AI: which jobs and sectors are at risk

By , on
AI: which jobs and sectors are at risk

Even accountants' jobs are at risk from artificial intelligence, but there is a potential upside, according to this special report.

Cloud services, big data and mobile devices are just some of recent disruptive technologies that have presented both great threats and opportunities in business and for careers – but artificial intelligence promises to dwarf them in all in scale.

AI-led automation can potentially add up to $2.2 trillion in value to Australia’s economy by 2030, according to The Automation Advantage, a new study by economics and strategy consulting firm AlphaBeta.

However, the study also acknowledges that up to 3.5 million of Australia’s 12 million workers are at “high risk of being displaced by automation in coming years”.

Highly manual and/or repetitive jobs, such as labourers and cleaners, are identified in most studies as those most likely to be replaced by AI and robots, but it’s not just blue-collar workers facing an uncertain future. High-routine white-collar jobs such as clerks and even accountants are also at risk, according to Artificial Intelligence and Robotics and Their Impact on the Workplace, a study by the International Bar Association.

Musk vs Zuckerberg

Not surprisingly, AI is one of the most divisive technology-related issues of our time. The most recent high-profile evidence of its divisiveness came last month when Elon Musk, CEO and founder of Tesla, slammed Mark Zuckerberg for his “limited” understanding of AI, following comments made by the Facebook boss.

The argument began following an interview last month in which Musk reiterated his bleak view of the emerging technology, describing it as an “existential risk to human civilisation”.

“I have exposure to the very cutting-edge AI, and I think people should be really concerned about it,” Musk told a US National Governors Association meeting. “I keep sounding the alarm bell, but until people see robots going down the street killing people, they don't know how to react, because it seems so ethereal.

“AI is a rare case where we need to be proactive about regulation instead of reactive. Because I think by the time we are reactive in AI regulation, it's too late.”

Musk believes the current approach to regulation, which often reacts when “a whole bunch of bad things happen”, such as self-driving car crashes, is not quick enough to deal with the problems AI poses to society.

“AI is a fundamental existential risk to human civilisation, and I don't think people appreciate that," he added. “There will certainly be a lot of job disruption, as robots will be able to do everything better than us – that includes all of us. It really is the scariest problem for me.”

A few days later, during a Facebook livestream with fans, Zuckerberg labelled Musk’s comments as “pretty irresponsible”, saying that he remained “optimistic” and couldn't understand why naysayers would “try and drum up these doomsday scenarios”.

“In the next five to 10 years, AI is going to deliver so many improvements in the quality of our lives,” added Zuckerberg.

A Twitter user then posted a link to Zuckerberg’s comments, which spurred Musk to reply: “I’ve talked to Mark about this. His understanding of the subject is limited.”

Gates and Hawking weigh in

Musk isn’t the only high-profile expert cautious about an AI-fuelled future. Stephen Hawking believes the creation of advanced AI will either be “the best or the worst thing, ever to happen to humanity”.

The scientist admitted that “the potential benefits of creating intelligence are huge”, and he particularly mentioned the positive effects AI could have on disease, poverty and the environment.

“Every aspect of our lives will be transformed. In short, success in creating AI could be the biggest event in the history of our civilisation,” Hawking said last year.

However, he also highlighted his concerns as to how we cannot predict the consequences of having “our own minds amplified by AI”.

Bill Gates argues that if robots are to take jobs away from humans, it stands to reason they should start paying their fair share in tax. Governments should be able to tax companies that make use of robots, as a way of slowing down the growth of automation and to provide funds for creating employment elsewhere, the former Microsoft boss said earlier this year.

“Certainly, there will be taxes that relate to automation,” said Gates, adding that unchecked automation may lead to massively reduced income tax revenue.

“Right now, the human worker who does, say, $50,000 worth of work in a factory, that income is taxed and you get income tax, social security tax, all those things. If a robot comes in to do the same thing, you'd think that we'd tax the robot at a similar level.”

“There are many ways to take that extra productivity and generate more taxes,” added Gates. “Some of it can come on the profits that are generated by the labour-saving efficiency there, some of it can come directly in some type of robot tax. I don't think the robot companies are going to be outraged that there might be a tax.”

Gates has spoken out several times against the prospect of intelligent machines, saying that some decades after they are capable of doing human jobs, they may become a threat.

Gates maintained that given many manual jobs such as warehouse work and cleaning roles will become automated over the next few decades, it is important to have policies that deal with the changing nature of the workforce. 

“It is really bad if people overall have more fear about what innovation is going to do than they have enthusiasm. That means they won't shape it for the positive things it can do. Taxation is certainly a better way to handle it than just banning some elements of it, but innovation appears in many forms,” he said.

