You can hardly read or watch the news these days without hearing at least one story about how AI is coming for all of our jobs. The truth is obviously more nuanced, and I’m not sure we’re as close to this particular apocalypse as some people want us to believe, but it’s better to be prepared and all of that, right?

Some jobs will be first on the chopping block when the AI tech gets there, so if you’re curious about which ones, let’s sort it out.

A group of professors from New York University, Princeton, and the University of Pennsylvania wrote a paper (How Will Language Modelers like ChatGPT Affect Occupations and Industries) that’s going to help.

“We find that the top occupations exposed to language modeling include telemarketers and a variety of post-secondary teachers such as English language and literature, foreign language and literature, and history teachers. We find the top industries exposed to advances in language modeling are legal services and securities, commodities, and investments.”

This despite the fact that AI has been shown to do poorly when it comes to facts and also math.

🚨New Paper Alert!!🚨

Which occupations and industries are most exposed to advances in AI language modeling, as exemplified by #ChatGPT?

In a new paper with @EdFelten and Manav Raj, we use a framework to systematically address this question.https://t.co/SxbJUSd7wB

1/10

— Rob Seamans (@robseamans) March 6, 2023

They came to these conclusions by figuring out which job descriptions had already changed significantly due to advances in AI.

“The key insight was to map ten areas in which AI was advancing (imagine recognition, speech recognition, language modeling, abstract strategy games, et al) to 52 abilities used in occupations. The result was an AI Occupational Exposure score, which we call AIOE.”

Once we had that mapping, we could then use @USDOL O*NET’s information about the importance and prevalence of each ability in each of 800+ occupations to map from AI to each occupation. The result was an AI Occupational Exposure score – which we call AIOE.

5/10

— Rob Seamans (@robseamans) March 6, 2023

The higher the AIOE the greater the chance that automation takes over your industry.

The way these different occupations received high scores varied, with educators ranking high because of how language modeling AI could change the way they “assign work, detect cheating, or use programs to develop teaching materials.”

When it comes to telemarketers, though, it will be more about the bottom dollar than anything else.

“One might imagine that human telemarketers could benefit from language modeling being used to augment their work. For example, customer responses can be fed into a language modeling engine in real time and relevant, customer-specific prompts quickly fed to the telemarketer. Or, one might imagine that human telemarketers are substituted with language modeling enabled bots.”

The key insight was to map 10 areas in which AI was advancing (image recognition, speech recognition, language modeling, abstract strategy games, etc) to 52 abilities used in occupations.

4/10

— Rob Seamans (@robseamans) March 6, 2023

No word yet on how employers and programmers will collaborate to handle AI that doesn’t do its job well or to existing standards.

I guess we’re planning to just figure that out as we go.