(Photo credit: Andrea De Santis/Unsplash)

Telemarketers will be the hardest hit by artificial intelligence (AI), and textile and garment workers—along with other professions involving manual labor—will be the least impacted. Those are the findings of a recent Princeton study examining more than 800 job classes measured against their durability to generative AI and its disruptive potential. The graph below ranks jobs based on their predicted level of exposure to language modeling, one of two capabilities of generative AI that can replace or augment human work—the other being image generation. In other words, generative AI (artificial intelligence that can generate content based on patterns it has seen in training data) can write sentences and create images.

Methodology

Language modeling and image generation were linked to and compared with 52 human abilities such as oral comprehension, oral expression, inductive reasoning, and arm-hand steadiness. These 52 abilities were then connected to more than 800 occupations using the Occupational Information Network (O*NET) database developed by the United States Department of Labor. Based on the closeness of these relationships, a score—the AI occupational exposure (AIOE) score—was assigned to each occupation indicating how much it will be affected by the language-producing and image-creating capabilities of generative AI.

In general, higher-earning, white-collar jobs were most exposed to the disruptive potential of generative AI, according to study authors Ed Felten (Princeton), Manav Raj (University of Pennsylvania), and Robert Seamans (New York University). Researchers also found that female, White, and Asian workers could be most affected by labour shifts triggered by generative AI based on the professions they held.

In terms of image generation (shown in the graph above), interior designers were the most at risk for labor disruption because of artificial intelligence; dancers, the least. “Many of the occupations most exposed to advances in image generation are occupations where spatial orientation is important,” write Felten, Raj, and Seamans.

“ChatGPT is the fastest growing consumer application in history, having reportedly reached 100 million active users within two months of its public launch.”

Ed Felten, Manav Raj, Robert Seamans

With the meteoric rise in popularity and attention paid to ChatGPT and generative AI—fueled by the media spotlight and clamor of tech insiders—a sense of anxiety is understandable. But many researchers have echoed the belief that exposure to generative AI doesn’t necessarily always mean job replacement. It may lead to increased efficiency and the enhancement of certain job functions, which can ultimately spark more economic growth. This in turn could lead to job creation. One estimate by Goldman Sachs indicates that global economic growth could increase by 7%.

In their study, The Potential Large Effects of Artificial Intelligence on Economic Growth, authors Joseph Briggs and Devesh Kodnani point out that innovation can spawn the creation of new job classes; for example, the internet created the need for web designers, developers, and network specialists. AI ethicists, AI policy experts, and AI-enhanced content creators may all become in-demand professions in the future.

Big changes to Google signal headwinds for online publishers

In recent weeks, a tech arms race between industry giants trying to carve out the largest or most valuable plot of real estate in the AI landscape has intensified. Companies are trying to leverage generative AI to improve their products and services to better meet the needs of digital consumers. One change by Google signals a significant shift that could dramatically affect online publishers.

“It will be like dropping a nuclear bomb on an online publishing industry that’s already struggling to survive.”

Matt Novak, Forbes

Essentially, Google is embedding Bard—the company’s answer to ChatGPT—into its browser. Now, when searching (“Googling”) a topic, the search engine will output the entire answer rather than providing links to publishers or content producers who offer the information. In effect, this is creating a bridge between user and information that bypasses content producers entirely. As reporter Matt Novak writes in Forbes, “It will be like dropping a nuclear bomb on an online publishing industry that’s already struggling to survive.”

What could this look like in practice? Let’s say you Google, “How to change a HEPA filter on your Dyson?” In addition to returning links to instructional blog posts or web pages published by Dyson showing you how to do this, Google would—using generative AI—output the answer directly to its browser so that you wouldn’t have to click on Dyson’s link. It would simply provide you with the answer so that you wouldn’t have to look any further. This could have significant implications for SEO-optimized content designed to answer questions based on key search terms and phrases.

Applying that same thinking to this blog post, I decided against titling it: “These are the jobs that will be affected by artificial intelligence.” Arguably, that is the more SEO-friendly headline; however, when the changes described above take effect, Google will simply present this list of jobs as a search result—likely with some explanatory copy—and there would be no need to read any further.

Implications

It’s hard to predict exactly what the outcomes of this change could be. But it’s safe to infer the following:

• Tutorial-style “how to” headlines and articles will receive less traffic.
• Informational content whose headlines are wholly written for SEO will lose traffic, e.g., “Best Places to Visit in Canada.” (Google can simply answers these questions outright.)
• Informational content in general will receive less traffic because text can be summarized and outputed by Google’s generative AI.
• As a consequence of the above, analysis and opinion-based content will become more valuable as readers look for something more than just information.
• Listicles will become irrelevant.
• Video and photos will become more important for enticing readers. I may read Google’s list of “Top 10 places to visit in Canada,” but if an article promises to show me beautiful photos of them, I’m still likely to read the original article.

Throughout history, technological advances have reshaped society. The steam engine, internet, smartphone—society’s march toward innovation is inexorable. The wrong approach would be to fear and resist these changes. Instead, we ought to consider their ethical implications and implement appropriate regulations that maximize the good while minimizing the harm. What if we could harness artificial intelligence to cure disease, end poverty, and tackle the pernicious influence of disinformation rather than fueling layoffs? Either way, a technological tidal wave is coming. Instead of heading to higher ground, we should be learning to surf—not just to survive, but thrive.