'We are not valuing expertise or quality, but now we are valuing efficiency': Canadian AI researcher on the loss of hands-on training and mentorship
Artificial intelligence (AI) tools are not only changing how work gets done, some experts warn they are also upending traditional hierarchies and how employees advance their careers.
“Some young workers I spoke to are using their experience with A.I. to vault themselves ahead of more senior colleagues,” wrote New York Times columnist Kevin Roose, in a piece examining how AI is affecting the early-career job market.
Even while junior employees experiment with AI, organizations that cut entry-level roles risk losing more than just early-career talent; as Dilan Eren, professor of strategy at Ivey Business School, explains, without opportunities for hands-on work and senior mentorship, future employees may miss the chance to develop critical skills—potentially weakening the organization’s internal training pipeline and ultimately, its future success.
“I do think that in this new world of work, expertise-based hierarchies, they will get an update,” says Eren.
“All of a sudden, now we are not valuing expertise or quality, but now we are valuing efficiency … this could potentially create a lot of problems within organizations.”
The catch-22 of AI implementation
This situation presents a conundrum to employers, Eren points out, as while it is important to be cautious when implementing AI into workflows, at the same time it is essential for companies to keep up with technology to stay competitive.
To address this catch-22, she suggests a more balanced approach to workforce adaptation, rather than relying on entry-level job cuts to cut costs. This way, senior level expertise and early career enthusiasm can work together instead of against each other.
“We cannot actually lose sight of the importance of expertise,” Eren says.
“But on the other hand, we cannot really afford not keeping up with technological change, right? Which means efforts have to be double-sided here. We need to make sure that juniors actually still develop expertise, but also seniors develop AI skills.”
Anthropic CEO’s grave warnings about AI job losses
In a much-talked-about May 28 Axios column – dramatically titled “Behind the Curtain: A white-collar bloodbath” – Anthropic CEO Dario Amodei warned that AI could eliminate up to 50 per cent of entry-level white-collar jobs within five years.
He pointed to CEO apathy as a large part of the problem, telling Axios, “Most of them are unaware that this is about to happen … It sounds crazy, and people just don't believe it.”
For Eren, this avoidance is going to lead to big problems down the road; replacing junior roles with AI tools may seem like a win in the short term, but she warns that doing so eliminates a crucial element of workplace development.
“What happens when these people get retired in the next ten years, 20 years? ‘I haven’t hired any junior people, I haven’t trained them, so what is going to happen?’” Eren says.
“This will create problems in the long run, even within the same organization. For now, it is cutting costs, but in a few years, maybe it will be even more expensive for them to find seniors, and more experts that can actually do the job.”
Traditionally, entry-level workers learn from senior colleagues by working alongside them – without that exposure, Eren posits, the development of future experts is at risk.
Amodei also addressed this concern, telling Axios, “Even those who are optimistic AI will unleash unthinkable cures and unimaginable economic growth fear dangerous short-term pain – and a possible job bloodbath.”
False efficiencies and missing mentorship
Not everyone agrees with Amodei’s grim vision – Ravi Kumar, CEO of Cognizant, told Business Insider (BI) that he believes the opposite is true: that AI will create more opportunities for recent graduates, not less.
"More things will be reimagined, and more human labor will be needed to supplement the AI digital labor," Kumar told BI.
"So this whole paradigm opens up more embrace of software, because you're doing more for less, and when you do more for less, the adoption of software is going to go up."
The image of universal AI adoption may seem appealing to some, but it comes with complications, Eren says. If senior workers begin using AI to complete even the most basic tasks, there may be no work left to assign to juniors – and no mentorship opportunities left either.
“The problem here is that I think we should be careful what we wish for,” says Eren.
“If I as a senior, let's say I start to use AI, which means I am working on my own, I am not training anyone else. What happens when I retire? How am I going to make sure that actually juniors are not only using AI tools … but also learning the really important thing about our jobs – the expertise.”
That concern mirrors findings in the New York Times, which noted that some firms now only hire mid-level employees or above. One executive cited in the article said his company had stopped hiring below the L5 software engineer level because lower-level tasks were handled entirely by AI.
Building hybrid talent strategies
To mitigate risks of over-reliance on AI at the cost of expertise, Eren recommends a dual development approach, emphasizing that to be effective it must be accompanied by an intentional culture of collaboration.
“We need to make sure that juniors actually still develop expertise, but also seniors develop AI skills,” she says.
“We are not enemies. We are not replacing one another – seniors, juniors. We are actually here to find a new way to work together collaboratively, and make sure that we are learning from each other.”
Employers also need to be wary of overestimating AI’s long-term capacity to replace human judgment; Eren points out that for HR professionals tasked with workforce planning, the key may lie in finding ways to blend hierarchical expertise with the efficiencies of AI.
Failing to do so could be “an exponentially bad move,” warns Eren.
“We really need to think about the quality of the solutions we are getting,” she says.
“It is not true that you don’t need these people ... you are just delegating some of the tasks to AI.”
