Are all our jobs going to be taken over by robots? Will any skill or profession be safe? What are the implications for how we are schooled, governed and treated in hospitals? And what on earth are we going to do in our cars once they are driverless and don’t require any human intervention? The brave new world of artificial intelligence holds the answer.
In some quarters, it is being called the Fourth Revolution, following those previously brought about by agriculture, industry and the Internet but it is not that new. As a phrase, artificial intelligence (AI) dates back to 1956 and a conference at Dartmouth College in New Hampshire, USA.
As an idea, the notion of trying to build a machine to perform useful reasoning may have begun with the Blessed Ramon Llull, a 13th Century philosopher and logician from Majorca. In fiction, Mary Shelley’s Frankenstein of 1818 envisaged the creation of a human-made grotesque but sapient monster; Aldous Huxley’s Brave New World of 1931 anticipates the development of reproductive techniques that obviate the need for human parenting.
Technologically, Alan Turing’s theory of computation in the 1940s led to the Church-Turing thesis, positing that digital computers could simulate any formal reasoning. This led scientists to consider the possibility of building a human brain, with the first work that is now generally recognised as AI coming in 1943 with McCullouch and Pitts’ design for artificial neurons.
Now, AI is defined variously as intelligence exhibited by machines, the mimicking of human learning and problem-solving by machines and even the eventual replacement of mental faculties by computation and machine learning. In current parlance, however the debate around AI largely sees it as the science that enables computer systems to complete tasks that traditionally would have required human action or intervention.
From drones and cars that don’t need drivers to machines that understand and utter human speech and Internet content delivery networks that provide their own intelligent routing for information, this is a world where functions that have until now been carried out by humans are under threat from automation and machinery.
Advancements in manufacturing, robotics and building infrastructure will make the process of building a car almost completely automated. Whole buildings and factories will be able to self-regulate. Robotic machines will work without the presence of humans and understand is something is wrong. In healthcare, IBM’s Watson technology can already diagnose cancers and some other diseases, while AI is also enabling drugs companies to identify missed patterns and correlations that could identify new treatments.
‘Make no mistake,’ warns Shelly Palmer, chief executive of technology consultancy The Palmer Group. ‘At some level, every job can and will be done by machine. It’s not a question of if. It’s just a question of when.’
Defining robots as machine-learning algorithms running on purpose-built computer platforms that are trained to perform tasks that currently require human beings, Palmer believes the first five jobs that AI will take will be in middle management, commodity sales, report-writing, accounting and bookkeeping and health diagnostics and surgery.
Bernard Marr, author of Data Strategy, goes even further, naming ten jobs most at risk of automation, including insurance broker and underwriter, journalist, architect, banker, human resources director, marketing or advertising executive, lawyer and even police officer.
Indeed, only recently Devon and Cornwall Police announced plans to launch the first 24 hour drone unit in the UK, to help police officers in their investigations. Academic studies and experts agree. Boston Consulting Group has predicted that by 2025, up to 25 per cent of all jobs that are currently available will be replaced by either robots or smart software.
Research at Oxford University has put the figure higher, forecasting that 35 per cent of existing UK jobs could be at risk of automation over the next 20 years.
The sequence, says Palmer, will be that predictable manual repetitive jobs like production line work and cognitive repetitive positions, such as accounting roles, will disappear first, with non-repetitive functions such as those in C-suite management proving more resilient. He believes that the last five jobs that will be taken by machine will be those of teachers, professional athletes, politicians, judges and mental health workers.
‘The last jobs that robots will take share a common thread: humanity,’ he concludes. ‘Jobs requiring a unique combination or human intuition, reasoning, empathy and emotion will be difficult for an AI machine to train for.’ Already there is plenty of evidence of the trend, whether it is referred to as AI, machine learning or the more innocent-sounding Internet of Things.
All essentially replace human functions with automated ones. In journalism, The Associated Press newswire is using the Wordsmith AI platform to compile sports reports.
Donald Trump’s US presidential campaign meanwhile utilised technology called Visitone to automatically scan speeches and statements made by rival candidate Hillary Clinton.
In industry, German engineering group Siemens is automating its 250 factories so that, rather than having supervisors walking around factories with clipboards to record down time and the reasons for it, the company’s industrial machines themselves report such data into a central information repository. ‘Getting this information before has carried a high cost and integration burden,’ says Brian Holliday, UK managing director of Siemens Digital Factory, the division driving the group’s digitalisation.
