Increased investment in AI has lead to companies restructuring – opting for more educated junior employees and cutting out middle-management and senior positions.

AI adoption has sparked fears across that it’s adoption within industry will lead to major job losses, with lower-level positions expected to take the greatest hit.

“Most execs will take the opportunity to rethink how their organisations are structured and where the greatest savings and productivity gains will be had,” Theo Priestly, a futurist, commented.

He added: “This means that most low-level work can be automated using AI agents to complete simple to moderately complex tasks.”

But a research paper painted a much different picture occurring in the US among companies integrating AI into their systems.

The paper, titled Firm Investments in Artificial Intelligence Technologies and Changes in Workforce Composition investigated job listings and company AI investment profiles in the US, and a CV dataset of 535 million individuals across the globe, as well as wage and education data from the US, to determine how AI is truly re-shaping the workforce.

The paper notes how emerging technologies have transformed working modalities in the past: computer software and robots displaced low and medium-skilled workers, replacing human ‘calculators’ and many manufacturing jobs. Other technologies, like electricity, completely re-organised production processes.

Cloud technologies also transformed businesses and the way data was stored and managed, but even this technology holds no bearing to the impacts of AI.

“The major impact of cloud was enabling easier remote access, and while it changes the way systems are hosted and run, it doesn’t change how end users work with their systems,” Richard Marshall, PhD, Senior Analyst at The Skills Connection and a former Gartner analyst said.

He added: “AI, however, changes how the work is done. This is equivalent to the introduction of power tools to the construction industry – AI is revolutionising how the actual steps or work are done, not how you access existing tools.”

And this restructuring of the process appears to result in a restructuring of the workforce.

The research team, composed of Tania Babina (Columbia Business School), Anastassia Fedyk (University of California at Berkeley), Alex He (University of Maryland Business School), and James Hodson (AI for Good), found that with increased AI investment came an increase in structural changes.

Two of the most notable changes found were: hiring a more educated junior level workforce, and less emphasis on middle-management and senior level roles.

This would create a more educated and flatter workforce structure, which is in line with what AI requires.

As a predictive technology, AI can help in decision making – we have already seen its capabilities in recruitment and hiring processes, concerning as they may be.

Results

Firms which invested in AI tended to adopt a more flat structure, with less senior and middle management level roles, and a widening bottom-level.

The paper described the shift as fast, and likely to do with AI adoption, as this was  not an overall trend across US companies.

To explain the predicted increase in junior level roles, as well as the reduction in middle management, the paper claimed that “technologies that improve predication and decision-making, such as AI, will give lower-level workers more autonomy and require fewer managerial layers in firms.”

Further, the paper found that junior-level hires tended to be more educated: an increase of one standard deviation in AI investment was found in association with a 3.7% increase in college-educated workforce, a 2.9% increase in employees with a masters degree, followed by a 0.6% increase in those with doctoral degrees.

In correlation, the amount of employees with no college education dropped by 7.3%.

The researchers claim that the overall workforce does not reflect the same level of upskilling – the amount of workers with advanced degrees, masters and doctoral – is relatively flat compared to the amount of higher-educated employees in AI.

In addition, it seemed that technical skills were also in demand among junior-level employees, with an emphasis on STEM and IT/analytical degrees and skills.

The resulting workforce structure consisted of a well-educated, bottom-heavy foundation with a flattening hierarchical structure.

Recommended

• Cloud First ‘23 | The Whats, Whys, and Hows of Cloud Centres of Excellence
• Scots Programme to Upskill Ex-Military to Fill Cyber Skills Gap
• Comment: 5G Could Supercharge Economy – If We Unlock Its Potential

Can we expect to see these results in the UK, and across the world, as AI’s implementation continues?

“The roll out of AI is going to take decades – Bill Gates famously and accurately said that we overestimate the immediate impact of technology while underestimating the long-term impact,” Dr Marshall said.

“I sincerely hope that top-down structures will vanish, empowering all levels of staff to create more fulfilling jobs, but it will take a long time, and surely some other changes will come along. Just as there are people invested in commercial real estate loudly complaining that all their minions should be back in the office, there will be those who call for immediate return to their old power roles.

“This is an opportunity for true meritocracy to be put in place, for new corporate structures to evolve with increased flexibility. There are always insecure management that feel threatened by flexibility.”

And what does this mean for education levels across the board?

“Less educated graduates, or those leaving school before higher education will struggle to enter the workforce in many industries as a result of AI and automation removing those entry level roles and jobs,” Priestly predicts.

He adds: “It may mean that blue collar work becomes the only safe haven and job choice for many until general purpose robotics start to enter the world. What governments need to consider is how to support those willing to learn and to cope with the disruption.”

Dr Marshall does not expect the shift to a more educated workforce to continue to be so drastic, however.

“Education is always necessary, and the collapse of continuing education/training in the workforce is a major problem that has been building for the last few decades. Training has been seen as a cost by the accountants while it should be considered an investment. However I would not rush to the conclusion that the introduction of AI will change the people being employed to do the work.

“Introducing AI and managing the data requires new skills, but in most cases work will evolve to include AI just as we’ve all accepted the use of autocorrect. As with all kinds of technology changes some jobs will die out, others will appear. Beware companies using AI – and remote work – as an excuse to shed workers.”

Marshall also challenges that AI is the only culprit leading to a thinning of the managerial herd.

“I think we are in the middle of several huge changes in the structure of businesses and AI is only one of the factors,” Marshall said. “Remote work has exposed that there is a huge layer of management that does little except sit in meetings, and this is driving structural change.”

Marshall stops short of calling this shift in workplace structure a revolution, so far: “It is more similar to offshoring of manufacturing than to revolution.

“There will be people who struggle to cope with new arrangements, some comfortable people will be unquestionably disturbed, but there are also many opportunities to balance it out.”