written by generative AI chatbots stirred up chaos in the newsroom, riddled with errors about the expansive universe. It was an event that highlighted one of the challenges of generative AI and raised serious questions about plans of some companies to automate complex forms of writing.
For the study, authors Shakked Noy and Whitney Zhang, both PhD candidates at the Massachusetts Institute of Technology, enlisted 453 college-educated people to do generalized writing tasks. They included writing a press release for a new product, a company-wide email about a new policy, or a consulting report about a new trend. Each person was asked to do two assignments that a professional would do in a given field .
“When we first started playing with ChatGPT, it was clear that it was a new breakthrough unlike anything we've seen before,” Noy toldThe study’s finding that productivity went up may not be surprising—first drafts are often time-consuming, and generating one at the click of a button is a powerful tool—but the finding that quality also rose may be unexpected.
The study comes as generative AI apps like ChatGPT, Google’s Bard chatbot and OpenAI’s DALL-E II image creation tool have set off intense excitement and dismay over how AI may change the future of work. McKinseygenerative AI’s boost on productivity could add between $2.6 trillion and $4.4 trillion in value annually to the global economy. OpenAI estimates 80 percent of jobs can incorporate generative AI capabilities into their workflows.