Artificial Intelligence (AI) looks set to bring huge change to workplaces, including in medicine, the law and professional services, a leading expert in the field says.
Jonathan Ruane says the impact of AI is already being felt broadly: the number of job postings in the US for software developers has declined 60 per cent since an explosion in AI services a couple of years ago.
“I don’t want to suggest this is all down to ChatGPT,” he says, referring to the generative artificial intelligence chatbot launched in 2022.
“We don’t know all of the reasons, but it is such a stark number that, combined the other available evidence, it gives a sense of how quickly things are changing.”
Sligo native Ruane cofounded courses at the Massachusetts Institute of Technology Management School in the areas of AI and Robotics and is also an adjunct professor at Trinity College Dublin.
He was previously chief executive of a tech start-up and recently chaired the employment subcommittee of the Government-appointed AI Advisory Council.
The council has recommended, among other things, the establishment of a national AI Observatory to monitor the impact the technology in workplaces.
This is required, he says, in part because AI’s impact is in many cases so incremental as to be hard to monitor in a way that informs the debate around it.
“It’s better not to think about this in terms of roles, but rather in terms of tasks. Our jobs all involved a number of tasks and if one of those tasks is automated it really doesn’t affect us, we just sort of adapt,” he says.
“But if seven, eight, nine of those tasks are automated, then things are going to shift in your occupation and hiring might well change.”
Ruane, who is also a member of the RTÉ board, says the precise scale and nature of the impact is impossible to predict.
“It’s happening so fast it’s incredible. And we get corporations, or their CEOs, coming to us and asking the question, ‘tell me where this is going to be in five years so I can work back?’” he says.
“What we’re consistently saying to them is that that’s not an option right now. You have got to take a more dynamic approach, aim to run fast and still be able to change direction.”
He says the process of change would be slowed in many instances by the caution of decision makers or organisations as well as by regulation. But he suggests the latter could, ultimately, become a driving force for change in sectors such as medicine where AI can already outperform many doctors at particular tasks.
What future applications can you see with artificial intelligence (AI)?
— The Nobel Prize (@NobelPrize) March 21, 2025
Hear 2013 medicine laureate Randy Schekman speak about the power of AI and future applications of the tool.
Schekman identified three classes of genes that control different facets of the cell's… pic.twitter.com/bzUh27BXLd
He cites the example of software widely in use that allows consultations with patients to be recorded so referral letters can be generated automatically.
“Why would we not use a system that listens to that conversation and tries to find errors in the doctor’s diagnosis because that is technologically possible,” he says.
Research suggests the technology outperforms all but the very best doctors, he continues, and “in medicine, we care about the average. So at what is it ethical that you would be a consultant and not use this stuff?”
A key problem in many sectors, he suggests, was the development and retention of really high-level skills that would still be required in various areas.
Some surgical procedures, he adds, have been shown to have better outcomes when performed by robotics.
“Here’s the problem, though: today, the way surgeons are trained, you come up and you do this, call it the manual version, the scalpel version, and then you become the robotic person. That makes sense,” he says.
“But what about the one time in 1,000 the robotic procedure goes wrong in the future?”
Referencing an existing procedure – the removal of kidneys – already performed using robotics, he says “that surgeon has three to five minutes to walk over to the machine, pull all the devices out and use his or her best scalpel skills to save your life. Now, what if that surgeon hasn’t used a scalpel in several years?”
[ New Bill targets AI ‘deepfakes’ and identity ‘hijacking’Opens in new window ]
A different problem may be faced by the legal and professional services sectors where thousands of junior staff are hired to perform more mundane tasks from which they learn,” he adds.
“Imagine a world [of accounting] in which an AI looks at the task of checking invoices, which is currently done by a junior, but it doesn’t check some invoices.
“It checks all the invoices. It doesn’t check some processes. It checks all the processes. What do we need the junior for?” he asks.
“But the question then is, how does anybody ever become the senior if they haven’t been trained up to be the intermediate skills? I think you’re going to see a lot of this, the issue of skills acquisition or atrophy. And I think a lot of industries are going to be affected.”