The onus is on government to figure this out, according to Gates, and nations cannot rely on businesses coming up with plans for automation. With a growing displacement in the workforce, this excess labour can be used to bolster social services and education, but it requires government oversight to make that happen.

Next: the economic case for AI

The economic case for AI

However, not all agree that AI will necessarily be a terrible thing. According to Sizing the Prize, a new study by global accounting firm PwC, global GDP could rise as much as 14% by 2030 – a whopping US$15.7 trillion boost to the world’s economic output – as a result of AI.

Half of that is down to labour productivity improvements – presumably including job losses – as well as “consumer demand resulting from AI-enabled product enhancements”.

China stands to win the most, with GDP up 26%, followed by the US with 14.5%, PWC predicted. Europe and developed parts of Asia will clock up gains of 9% to 12%.

The gains aren’t only in automation, which is expected to be what kills off up to a third of all currently existing jobs, but in expanding what people are capable of doing.

“The analysis highlights how the value of AI enhancing and augmenting what enterprises can do is large, if not larger than automation,” said Anand Rao, global leader of AI at PwC. “It demonstrates how big a game changer AI is likely to be – transforming our lives as individuals, enterprises and as a society.”

What about Australia?

AI-led automation could also greatly boost Australia’s productivity and national income, according to AlphaBeta’s study – but only “if we get it right”.

“This opportunity will not land in Australia’s lap,” AlphaBeta says. “To unlock the benefits of automation we must be bold enough to lead changes. This means embracing technology’s potential to make our workplaces more productive, while taking steps to prevent Australia’s most vulnerable workers from sliding into unemployment.”

To unlock all potential gains, two conditions need to be fulfilled, the firm says. First, Australia needs a strong policy framework to ensure workers at risk of being displaced are redeployed. And second, governments must encourage more businesses to intensify their automation efforts.

However, Australian businesses have some catching up to do. “Only 9 per cent of Australia’s listed companies are making sustained investments in automation, compared with more than 20 per cent in the United States and nearly 14 per cent in leading automation nations globally,” the study says. “This low rate of investment in automation technology acts as a handbrake on our productivity growth that will ultimately reduce our national income.

Source: The Automation Advantage, AlphaBeta.

“If Australia accelerated its automation uptake, it would stand to gain up to another $1 trillion over the next 15 years. Both scenarios together – successfully moving all workers affected by automation into new employment ($1.2 trillion) and accelerating the rate of automation ($1 trillion) – represent a $2.2 trillion opportunity for Australia by 2030.”

Of these two goals, transitioning displaced workers would seem to be the most difficult and arguably the most important. Reskilling workers will be critical, according to Professor Stephen Martin, chief executive of the Committee for Economic Development of Australia at the time of CEDA’s 2015 report, Australia’s Future Workforce.

The CEDA report pointed to Denmark’s policy approach as a potential solution. “The Danish approach is three pronged – greater flexibility around hiring and firing, generous unemployment benefits and substantial programs to help unemployed people gain new skills. Often these programs start before a person is even retrenched,” Martin said.

“The Danish policy, while more expensive initially, makes long-term economic sense because it ensures people return to the workforce more quickly and with the skills the economy needs.”

Next: Which jobs are at risk?

When machines will outperform humans

Debates about AI capabilities often boil down to whether or not ‘Artificial General Intelligence’ (AGI) can be achieved. The definition of AGI is heavily debated, but primarily refers to a machine with the same intellectual capabilities as a human.

Also known as strong AI, such a technology would possess the complete range of human cognitive abilities, including self-awareness, sentience and consciousness. On the other hand, weak AI, which is what we have now, is non-sentient. Human brains and Weak AI can both compute, but only the former can understand what it computes. Many researchers doubt that AGI is even possible and without it, machines will never truly achieve genuine human intelligence.

It seems like the dystopia of The Matrix isn’t likely to become a reality anytime soon, so for the time being you can stop having nightmares about Agent Smith.

However, while machines are unlikely to become human-like any time soon, they will be able to outperform humans in many activities over the next ten years, according to When Will AI Exceed Human Performance, a global survey of AI experts conducted by the Future of Humanity Institute at the University of Oxford and the Department of Political Science at Yale University.

The study predicts that machines will soon be better at translating languages (by 2024), writing essays (by 2026), driving a truck (by 2027), working in retail (by 2031), writing a bestselling book (by 2049), and working as a surgeon (by 2053). It also predicts there’s a 50% chance of AI outperforming humans in all tasks within 45 years and of automating all human jobs by 2040.

Most affected industries

The industries most affected by AI will be retail, finance and healthcare, according to PwC’s Sizing the Prize.

In retail, the study predicts personalised design and production, consumer insights, and inventory and delivery management will be most impacted. Finance will see personalised financial planning, fraud detection and automated transactions, while healthcare changes will include AI diagnostics, pandemic identification, and imaging, such as radiology and pathology.

