OpenAI chairman reflects on emotional shift as AI automates coding

Bret Taylor, chairman of OpenAI, says he still finds it emotionally difficult to let artificial intelligence take over coding tasks, even as the technology rapidly advances.

Speaking during an episode of the “Cheeky Pint” podcast hosted by John Collison, co-founder of Stripe, Taylor said he is trying to adapt to a future where software engineers may rely heavily on AI to write code.

“I am trying to get to a world where I’m not writing code,” Taylor said. “It’s hard emotionally, if that makes any sense. I have a hard time not caring.”

Taylor explained that he has long been invested in the accuracy and quality of the code he produces, which makes it difficult to hand over the process entirely to automated tools. 

However, he acknowledged that engineers may need to adjust their mindset as AI systems become more capable.

His comments come as technology companies race to develop powerful AI coding assistants. Tools such as Codex from OpenAI and similar products from competitors are increasingly able to generate and debug complex code with minimal human input.

Some developers have begun questioning what their role will be if much of the coding process becomes automated. 

Others suggest that writing code manually could eventually resemble a craft or hobby rather than a core part of software production.

David Heinemeier Hansson, creator of the Ruby on Rails framework, previously compared traditional coding to artisanal work, saying developers may continue writing code by hand for personal satisfaction even as automation grows.

Taylor said he frequently interacts with AI tools but remains curious about how future development environments will evolve as the technology matures.