While Irish retailers still make a virtue in advertising showing staff carefully picking fresh produce from shelves in your local supermarket, British online delivery service Ocado sells a very different vision.
The online grocer recently closed what was its first “customer fulfilment centre” at Hatfield, north of London and replaced it with a “next-generation” centre in nearby Luton.
Cutting jobs was not the main aim, and the new tech did bring huge new gains in productivity. Thousands of robots whizz around a grid the size of six football pitches accessing about half of the 50,000 items in stock.
The machines can gather more than twice as many items per minute than was the case at Hatfield, where there was already a good deal of automation. The robots are overseen by a “hive mind” centralised computer employing artificial intelligence (AI), while a handful of humans in a control centre look out for glitches.
At the company’s centre in Erith, east London, which opened a few years back, the company reckoned an order of 50 items could be filled in less than 10 minutes, a considerable improvement on what people walking the aisles could do.
The company hopes to automate more of each product’s journey from farm or factory to fridge. It has struggled to turn a profit, but it believes plants such as Luton will help it get there. In the meantime it has sold its technology to businesses in the US, Japan, Canada and France.
On this side of the Irish Sea, trade unions are uncertain about what advances in AI will mean for jobs.
Brian McGann, head of organisational development at Siptu, cites Ocado as evidence of AI’s existing ability to drive dramatic changes in workplaces.
He sees huge potential benefits but like many with a professional interest in watching its development, he admits to being uncertain about what is coming down the line.
“We’re just not really sure where it’s all going to take us,” he says.
The unions are not burying their heads in the sand, he says, but there have been enormous advances in the technology in the past year. There is generative AI, which can produce text, images and audio, and applications such as ChatGPT, created by the Microsoft-funded OpenAI, which can be prompted in ordinary language rather than code. There is an inevitable concern among union members that these advances could accelerate the “disruption” to jobs, he believes.
He says that at a recent workshop for activists and officials there was concern that people with lower skills “might just be wiped away because AI is often seen as an opportunity to automate repetitive tasks”.
“There was a fear that people whose education level may not be so high could be the most disadvantaged,” says the Siptu official.
The unions are not alone in this concern. A survey of more than 1,200 lawyers, tax specialists and accountants from across three continents carried out by media group Thomson Reuters last summer found that a third of respondents thought the impact of AI on their professions would be “transformational” over the next five years. A similar number thought the level of change would be far more modest.
Two-thirds believed their skills would become more highly prized over the coming five years, but a third suggested their professions would be effectively wiped out in that time frame.
The study highlighted serious concerns over prices that can be charged for services in a world where clients can, or believe they can, get legal information from an AI system or have it draft documents for them.
Concerns were also raised about training younger workers or apprentices who undertake work that might be taken over by the computers.
Aidan Connolly, chief executive at Dublin-based Idiro Analytics, which advises on the use of AI, says many employers are still weighing up how the technology might change the way they do things.
A growing number are using AI for basic functions such as letter writing or blog posts while avoiding deeper use of the technology until glitches in big commercial systems are ironed out.
They are very conscious, says Connolly, of being left behind. He believes big changes are inevitable across a range of sectors in the coming years.
“I think there is going to be a lot of job disruption,” he says. “Irish companies compete with international ones, and if they’re not efficient, they’re going to lose.”
He points to areas such as coding, which newer AI applications are adept at, and customer call centres, where AI has been used for a long time, as obvious areas for increased deployment.
“The tech is only getting better and if you’re [a] software development company with a team of 100, I don’t see why that same team wouldn’t be 50 in two years’ time,” he says.
He says companies that set up call centres overseas with “a couple of hundred people” could move them back to Ireland and run them with new AI technology and a team of 10.
Many firms will retain jobs while doing new or better things, he says, but overall he is pessimistic.
John McCarthy, chief economist at the Department of Finance, recently warned that substantial Government investment to mitigate large-scale job losses caused by AI would be necessary, citing the technology as a “severe” vulnerability to the Irish economy.
Erik O’Donovan, head of digital policy at employers’ group Ibec, says much of the existing research suggests AI will create at least as many new jobs as it makes redundant, as has been the case with previous revolutions in the workplace.
Ibec echoes McCarthy’s view that substantial investment will be required to help businesses and employees adapt so existing jobs can evolve with the employment landscape.
“It’s a net positive story,” he says “but it’s a scenario where you can’t be complacent; you have to manage that evolution by supporting skills and supporting companies.”
He suggests the €1.5 billion surplus in the State’s National Training Fund could be used to retrain businesses and employees in the skills they need, and an accelerator fund could be used to invest in innovation.
The sort of organisational evolution needed may be close to what German software giant SAP announced last week. Analysts suggested that after 8,000 jobs were cut across the company’s global operations, the company would be left with an unchanged headcount at the end of this year with significant numbers of staff retrained for new roles, many related to AI.
In an attempt to provide expertise to its various departments, the Government recently assembled an advisory group comprising academics, entrepreneurs and others, that has held its first meetings in recent weeks.
Dr Patricia Scanlon, the chairwoman of the group, who sold her AI start up SoapBox Labs last year, says the committee’s members have been considering how they might best work together but suggests there is plenty of work to be done given the scale of what is happening and the uncertainty of where it might lead.
Among the issues the group is considering are AI’s impact on employment, the ethics around its use, the application of EU legislation and whether there is any requirement for local regulation.
“The real goal here is to take advantage of AI but mitigate the risks,” she says. “In some companies this will all mean lay-offs and in others there will be more innovation, better customer service or new opportunities.”
“But every boardroom, every executive office, every manager, every employee is thinking about this – it is going to have a huge impact. We need to have a conversation about that.”