Anyone who has spent a few minutes playing around with ChatGPT, one will understand the worries and hopes that such technology generates when it comes to white-collar jobs. The chatbot is able to answer a wide variety of queries—from coding problems to legal conundrums to historical questions—with remarkable eloquence.
Assuming companies can overcome the problem that these models tend to “hallucinate” incorrect information, it’s not hard to imagine them replacing customer support agents, law clerks or history tutors. Such expectations are fueled by research and media reports that claim ChatGPT can get a passing grade on some law, medical and business exams. With companies like Microsoft, Slack, and Salesforce adding ChatGPT or similar AI tools to their products, we’re likely to see an impact on office life soon.
Several research papers published online this week suggest that ChatGPT and similar chatbots can be very disruptive, but not necessarily in the ways you might expect.
First, Edward Felton and his colleagues at Princeton University are trying to identify the professions most likely to be affected by ChatGPT. They used a test called Occupational Exposure to AI, which compares occupational tasks to the capabilities of various AI programs, to find out which jobs are most vulnerable to chatbots with good language skills.
The results suggest there could be big changes ahead for those in some occupations, including telemarketers, history teachers and social scientists, while people in more physical occupations such as bricklayers, dancers and textile workers may not need to worry about ChatGPT showing up at their workplace.
But a second study suggests that people in language-oriented jobs aren’t necessarily being replaced. Shaked Noah and Whitney Zhang, graduate students at MIT, looked at what happens when you put ChatGPT in the hands of office workers. They asked 444 college-educated professionals to complete a series of simple office tasks, including writing press releases and briefs, composing emails, and creating analysis plans. Half of them used ChatGPT.
The study found that people with access to a chatbot were able to complete tasks in 17 minutes, compared to an average of 27 minutes for those without a bot, and that the quality of their work improved significantly. Participants who used ChatGPT also reported higher job satisfaction. Although the study included asking experts to rate the quality of participants’ work, the paper does not say whether this included looking for the types of “hallucinatory” errors that might creep into the ChatGPT result.
These two studies hint at how things might play out, but they’re just early (and not yet peer-reviewed) attempts to figure out where ChatGPT is taking us. It is notoriously difficult to predict how new technologies will affect work, and economic studies related to ChatGPT are emerging quickly.
It is also ironic that textile workers have been identified as potentially immune to ChatGPT, as those concerned about the impact of artificial intelligence on work are sometimes called Luddites in reference to the nineteenth-century movement in which English textile workers broke their looms to protest automation.
In fact, according to some accounts, the Luddites were more concerned about who controlled automation than the existence of automation itself, directing their fury at employers who used automation to avoid paying workers fairly.
It might be a good idea for workers to take the initiative and start using ChatGPT to make themselves more productive. Just don’t tell my boss, okay? (Just kidding—WIRED just published a new generative AI policy that says we won’t publish AI-generated text unless it’s part of a story).
My first attempt at automating my own work was a false start. When I asked ChatGPT to find some links for this week’s newsletter, it suggested a bunch of stories from 2021, which makes sense when you remember that the AI ​​model was trained on data taken from the internet some time ago 🙄. Newsletter writers may not yet see a significant increase in their productivity.
