The one condition keeping AI from taking our jobs


Replacement of workers by AI will only happen once a recession strikes, according to the IMF.

In the last few years, there has been a narrative that AI is coming to take everyone's job. As the hype dips, the ROI in many expensive AI projects is yet to materialize. Thankfully, many bosses realize their human workforce delivers much more than regurgitating information or filling documents with word salad.

Business is booming during the so-called AI gold rush, but could this change during an economic downturn?

ADVERTISEMENT

Why recessions could trigger a wave of AI-driven job losses

Earlier this year, Kristalina Georgieva, Managing Director at the International Monetary Fund (IMF), predicted that AI would impact 60% of jobs in advanced economies and 40% worldwide. We can all agree that AI is here to stay. Although we have seen job cuts, thankfully, they have not been on the scale that many have predicted.

However, the IMF's Gita Gopinath recently told attendees at the Fortune Global Forum in New York that "the replacement of workers by AI won't happen until a recession strikes." She warned that companies will hold on to staff members during the good times when profits are on the rise, but that changes when a recession hits, forcing the company to lay off workers.

"Just because you're not seeing it now doesn't mean you won't see an abrupt shift in a recession." Gita Gopinath.

Gopinath warns that a recession will increase AI adoption, and this is the moment when automation replaces jobs. When everything returns to normal, a so-called jobless recovery means many roles will be retired forever. But this is not the first time that Gopinath has delivered this warning.

Earlier this year, at the AI summit in Switzerland, Gopinath warned that while many focused solely on the risks of AI around misinformation, privacy, and security, they were ignoring how AI could amplify the next recession. Any spike in long-term unemployment in traditional industries would cause many displaced workers to struggle to acquire the skills needed to thrive in an economy increasingly shaped by AI.

AI's hidden price tag on workers' resources and the environment

ADVERTISEMENT

Every tech conference will feature a keynote on how AI augments human capabilities rather than replacing them. The AI co-pilot is sold as a solution that will work alongside your employees, but there is no hiding from the fact that AI is expensive.

Microsoft has invested $13 billion in OpenAI, proving expensive in more ways than one. Despite declaring it would be carbon-negative by 2030, AI has helped push Microsoft's emissions up by 40%.

Microsoft's data centers are expected to consume over 50 million gallons of drinking water every year, and even a quick ChatGPT prompt uses 10 times more electricity than Googling something. But who is paying the price for big businesses' obsession with all things AI?

Last year, Microsoft revealed that with the arrival of AI, they were focusing on short- and long-term opportunities. However, this new focus reduced its workforce by 10,000 jobs. After axing over 27,000 jobs, Amazon declared it would continue to maximize its resources and efforts in generative AI.

Similarly, Cisco started the year laying off 4,000 employees, followed by a second round of layoffs that included the axing of nearly 6,000 workers. The company went on to announce a $1B global investment fund to expand and develop secure, reliable, and trustworthy AI solutions.

A recent study also revealed three in ten companies are blowing budgets on massive AI investments while failing to upskill their employees. At the same time, Gartner is warming enterprises that 80% of the software engineering workforce must upskill to meet the demand for new roles created by generative AI.

AI ambitions come at a massive cost, but who is paying? In the last few years, workforces have been sacrificed so enterprises can increase their spending on and invest in all things AI. For many, this ongoing threat to their livelihoods is pushing them towards creating their own opportunities.

The recession and environmental risks of an AI-dominated future

While AI might not be the direct cause of an economic crisis and massive job losses, it could exacerbate the problem. It's important to highlight how Gopinath outlines a significant risk that seems missing from the broader conversation rather than making a prediction. But many businesses are arguably flirting with this scenario already.

ADVERTISEMENT

Due to industry transformation, 23% of global jobs are expected to change over the next few years. AI ambitions come at a massive cost, but who is paying?

In the last few years, workforces have been sacrificed so enterprises can increase their spending on and invest in AI. Despite AI's well-documented environmental impact, many enterprises have pushed back their climate commitments with future promises that nuclear energy will solve the problem in 5-10 years.

Microsoft is reviving the infamous Three Mile Island nuclear plant to power its increasing AI needs. Google and Amazon are also plotting a similar path toward atomic power. Despite environmental concerns, many in the tech community are quick to dismiss the fearmongering that follows any suggestion of going nuclear.

While AI continues to reshape industries and promises to augment human capabilities, the economic and environmental consequences cannot be ignored. Gita Gopinath's stark warning at the Fortune Global Forum underscores a critical reflection point.

The real impact of AI on jobs could only fully manifest once an economic downturn occurs. This potential for a significant shift during a recession highlights the need for businesses to consider more sustainable and employee-friendly approaches to AI integration.

Enterprises appear happy to ignore environmental concerns and lay off workers during the race toward the AI dream. The IMF warning, combined with going all in on nuclear power to further fuel AI ambitions, could lead us toward an uncertain future that benefits the few rather than the many.