The jobs at risk from ChatGPT
Skip to content
The Insightful Leader Live: How to Talk about Your Work (and Yourself) So People Listen | Register
The Insightful Leader Logo The Insightful Leader Sent to subscribers on February 1, 2023
The jobs at risk from ChatGPT

It’s hard to avoid news stories about ChatGPT and similar AI-driven chatbots these days, particularly think pieces analyzing whether the robots are coming for our jobs.

Opinions range from insisting that the takeover is imminent to “eh, there’s not too much to worry about.” So I reached out to an assistant professor of finance, Bryan Seegmiller, who has studied the impact of disruptive technology on both blue- and white-collar jobs, for his take.

We’ll hear from him today. And we’ll also get some thoughts on the state of the economy from professor Sergio Rebelo.

Which jobs will be most disrupted by ChatGPT?

Seegmiller, along with finance professor Dimitris Papanikolaou and colleagues, conducted recent research that looked at which workers have historically been most likely to lose jobs or have wages reduced when new technologies emerge.

Not surprisingly, they found that manual laborers had the highest exposure to emerging technologies, especially from 1850 to 1970. But other patterns were more surprising. In the 1970s, occupations in which people performed routine “cognitive” tasks, such as clerks, technicians, and programmers, also began to face much larger exposures to technology. And when new inventions showed up, workers who earned the highest salaries within the affected occupations—that is, those with the most advanced skills—saw the biggest slowdowns in their wages. Additionally, older workers’ wages generally took a bigger hit, too.

So I asked Seegmiller what he thought the impact of ChatGPT and similar bots would be on different types of jobs. Here’s what he had to say:

Seegmiller:
ChatGPT and other types of AI are excellent at analyzing very large datasets to detect patterns and make predictions. For predictions to be good, the patterns AI detects must hold up outside of the sample data it was trained on. Because of this, the workers who are currently the most at risk of displacement by an AI technology like ChatGPT are those who specialize in retrieving or analyzing lots of information, but usually without requiring much judgment or interpretation. This may include some jobs specializing in data analysis.

So where is the room for humans? Current AI seems to do poorly in making decisions or predictions when data are scarce; rules are fluid or changing; or creativity and new ways of thinking are needed. These are all comparative advantages of humans, because while AI spits out predictions based on past relationships in data, humans have a much better ability to make judgments based on small amounts of information or to apply lessons from things they’ve learned to new, yet-unforeseen situations. This underscores the need for workers to develop their critical-thinking skills rather than to just spit out memorized information. This is easier said than done, of course—according to my own research, workers who are really good at tasks displaced by AI, or who are older, are likely to find it hard to adapt to these new skill demands of an AI-intensive economy.

Finally, it’s important to note that ChatGPT and other AI are far from being able to replicate human-to-human interaction. The value of people skills, including working in teams, communicating clearly, etc., is only going to increase in importance going forward.

The latest on inflation and the economy

If you were hoping that by now, three years since the start of Covid, we’d have a clear sense of the economy’s trajectory, then you have surely been disappointed. Are we headed toward a recession? Will grocery-store prices ever fall back to earth? A lot of questions remain. But, thankfully, finance professor Sergio Rebelo is always willing to share his perspective on the latest economic indicators, as he did in a recent The Insightful Leader Live webinar.

He offered up six takeaways on the state of the economy right now. Here are a couple:

Worker shortages are unlikely to be a short-term problem: The rebounding economy is now constrained by long-term demographic trends. Thanks to decades of lower birth rates and slowing worker migration, the population of active workers aged 25–54 has plateaued in developed economies. “Fertility rates are currently below the replacement rate, so we are sowing the seeds of a future labor shortage,” Rebelo says. And while before 2016, the U.S. made up for declining fertility rates with immigration, “that channel has been greatly reduced,” dropping from about a million immigrants a year to just 200,000.

Don’t expect prices to come down anytime soon, either: The retail prices for food and energy are not coming down as quickly as falling commodity prices would indicate. That’s due to the so-called “rockets and feathers” phenomenon, Rebelo explains. Retailers are quick to raise prices when their costs go up but act much more slowly when those costs come down.

You can read the full recap of his talk here.

“If a team fails, good leaders will say, ‘okay, team, let’s look at what happened. Let’s identify the issue.’ It should never be, ‘oh, you guys suck. You’re the team that failed me and that’s why I got to come in and do all the work.’” 

— Colonel Fredric Maddox, an assistant professor at the U.S. Army War College and senior fellow at Kellogg, on The Insightful Leader podcast, on how letting teams fail can help them succeed.