring this initiative to their employees. Technological anxiety is not a new phenomenon and neither is the displacement of work. A quick glance at history shows us the evolution of work and the eventual displacement of skills. Our agrarian societies have given way to a manufacturing economy. Over time that dependence on manufacturing jobs has been replaced by service sector economies. If you are a factory worker, the struggle against both outsourcing and automation has been a way of life for some time. This fear is now just catching up with the service economy. Our education systems have also followed suit, based on the shift in economic activity. Society has never known what the job of the future is but there is a tendency of developing what kind of a human it will need. When the Industrial Revolution began, education moved to support the assembly line worker. We were taught order and rules. As the need moved towards services, the education system upheld the opportunities for specialisations and the emphasis on higher education increased. Through this process emerged some winners and some losers. Today, more than 30% of US workers believe that their jobs will be replaced in their lifetime. Our insecurity and anxiety is not linked to the future of our children but a direct fear for ourselves. Building on these fears, we also saw signs of protest at Davos. Signs saying “Eat the rich” and discussions on the responsibility of companies to stakeholders held centre stage. Underpinning these initiatives is the acknowledgement of what is now called the Fourth Industrial Revolution. From academic Klaus Schwab to Alphabet chief executive Sundar Pichai, the effect of the progress in machine learning and technology on the social equality is a topic of reality and not fiction.  So what is a job? To understand the implication of technology on the labour markets, we need to understand the meaning of the concept of a job. No job is absolute in definition. A job is usually defined by a collection of tasks. If we look at job descriptions, we will find that there are some qualifications to achieve these tasks but not every task requires that qualification. The separation of these tasks sometimes results in various positions of the same job. If the question you want answered is “will my job be safe?” then the resolute answer to that question is – yes. However, technology does have the ability to take on tasks within a job which may convert the nature of that job itself. Economists have studied the impact of technology on jobs since the first wave of AI in the 1960s. In the first wave there was a belief that computers could automate tasks by replicating a human. They would find an expert and ask them to explain what logical steps they were taking to complete a task. Economists looked for two requirements to judge if a task could be automated: Does the task have a defined goal? And can a human explain the process of how to get to that goal? These tasks were defined as “routine”. It does not mean that the job is boring. It just means that it is explainable. The adoption of software and technology to automate the “routine” tasks is the first wave of AI. In The World Without Work, economist Daniel Susskind defines this as “task encroachment”. The tasks that comprise our job are slowing, being encroached by complementary technology. ‘Best guess’ solutions The role of technology has been to thus make the human more efficient. Drivers can use satnavs to get to places without memory, surgeons can operate with guided robots and architects can use software to design complex spaces. However, there are certain jobs that we assumed machines couldn’t do. These are the tasks that economists defined as “non routine”. When an expert was unable to explain a process, when a task was not logic or process-driven, we assumed those jobs would always remain with the human. However, the advances in machine learning makes this statement untrue. As long as a task has a defined goal and available data, machine learning no longer requires a defined process. It utilises probabilistic models to achieve to the best guess solution. The “non-routine” tasks are now available for task encroachment. As the second wave of AI takes on more tasks, the human is now needed for tasks without a clear goal or data. The world as we know it is siloed into roles of occupation. We always want to know if “accountants” will be safe or if we should be moving our skills become a “developer”. The skill of coding is already being replaced by data science. The future of work is not linked to roles but to tasks that are innately human. In order to be successful, we may need to be better people. In September 2013, economists Carl Benedik Frey and Michael Osborne released a paper suggesting that nearly half of all US jobs were at risk of automation. In 2017, McKinsey released another study on the jobs whose share of activity could be automated in the future. The tasks that showed the most unpredictability were jobs linked to managing and developing people. Tasks that involved social co-operation, empathy, ethics, creativity, curiosity and innovation become “human critical”. Basically the human is needed for the unknown. It also suggest that traditional roles, ironically traditionally female roles, such as kindergarten teachers or nurses, will continue to be in demand. These roles may not be highly paid but still need interpersonal skills. But herein lies the conundrum that our society faces. The technological encroachment sets the seeds for a very large wage gap. Jobs are separated out into high skill, high wage and high skill, low wage groups. There will be many more jobs in both these categories but it will also lead to a growing social inequality. UK manufacturing produces 150% more output today than it did in 1948 but it employs 60% fewer workers. In 2018, the most valuable company in the world was Apple with 138,000 workers compared with AT&T in 1964, which employed more than 750,000 workers. The economic contribution per employee to the growing pie is higher but the volume of opportunities are lower. Riding the first wave According to the British Property Federation, the commercial real estate sector in the UK directly employs more than 1m people and supports an additional 1m jobs. There is an onus on the industry to prepare its existing workforce for redeployment within or outside the organisation. The real estate sector mostly believes that it is still a participant in the first wave of AI. There is an enormous knowledge gap in our understanding of technology and it seems to believe that the “non-routine” nature of most of its tasks make its safe from massive technological encroachment. I work in an environment where we engage with the breakthroughs in AI as part of our everyday tasks. This mindset concerns me about the lack of responsibility to the future of our employees. In 2017 when we first launched our product we were restricted to one single job function for property managers. In just two years that  scope has expanded to wider functionalities of leasing and resident requests. Whereas in the past, when we could only automate half of the interactions for a very specific function, we can now achieve higher metrics for a wider variety of functions. In such a situation the role of the human needs to be defined with new goals. The old performance metrics don’t apply anymore. At one organisation, an employee requested that the AI handle the critical path when they were on holiday but on working days add them back to the critical path for “approval”. The role of the employee in this situation is not clear and the organisation is doing nothing to prepare them for their future. Technology is not just an economic driver. It is also a social one. The labour markets are the conduits for economic contribution and social mobility. Unemployment or “under-employment” due to economic conditions can be resolved but unemployment due to structural technological change requires us to rethink the nature of work itself. It also requires employers to engage in a real conversation about the nature of tasks available to their workforce instead of denying the impact of technology. Tripty Arya is founder of Travtus