Raphael ThysThe adoption of AI technologies won't replace everyone's jobs. But some workers could be most at risk. (Left) VioletaStoimenova via Getty Images, (Right) GoodLifeStudio via Getty Images
- The adoption of AI technologies could make some workers more productive and replace others.
- A White House report found that 20% of Americans had jobs that were particularly likely to be impacted by AI.
- Lower-income workers without a college degree could be at the most risk of AI job replacement.
Nothing is certain about the future of AI technologies, but three things are becoming more clear.
1. AI is likely going to impact or change millions of jobs in the years to come.
2. It's likely going to make some workers more productive.
3. It's likely going to replace some workers.
The million-dollar question is, how many workers — and what jobs — are most likely to see the positive and negative impacts of AI?
Twenty percent of Americans worked in "high exposure" jobs that were most likely to be impacted by the adoption of AI technologies, according to a White House report released in March by the Council of Economic Advisors.
To be sure, the authors weren't suggesting that 20% of Americans were likely to see their jobs replaced by AI. That's because AI-related job changes could impact one job task but leave all others untouched. Additionally, these changes could bring about positive outcomes for workers.
That said, if and when some AI job replacement does come, some workers could be more at risk than others.
"Some workers typically benefit from technological change, either because the evolving technology provides new labor market opportunities for them or because it enhances their productivity in their current job," the council wrote. "Conversely, some are harmed, typically due to job displacement."
The council's findings highlight what could be a reality of the coming AI boom: It'll be good for some workers and bad for others. While the adoption of AI technologies could help some workers become more productive, spend less time on boring tasks, earn higher wages, and even have a four-day workweek, others could face more competition, earn lower wages, or even see AI replace their jobs.
Lower-wage jobs could be at the most risk of being replaced by AI
The report didn't highlight specific jobs or industries that are most likely to be negatively impacted by AI.
But the council wrote that workers at the highest risk of replacement fit two criteria: Their jobs were highly exposed to AI and had lower "performance requirements" — meaning their job tasks had a lower degree of "difficulty or complexity" that may be more likely to be automated.
Ten percent of US workers fit both criteria, the council found. They tended to be lower-income workers without a college degree.
Related stories
The findings "suggest that AI may be a skill-biased technology, increasing relative demand for workers with high levels of education in high-earning occupations," the council wrote. "They also suggest that AI could exacerbate aggregate income inequality if it substitutes for employment in lower-wage jobs and complements higher-wage jobs."
Perhaps counterintuitively, the council found that the cohort of workers whose jobs had among the highest exposure to AI — people with a bachelor's degree — were the least likely to have a high-exposure job with low-performance requirements, the combination of roles at most risk of replacement.
Twenty-one percent of people with a bachelor's degree worked in jobs that are highly exposed to AI, the council found. However, only 6% worked in high-exposure jobs that also have low performance requirements. Conversely, 17% of high school graduates had jobs with high AI exposure, and 14% had jobs with both high AI exposure and low performance requirements.
This suggests that college-educated workers could be more likely to see the productivity benefits of AI, rather than see their jobs replaced by these technologies, the council theorized.
Meanwhile, 20% of women had jobs that are highly exposed to AI, compared to 19% of men. Twelve percent of women had jobs that fit both the high exposure and low performance requirement criteria, compared to 9% of men.
The world isn't black-and-white, so it's unlikely AI's impacts on workers will be either. That means it's unlikely that every worker whose job is changed due to AI will be able to be clearly placed in a "good change" or "bad change" bucket.
The council cited the hypothetical example of a school bus that can drive itself, which, in theory, could render a bus driver's job obsolete. In this scenario, the bus would likely need an adult on the bus to supervise the children.
"AI-led automation might fundamentally change the school bus driver's job, but it is unlikely to eliminate the job," the council wrote.
In this example, the bus driver gets to keep their job — a good outcome — but the nature of their job has changed substantially.
Major economic changes tend to help some workers and hurt others
The AI boom wouldn't be the first major shift in the global economy to have divergent impacts on workers.
When China emerged as a major player in global trade in the 1980s, some economists argued that the deluge of low-cost products was a net positive for the US, even though some domestic manufacturing jobs were lost in the process.
"The people who lost their jobs lose money to the China shock, but the rest of us get cheap goods at Walmart and Target or whatever," Nobel Memorial Prize-winning economist Angus Deaton previously told Business Insider. "And the theorem says the value of what we gain is more than the value of what they lose."
Deaton said he's grown increasingly uncertain that this trade-off has been worth it.
It remains to be seen whether the tradeoffs that come from the AI boom will be a net positive for Americans. Re-training workers who lose their jobs due to AI could help move the needle in a more positive direction.
But as Deaton's globalization example illustrates, the US doesn't have a great track record of helping displaced workers find new jobs.
Has your job been impacted by AI technologies for better or worse? Are you willing to share your story? If so, reach out to this reporter at jzinkula@businessinsider.com.
Axel Springer, Business Insider's parent company, has a global deal to allow OpenAI to train its models on its media brands' reporting.