AI hasn’t yet become the promised panacea for companies and their workers, but it’s not because the technology isn’t living up to expectations, according to OpenAI’s chief economist Ronnie Chatterji. It’s because workers aren’t using it correctly, he said.
At a conference hosted this week by workplace-focused media company Charter, Chatterji described a Goldilocks problem where employees give AI tasks that are too small (using chatbots as a search engine) or too big (trying to solve billion-dollar problems). “There’s a difference between what our models can do and how people use them, and that’s the biggest story in AI in 2026 — closing that gap,” he said.
It was a fascinating admission by one of the world’s largest AI companies, and one that may come back to bite the ChatGPT owner. The tech industry is littered with examples of companies that had superior technology but couldn’t master the art of getting widespread adoption (remember Palm Pilots? Betamax?). So many of these frontier AI companies have been so laser-focused on coming up with cutting-edge tech that only now are they coming around to understand they also need to make sure people use it.