Economic impacts research at OpenAI

OpenAI has opened a call for expressions of interest from researchers who want to dig into what large language models are actually doing to the economy. Not the speculative version you get in op-eds or panicked LinkedIn posts, but the kind of grounded, data-backed work that could tell us whether AI is quietly reshaping wages, hiring, and productivity right now, in 2024, rather than in some hazy future.

The request is notably open-ended. OpenAI isn't dictating a single methodology or a narrow set of approved conclusions. Instead the company is asking economists, labor researchers, and data scientists to propose their own angles on the problem, whether that means tracking how customer service roles change after chatbot adoption, measuring productivity shifts among coders using tools like Copilot, or examining wage pressure in industries where LLMs are creeping into daily workflows.

What's interesting is the timing. Most of the loudest claims about AI and jobs so far have come from consulting firms with an incentive to sell AI transformation services, or from labor advocates warning about mass displacement. Actual peer-reviewed, empirical research on this specific technology wave has been thin. OpenAI positioning itself as a funder of that research is a notable move for a company that's also the one building the models being studied, and it's worth watching whether the selected work ends up independent in practice or just independent on paper.

Still, the underlying problem is real and underserved. Companies making hiring decisions, governments drafting labor policy, and workers trying to plan careers are all operating on vibes right now. Nobody has solid numbers on how many jobs have actually shifted because of ChatGPT-class tools versus how many shifted for unrelated reasons. If this call for proposals produces even a handful of rigorous studies, that's more empirical ground than the field currently has.