While those changes will have the biggest impact on GDP, other changes PwC forecasts include autonomous car technology for ride-sharing, delivery, and traffic control, as well as on-demand production in manufacturing and even AI content creation in films, music and marketing.

“No sector or business is in any way immune from the impact of AI,” said Gerard Verweij, global data & analytics leader at PwC.

“The impact on productivity alone could be competitively transformational and even disruptive. Businesses that fail to apply AI, could quickly find themselves being undercut on turnaround times as well as costs and experience, and may lose a significant amount of their market share as a result.”

How work will change

AlphaBeta’s study analysed how Australians spent a total of 20 billion work hours each year, assigning each of those hours to one of more than 2,000 different work tasks – observing how technology has affected the time spent on those tasks over a 15-year period and then projecting future change to 2030.

It then divided the 2,000 tasks into six “task groups”:

  1. Interpersonal tasks – engaging with other people, such as sales, training and teaching
  2. Creative and decision-making tasks – such as designing artwork, software development and strategic management
  3. Information synthesis – interpreting and extract meaning from simple data points, like an analyst
  4. Information analysis – gathering and processing data, such as a cashier calculating daily sales values
  5. Predictable physical tasks – repetitive and routine physical work, such as assembly line workers packaging equipment or agricultural workers picking fruit
  6. Unpredictable physical tasks – such as a mechanic repairing a car or a doctor performing surgery.

“Activities in the first three tasks groups – interpersonal , creative and decision-making, and information synthesis – are generally the least likely to be rapidly replaced by machines,” the study says. “However, activities in the last three tasks groups – information analysis, predictable physical and unpredictable physical – are expected to experience workplace change driven by automation in the near future.”

Source: The Automation Advantage, AlphaBeta.

AlphaBeta contends that these trends have “the potential to improve the work lives of every single Australian in a very tangible way.”

“Tasks lost to automation are typically the most dangerous, least enjoyable and the least likely to be associated with high pay,” the study says. “As automation shifts these dangerous, tedious and less valuable tasks from people to machines, work injuries are set to fall and work satisfaction levels bound to rise as workers can focus on more creative and interpersonal activities. Automation will make work safer, more meaningful and more valuable.”

In fact, the study says that workplace injuries will fall by 11% over the next 15 years as dangerous manual tasks are automated, 62% of low-skilled workers will experience improved satisfaction, and wages for what it considers “non-automatable” work are 20% higher than for “automatable” work.

In addition, it found that over the next 15 years, the average Australian worker will spend two hours per week less on manual and routine tasks – and more time on engaging with other people, creativity and decision-making, and interpreting or extracting meaning from data.

The jobs at most risk

AlphaBeta argues that 71% of the total expected reduction in work time will come from people doing the same job but completing fewer manual and routine tasks. However, that still leaves 29% likely to require workers to change jobs.

CEDA's Australia’s Future Workforce was even less optimistic. It found that five million jobs – almost 40% of the Australian workforce – had a moderate to high likelihood of disappearing in the next 10 to 15 years.

It should come as no surprise that those most affected will be jobs that are mostly manual and/or routine. According to AlphaBeta, those jobs most susceptible to automation are: construction and mining labourers, glaziers, plasterers, tilers, floor finishers and painters, food preparation assistants, cleaners and laundry workers.

Least susceptible to automation are: contract and project administrators, insurance agents, sales representatives, real estate agents, engineering professionals and ICT managers.

However, high-routine white-collar jobs are also at risk, according the International Bar Association’s study. For example, the study says: “There is a 98 per cent probability that the work of an accountant can be done by intelligent software.”

Want to know how susceptible your occupation is? The ABC has a great online app that allows you to easily search AlphaBeta’s data for your job.

Next: Who’s safe?

AI in action

Digging deeper into how AI is beginning to impact different sectors, many construction and manufacturing positions are obvious contenders for automation. Jobs in human transport are in danger from self-driving vehicles and storage services from automated warehouses like the one being used by Ocado Technologies. Even agriculture is feeling the heat as cow-milking robots and other machines pop up around the globe.

Bookkeeping and the routine parts of accountants’ jobs are under threat from rapidly improving accounting applications. Xero, for example, is introducing machine learning to automate tasks. Sage is aiming to provide “invisible accounting” and to completely eliminate admin for business owners and managers.

Xero is piloting a machine learning tool that will speed up invoice preparation.

Processes important to the ‘health’ part of healthcare are already being automated. Google DeepMind has already developed an app to help quickly diagnose kidney disorders in patients. More accurate diagnoses are not the only advantage of automated healthcare.