‘It’s about using machines to talk to machines to so we can understand better in real time all the information on which our factories are dependent and make much better decisions. We want to do for industry what Apple’s iOS operating system and Android have done for consumer electronics.’
Holliday sees Siemens’ automation having little effect on the company’s 350,000-strong workforce, though the jobs will require increasingly high skill levels. ‘We’ll always need a human touch,’ he says, ‘but our digital tools will help us much better capture the kind of capabilities that resided with experts in the plants in the past. We’re not automating for cost reductions; we’re automating for productivity, quality and output.’
In accountancy, Stephen Kelly, chief executive of accounting software group Sage, which last year launched the first Chatbot enabling business executives to file their expenses, believes that AI offers a business opportunity, rather than a threat. ‘AI is the game-changer for the next decade,’ he says. ‘Technology has made accounting business solutions much smarter, with administrative capabilities that manage your business enabling you to get on with your business and follow your dreams.’
Kathryn Parsons, co-founder of teach-yourself computer coding firm Decoded, also sees AI transforming education. ‘If we really are living through the Fourth Industrial Revolution, we have to acknowledge that our education system was invented in the last one,’ she says. ‘It will change. There will be a different education landscape in the next 20 years. Maybe the two or three most prominent brands in higher education will exist in 20 years’ time but pretty much everything else is up for grabs and up for being disrupted. The dream of hugely personalised, democratised, high-value, highquality education for everyone has not yet been realised but I don’t think that means that it will not be.’
In public relations and corporate communications, a recent report by PR agency group Weber Shandwick found that 55 per cent of chief marketing officers surveyed believe that AI will have a greater impact on marketing and communications than social media ever has.
However, practitioners are being encouraged to adopt a positive approach ‘AI will revolutionise lots of sectors, particularly in services, and will change communicators’ lives for sure,’ says John Machin, senior programme director for IT and infrastructure at Hotwire PR. ‘It’s going to change many of our jobs. Initially, it will be simple things. For example, there are already AI tools to do things like schedule meetings.
‘There are plenty of aspects of the job that are quite transactional and administration-based and I imagine that lots of those will be automated which will result in us being able to spend more time advising clients and doing the more strategic elements of the job. ‘Are robots going to take everybody’s jobs? I think that people will need to up-skill and re-skill but actually there will just be different types of jobs.’
Adam Hirsch, executive vice-president of Edelman Digital, adds that AI still has a long way to go to compete with the knowledge and abilities of sentient humans. ‘Turning the physical world into data is not easy. Computer vision has to understand the objects within images and videos,’ he says. ‘We’re not just talking about people, tables, chairs, doors and cars but also understanding the context and relevance of all objects, including 10,000 species of birds and nearly 100,000 species of insects. Inherently, the future of AI is murky. There’s a lot we don’t know yet but what we do know is there’s enormous potential to make the world around us autonomous.’
Some technologists are also sceptical for different reasons. Bradford Cross, founding partner at America’s DCVC, which claims to be the world’s leading machine learning and big data venture capital fund, predicts this year will see break-out successes from AI start-ups solving industry problems that require subject matter expertise and unique data to deliver its core value proposition.
However, he adds: ‘With AI in a full-fledged mania, 2017 will be the year of reckoning. Pure hype trends will reveal themselves to have no fundamentals behind them.’
Cross predicts that Bots will go bust, with mere conversational interfaces over voice and chat applications unable to replace human expression and connection, which after all is the main reason for Facebook’s success.
He is also not convinced about AI-powered personal assistants, but adds: ‘None of my reasons for the Bot bust state that the AI isn’t good enough yet. The issue with most systems like Apple’s Siri is more that they’re poorly implemented. We can build many interesting Bot interfaces using modern techniques but the bigger issue is that it’s not clear that humans want to use them.’
That may change once better platforms that are easier to use are developed and brand leaders emerge. The Weber Shandwick poll, for example, found that consumers most commonly associate AI with robots but few were able to name an AI brand leader.
Few technology experts, however, believe that AI
will not have a transformative impact across industry,
business and society. If that is indeed inevitable, the
focus then changes to how to prepare for the changes it
will bring and Palmer has no doubt about what
workers and managers need to do.
‘If you’re wondering where your job sits on the list
of Run for your life, the robots are coming, you have a
simple, singular mission,’ he says. ‘Learn how your job
is going to be automated. Learn everything about what
your job will become and become the very best
machine partner you can. Everyone will tell you that
none of this is happening any time soon. They are flat