Welcome to Antenna, our glance into the middle distance and the life-changing developments the near future will bring. We’ll also look at the investment implications which, as ever, are far from intuitive.
The world of work is becoming turbulent as AI scales the list of skills previously thought to be the preserve of humans. But when the dust settles, AI should enable us to fulfil our real potential.
Fear is proliferating through almost every profession as the realisation dawns that the aim of AI – to create machines that can perform any cognitive task a human can – seemingly puts every job in the sights of technologists. In reality, while there will be many casualties as the technology improves, AI should ultimately enhance what we can achieve and allow us to concentrate on the most human skills.
Research from Oxford University suggests that 47% of all jobs in the US could be automated within 20 years. “If your job is essentially routine and predictable, you’re at risk, while skilled manual trades like electricians, plumbers and nurses can’t be replaced,” says futurist Martin Ford, author of Rise of the Robots: Technology and the Threat of a Jobless Future. The latter set of jobs is protected by the difficulty and expense of building dexterous robots to give AI physical form, and the possibly insurmountable problem of creating machines capable of empathy.
Repetitive jobs have often been the first to be disrupted through history. But where physical work – be it on the production line or the telephone switchboard – was the target of automation in the past, AI makes it possible to automate brainpower. “AI broadens the range of jobs that machines can take over into almost everything involving data,” explains Tom Hulme, General Partner at Google’s venture capital arm, known as GV.
The kinds of jobs that will be affected first are rules-based and non-subjective, such as data collection and synthesis. One face of that change is robotic process automation (RPA): the software slots into a company’s existing IT system, logging on as if from a desk and mimicking the behaviour of a human employee. “With RPA you can take, say, 60% of your company’s rules-based repetitive tasks and get the robot to take the repetitive work away from employees,” explains David Wright, Director in Robotic and Cognitive Automation at Deloitte. “Software robots can open email attachments, complete e-forms, record and re-key data, and replace many middle- and back-office processing centres. Our analysis suggests the cost of a digital worker is about one-third the cost of an offshore worker.”
RPA-based systems also run non-stop, so it is easy to see the productivity potential. According to Raja Subramanian, Group Enterprise Architect at technology services company Xchanging, robots processing insurance-related tasks can clear over 30,000 cases per month, reducing processing time from five minutes to under 10 seconds.
Advances in natural language processing mean AI is even steadily encroaching on jobs involving human interaction. Already, Barclays is using conversational AI chatbots to deal with unbanked populations, including the illiterate. In education, AI is starting to replace routine jobs – for example, language learning platform Duolingo teaches 200 million people new languages by getting students to listen and repeat. And taxi drivers are starting to grow concerned for their jobs.
So tasks based on set rules and processes, even those that may seem on the surface very human, are already feeling the effect of AI. But the programs processing and analysing data still need human oversight and with the opportunities companies are seeing in unstructured data blooming, they are starting to employ more data scientists and advanced analytics and machine-learning experts.
Transforming the professions
It is not just low-paid routine work that is changing. The great strides that have been taken in deep learning (see page 4) mean that knowledge-based professions such as law and medicine are also set to be transformed. The work that UK-based AI company DeepMind has done with the NHS, building on lessons learnt from a computer-game playing algorithm, is at the forefront of this trend. In 2015, the company unveiled a deep learning system that could increase the score in Atari console games without information on the rules. DeepMind recently started applying this technology to develop diagnostic tools for head and neck cancer with University College London Hospitals, acute kidney injury with the Royal Free Hospital and age-related blindness with Moorfields. In an interview with The BMJ, DeepMind Co-founder Mustafa Suleyman explained that diagnosis is “actually very similar” to the Atari research. When preparing for radiotherapy, a patient’s CT scan is marked up by a radiotherapist to identify healthy and cancerous tissues for accurate targeting. “This takes about four hours for each radiographer to do, so there’s a very long delay from deciding on treatment to being treated,” he said. “We can train a model to do this faster and more accurately.”
In the surgeries and hospitals of the near future, routine check-ups, mammograms, lung cancer scans and even dementia diagnosis will be done by AI. In some cases, such as a pneumonia-detecting program developed by researchers at Stanford University, the software is even better than humans at spotting problems. And as machine learning advances, programs are starting to not just detect, but predict diseases. It is not hard to imagine a point when – as AI starts to reason for itself, unrestrained by human understanding – disease prediction will be vastly improved by the technology.
