Follow ZDNET: Add us as a preferred source on Google.
ZDNET’s key takeaways
- MIT’s “Project Iceberg” maps out the economic impact of AI.
- Companies overlook AI’s impact on 9.5% of the US workforce.
- AI’s disruptive potential will vary across industries and jobs.
When considering the specter of human workers losing their jobs to AI, most people will probably think of software engineers and other tech industry professionals in coastal cities like San Francisco and New York. But what if those roles were just the tip of the iceberg?
Also: AI could double the US economy’s growth rate over the next decade, says Anthropic
Massachusetts Institute of Technology (MIT) published a study last week indicating that, indeed, the scale and distribution of AI-fueled automation is likely to be much more extensive than previously believed. Titled “Project Iceberg,” the study found that current AI systems can already replace 11.7% of the US workforce, representing around $1.2 trillion of the total value of the country’s labor market (which is worth more than $9.4 trillion total, according to the report).
Key findings
The study was based on a large-scale computer simulation of the US labor force powered by the Frontier supercomputer at Oak Ridge National Laboratory in Tennessee. The researchers leveraged what they call “Large Population Models” — AI systems that can represent the interplay of 151 million workers across 3,000 US counties, and exhibiting 32,000 individual workplace skills.
The goal was to develop a detailed map of the particular jobs and industries AI can handle. This also provided the researchers with a snapshot of the actual economic output that can be generated by AI — a matter of supreme significance to tech investors and state officials amid ongoing tremors about a possible AI bubble that could send shockwaves throughout the national economy.
Also: AI buzzwords are wrecking your job search – here’s what really matters
All told, the study found that existing AI systems can automate around 11.7% of the total US workforce, while current adoption strategies focus on a relatively tiny number of tech jobs, cumulatively comprising only around 2.2% of the workforce.
“This fivefold exposure difference is geographically distributed nationwide rather than concentrated in coastal hubs, indicating that workforce preparation strategies based on visible technology-sector signals may substantially undercount transformation potential,” the authors wrote.
It won’t just be tech jobs; across all 50 states, roles that blend rote data-processing functions with human interaction — such as HR, finance, and office administration — will feel the effects of AI automation, the authors added.
Also: How to learn ChatGPT in under an hour using my favorite guides and videos – for free
Just as most of the mass of an iceberg is hidden from view beneath the waterline, the study found that the majority of workplace tasks that will be automated by AI are being excluded from mainstream discourse, which thus far has mainly focused on tech jobs. This, in turn, is creating a warped picture of what government officials and company executives should expect in the coming years.
Location is everything
The report also emphasizes a region-specific approach to preparing workers for the coming wave of automation.
For most businesses across the country, it won’t be enough to simply copy the AI strategies deployed by tech companies — whether that’s layoffs, upskilling, or a combination of the two. Rust Belt states like Ohio and Michigan, for example, have a modest tech presence but many white-collar manufacturing jobs compared to the rest of the country, including financial analysis and administrative coordination, according to the report.
Also: 3 ways AI agents will make your job unrecognizable in the next few years
Mainstream narratives about the potential for job disruption in these states (and others) are misleading, the researchers contend, in that they overemphasize the tech industry and overlook the many other industries that will be transformed by AI.
“States with small technology sectors underestimate the scale of their exposure, leaving them vulnerable when adoption accelerates in white-collar work,” the authors wrote.
A call to action
The MIT study is more than just a state-of-the-union report about AI’s potential for job displacement — it’s also intended to be a virtual sandbox in which concerned parties can analyze how their particular regions will be affected. After all, no two states, counties, or cities have identical labor markets; each is characterized by a unique fabric of industries and jobs that support their individual economic spheres. Understanding if and how AI tools can support or replace the workers in specific locations, therefore, will be an important step toward understanding the technology’s impact on a larger scale.
Also: Is AI a career killer? Not if you have these skills, McKinsey research shows
“By simulating how capabilities may spread under alternative scenarios, Project Iceberg enables policymakers and business leaders to identify exposure hotspots, prioritize training and infrastructure investments, and test interventions before committing billions to implementation,” the researchers wrote in their report, which was also coauthored by officials from the North Carolina General Assembly and the Utah Office of AI Policy, and other organizations.
In an AI “action plan” published last month, Tennessee state officials wrote that they would use Project Iceberg “to gain a deeper understanding of the impact of AI on the Tennessee workforce.” Utah policymakers are planning to follow suit, according to CNBC.
AI and the future of jobs
The MIT study arrives on the heels of other efforts to assess the fine-grained impacts of AI upon the labor market, zooming in from an industry-focused view to gauge the degree to which the technology can handle specific workplace tasks.
A new study from AI startup Anthropic, for example, found that people who used the company’s Claude chatbot for assistance completed various tasks about 80% faster on average than those who didn’t. The company concluded that those findings could cause the growth rate of the US economy to double over the next 10 years.
Also: AI may take jobs, but it makes starting a business easier than ever – here’s how
Research from job search site Indeed published in September, moreover, shows that AI is more likely to transform jobs rather than replace them wholesale.
The big takeaway from these recent findings, including the new MIT study, is that AI-powered automation will not take place along neat and predictable lines; rather, it will impact individual roles in unique ways, depending on a multitude of factors. This obviously complicates the picture for employers, since it basically negates the utility of a one-size-fits-all approach to AI upskilling for employees.
This varied ripple effect throughout the labor market, as market research firm Gartner predicted last month, will cause “jobs chaos” in the coming years as business leaders race to come to grips with the changes affected by the technology within their particular workplaces.
Comments are closed.