• Join Puck

    Directly Supporting Authors

    A new economic model in which writers are also partners in the business.

    Personalized Subscriptions

    Customize your settings to receive the newsletters you want from the authors you follow.

    Stay in the Know

    Connect directly with Puck talent through email and exclusive events.

  • What is puck? Newsletters Authors Podcasts Events Gift Puck Careers

Code to Nowhere

A new study confirmed what many artificial intelligence skeptics have long championed, and what its own researchers didn’t expect: A.I. coding tools don’t always speed up users’ workflow. In fact, the technology can slow them down. As one MIT researcher put it: “The need for people to gain deep expertise in the topics they work on is not going away.”

Computer Programmers
In benchmark tests, A.I. coding tools tend to perform incredibly well. And while there seems to be widespread real-world adoption and plenty of hype around the potential for these tools to improve coders’ workflow, there’s been little empirical evidence to support it. Photo: Patrick Lux/Getty Images
Ian Krietzberg
July 17, 2025

Last week, a relatively small A.I. safety research nonprofit called Metr published the results of a study exploring one of the more highly touted use cases for artificial intelligence: coding assistants, and particularly how they perform “in the wild.” The results didn’t necessarily break the internet, but they did garner millions of views and thousands of reposts on X. That’s because the main takeaway was that, among the software developers who participated—nearly all of whom expected A.I. coding tools to speed up their workflow by 20 percent—the tools in fact slowed them down by 19 percent.