For now, these developments do not mean specialist doctors, such as radiotherapists, ophthalmologists and nephrologists, will all lose their jobs. Indeed, it should make them more precise and free them up to concentrate on elements of their work that require judgement, creativity and empathy.
However, it is likely there will be some losses. Government adviser Richard Susskind, who looked at the impact of AI across law, education, audit, tax, consulting and architecture, points out that people tend to overestimate how much of their work demands judgement, creativity and empathy. When broken down into its constituent parts, the work of many professionals is routine and process-based. Taking a macro view, a recent OECD study looked at the susceptibility of tasks within jobs to automation and found that, across 21 countries, 9% of jobs are at risk. Compared with the 47% figure cited by the Oxford study, this may seem small, and shows how much debate there is around the topic, but this amounts to millions of jobs.
In the longer term, advances such as AlphaGo Zero (see page 4) show how AI might one day be capable of problem solving in ways we could not imagine. And with teams of researchers plugging away at the challenge of affective computing (understanding human emotions), natural language processing and reasoning, and machines already creating original artworks (see page 12), it is possible that many more tasks could be automated in the future. Even those plumbers and electricians might be out of a job once AI can cheaply be given humanlike form through advances in robotics. That future, of course, is very distant and reliant on technical breakthroughs that may not be possible.
Rewriting the work landscape
It is likely that people will try to resist progress, but ultimately, the advantages of AI, as with steam and electrification before it, will make a necessity of change. Technological revolutions completely rewrite the landscape so that it is not just jobs that are redefined, but the fundamentals of the organisations creating them. As the Massachusetts Institute of Technology’s Andrew McAfee and Erik Brynjolfsson have highlighted, in retrospect, it is clear electric power was so transformational to industry because it enabled the redesign of the production line in ways previously inconceivable, such as the introduction of conveyor belts. At first those advantages were not at all obvious, as the ‘curse of knowledge’ blinded incumbents to the possibilities of the technology, but eventually those who did not forget what they knew about business and go back to the drawing board, failed. Almost every industry will go through this kind of process as AI picks up pace.
It is impossible, standing where we are, to see what is beyond the horizon. But given the sheer scale and pace of change that is to come, everyone will need to be adaptable, from the individual worker to the businesses that employ them. But, while we are headed into turbulent times, AI should ultimately make our work more enjoyable. Waves of technological progress in the past have relieved humans from the drudgery of hard manual labour and created more fulfilling, knowledge-based roles. For now, we vastly outstrip machines at perception, reasoning, communication, creativity (be it coming up with theories or musical compositions) and anything involving empathy or social skills (including things like negotiation, persuasion or care). As Google’s Hulme says: “Sixty-five per cent of children entering primary school today will ultimately end up working in completely new job types that don’t yet exist. Those who truly excel will be able to harness technology for repetitive tasks and save their efforts for more creative work… it could be the best thing that’s ever happened to us.”
1. Rules-based, repetitive jobs will be the first to be replaced by AI, from back-office data processing to dealing with basic customer enquiries.
2. Highly skilled work will also be hit by the coming wave of AI, as the technology automates functions within medicine, law, audit, architecture, education, tax, consulting and more.
3. New technical jobs are being created to make the most of AI, and humans remain better at perception, reasoning, communication, creativity and social skills.
IMPORTANT NOTE: The value of your investments may go down as well as up. Past performance is not a guide to future performance. Any tax allowances or thresholds mentioned are based on personal circumstances and current legislation, which are subject to change. Some products or services may be affected by changes in currency exchange rates. If you invest in currencies other than your own, the value of your investment may move independently of the underlying asset. All information within this publication is for illustrative purposes only and is not intended as investment advice; no investment is suitable in all cases and if you have any doubts as to an investment’s suitability then you should contact us or your financial adviser.
We or a connected person may have positions in or options on the securities mentioned herein or may buy, sell or offer to make a purchase or sale of such securities from time to time. In addition we reserve the right to act as principal or agent with regard to the sale or purchase of any security mentioned in this document. For further information, please refer to our conflicts policy, which is available on request or can be accessed via our website at www.brewin.co.uk. The opinions expressed in this publication are not necessarily the views held throughout the Brewin Dolphin Group. No Director, representative or employee of the Brewin Dolphin Group accepts liability for any direct or consequential loss arising from the use of this document or its contents. The information contained in this publication is believed to be reliable and accurate, but without further investigation cannot be warranted as to accuracy or completeness.