A lot of people believe that the ‘care’ part of healthcare isn’t as easily automated and plenty of IT experts anticipate job safety for careers requiring emotional intelligence (EQ). Ian Pearson, leading futurologist of Futurizon and fellow of the World Academy for Arts and Science, is one of those experts.

“AI and robots can automate intellectual and physical tasks, but they won’t be human, and some tasks require the worker to be human,” said Pearson in a Futurizon blog. “A human will always be able to identify with another human on an emotional level better than a robot can.” According to Pearson, humans aren’t only better in EQ fields like therapy, but patients also prefer humans over robots to fill these roles.

Research gathered at the University of Southern California's (USC) Institute for Creative Technologies suggests otherwise, however. Ellie, an AI-powered virtual therapist, is part of a virtual-reality program at USC called SimSensei. The program aims to treat people with depression and post-traumatic stress disorder (PTSD). Ellie is programmed to detect and analyse verbal and nonverbal cues such as body language, facial expressions and eye contact that are linked with these conditions, offering more accurate diagnoses of emotions and mental illnesses. Test subjects who talked to Ellie told researchers ‘her’ voice was comforting and that talking to a computer made them feel less judged than when they spoke with a human therapist.

In education, computerised tutors relying on AI technologies like those provided by Third Space Learning are proof that AI is already infiltrating education. With more students leaning towards online learning environments anyway, there may not be many education jobs left to steal. Distance learning saves money and offers students the flexibility to complete work whenever and wherever they want. Plus, it works better. A report from the SRI International for the Department of Education found that, on average, “students in online learning conditions performed better than those receiving face-to-face instruction”.

Still, there could be people who will always prefer a traditional classroom with a real teacher over a computer program. But what if students can’t tell the difference? Jill Watson, a teaching assistant at the Georgia Institute of Technology, taught a class of graduate students for 5 months without any of them realising she was a robot. Jill answered questions and assisted students with their projects through casual email responses that were indistinguishable from those sent by human TAs for the class.

Even positions based on the science, technology, engineering and maths (STEM) disciplines are in danger. The global survey report estimates that in roughly 88 years, AIs will be researching and developing themselves.

Nevertheless, the STEM disciplines are becoming increasingly important. Representatives of Google’s AI research unit, DeepMind, told an inquiry by the UK government's Science and Technology Committee: “[One of the] most important steps we must take is [ensuring] that current and future workforces are sufficiently skilled and well-versed in digital skills and technologies, particularly STEM subjects.”

Who's safe?

Remember the jobs that our parents told us not to pursue? The ones that seemed like they might actually be enjoyable? As it turns out, you should’ve just gone for it. Few argue that robots will replace human artists, actors, athletes or musicians.

UK non-profit Nesta is publishing a report that explains why. “Tasks which involve a high degree of human manipulation and human perception – subtle tasks – other things being equal will be more difficult to automate,” report co-author Hasan Bakhshi told Fortune.

Bakhshi identifies creativity as a bottleneck to automating work, meaning that jobs which require it “have a very high degree of resistance to automation.”

DeLaval’s robotic milking system.

There are already robots composing and playing their own music, but people don’t want to listen to it. It's fascinating that a robot can play chess well, but if that robot is unbeatable then it's not really a game. Sports and games are entertaining because they are innately competitive and unpredictable. If one robot can be programmed to paint like Picasso, then fifty robots can be programmed the same way. People are different -- there can only ever be one Freddie Mercury. AIs may be considered successful, but they will never be considered talented.

How do I avoid redundancy?

For those of us who aren’t athletes or artists, articles like this can be frustrating. There must be something we can do! People working in fields vulnerable to automation have two options.

You could be the type of person that accepts the inevitable, resenting technology so passionately that one day you intentionally run over your own smartphone with your car – or you can rise to the challenge and retrain yourself.

You can focus on factors that can’t be automated – creativity and uniqueness – and develop those traits as they relate to your job. Consider the rise in automation as an opportunity to explore what makes you uncloneable.

Although AI research may pose a threat to your job, it also has benefits. The majority of AI technology currently being used isn’t intended to replace workers, but rather to augment them.

Don Schuerman, chief technology officer of Pegasystems, explains the ways in which the influx of AI will change the nature of the workplace.

“AI will take the repetitive nature out of jobs – or take the ‘robot’ out of the human,” Schuerman said. “It will add intelligence and insight that employees otherwise wouldn’t have. It’s important to remember that the best chess player in the world is not a human or a computer: it’s a human and computer playing together.”

This feature includes AI-related content on the employment impact, Elon Musk, Stephen Hawking, Bill Gates and PwC’s report that originally appeared at IT Pro.

Multi page
Copyright © BIT (Business IT). All rights reserved.

Most Read Articles


What would you like to see more of on BiT?
How To's
Photo Galleries
View poll archive

Log In

  |  Forgot your password?