The fourth Future of Jobs Report 2023 from the World Economic Forum (WEF) finds that the human-machine frontier has shifted, with businesses introducing automation into their operations at a slower rate than previously anticipated.
Organisations today predict only a 1 per cent increase in automation from the 2020 report in terms of physical and manual work. However, in the sectors involving reasoning, communicating and co-ordinating, artificial intelligence (AI) is expected to be adopted by nearly 75 per cent of surveyed companies and to lead to high churn – with 50 per cent of organisations expecting it to create job growth and 25 per cent expecting it to create job losses.
Erik O’Donovan, head of digital economy policy at Ibec, is not overly worried. He points out that humans have been automating work for at least 200 years. The great debate of technology – job replacement versus job creation – is ongoing.
“The WEF expects the impact of most technologies on jobs to be a net positive over the next five years. It predicts AI adoption will result in disruption but also net job creation,” says O’Donovan.
‘A gas emergency would quickly turn into an electricity emergency. It is low-risk, but high-consequence’
The secret to cooking a delicious, fuss free Christmas turkey? You just need a little help
How LEO Digital for Business is helping to boost small business competitiveness
‘I have to believe that this situation is not forever’: stress mounts in homeless parents and children living in claustrophobic one-room accommodation
The adoption of AI will impact skills needs. Trusted AI tools can be applied in an education setting to support teaching and learning. The WEF research expects that 42 per cent of companies it surveyed will be seeking to train staff in AI and big data.
“A recent report by Ireland’s Expert Group on Future Skills Needs (EGFSN) on the skills needed for Ireland to fully benefit from the opportunities presented by AI found it is not likely to bring about a net loss of jobs but it will replace certain tasks within many jobs over time‚” says O’Donovan.
“AI has the potential to bring substantial productivity increases. A broad range of skills is needed to ensure that this is realised in practice. For example, while AI developers and researchers may require specialist skills, anyone working in an organisation deploying AI will need an understanding of AI to work effectively with such systems or experts. Thus, there is a need for both organisations and individuals to prepare by seeking out the necessary education and training. The EGFSN report highlights numerous new courses in AI-specific skills, including industry-led programmes through Skillnet Ireland.”
Dave Feenan, director of Technology Ireland ICT Skillnet, sees this as an opportunity to upskill people.
I don’t see the headcount going down as much as new job classification and naming conventions
— Dave Feenan, Technology Ireland ICT Skillnet
“With every technology disruption there are some jobs that will come under pressure,” says Feenan. “Ultimately from a job classification point of view they will be replaced in the naming convention. Basically, job roles will be enhanced by the use of AI.
“An example might be proofreaders or translators where AI will undertake much of the heavy lifting but those original job titles might be upgraded to managers or directors. So, I don’t see the headcount going down as much as new job classification and naming conventions.”
Technical disruption is all about efficient enhancements – and what’s good for the goose can be good for the gander. Feenan sees AI being used by individuals, organisations and government agencies looking at AI as an enabler for best practices in society.
“However, there are good and bad actors out there looking to use AI to their advantage,” he says. “Good actors can learn from cyber infiltrations in the same way historically armies defending their walls see intrusions and repair the defending walls.”
Feenan sees examples of such real-life implementations across the world. He cites those of the French government using AI and Google Maps to find unauthorised buildings without planning permission, and the US government using sensors on municipal vehicles to collect details on unused buildings to be brought back into the housing stock.
Professor Cal Muckley, chair in operational risk, banking and finance at UCD College of Business, has been in the machine learning sector for many years. Today most of his research is about ethical AI, or trying to do good with AI.
“Oftentimes human decisions can be emulated impressively using machine learning techniques,” he says. “My work helps to identify fraud in bank transactions, deciding whether an applicant should receive a loan. Traditionally, these are heavily human expert oriented, but It’s impressive how effectively machine learning can replicate decisions based on the historical data.”
Muckley explains how, without human intervention, these decisions can become potential circles of discrimination. For example, if AI alone is used to analyse lead indicators in loan decisions, it might find concentrations of low-income families in an area. This geographical area might then be associated with long declines or higher premiums, so the AI decides to charge higher premiums based on area, not on the ability of the person to repay.
“This has the potential, with higher premiums, to feed into more defaults and a circle of impermissible discrimination against a geographical region,” says Muckley.
“Overall, I believe that the growth in jobs won’t be about data scientists – people like me; it will be about people who can manage projects actually related to AI, like the new loan managers.”
Cait Mulcahy is head of capability and automation at Three Ireland; she has found AI a boon for replacing repetitive tasks.
“Internally we would have had a highly skilled workforce who were doing a lot of copy and paste-type activities. it’s manual and repetitive. Accordingly, we use robotic process automation to automate those types of activities, allowing those individuals then to focus on higher value activities,” says Mulcahy.
“It allows our people to focus on solving problems, so it’s helping our data analysts and our network engineers plan, and allows us to be more proactive with our infrastructure.”