OpenAI Chairman Bret Taylor recently admitted that delegating coding tasks to artificial intelligence remains an emotionally difficult transition. Despite leading a company that develops some of the industry's most advanced AI coding agents, Taylor prefers to write code by hand. Speaking on a podcast hosted by Stripe co-founder John Collison, Taylor described the challenge of unlearning his deep attachment to the act of programming itself. He stated that while he aims to reach a future where he no longer writes code, he finds it hard not to care about the details. Taylor's sentiment reflects a broader anxiety within the tech sector as agentic AI tools become increasingly capable. Other industry figures have noted that AI coding assistants have effectively solved many routine programming tasks. For Taylor, the struggle is tied to his professional identity and his high standards for code correctness and robustness. He is consciously trying to detach his self-worth from the artifacts he creates, fearing that clinging to traditional coding might prevent him from becoming a self-actualized software engineer in an AI-dominated future. This perspective aligns with observations from other leaders who compare manual coding to artisanal crafts like saddlemaking in a world of automobiles. Ruby on Rails creator David Heinemeier Hansson previously suggested that while the main production of software is shifting to AI, handwritten code can still serve as a luxury skill for personal enjoyment. However, Taylor expressed uncertainty about the future of software development environments and the role of human input. He acknowledged that while he enjoys using tools like OpenAI's Codex for conversations and mockups, he has not yet fully visualized how integrated development environments will evolve or what the core focus of a developer will be. Taylor noted that he took great pride in the elegance of the code he wrote in the past. Now, he is grappling with a new question: if the craftsmanship remains important, what should that craftsmanship look like? He admitted he has not yet found the answer. While he jokes about producing bespoke, artisanal code, his primary goal is to adapt to an era where AI handles the bulk of programming. His experience highlights the psychological hurdles facing engineers as the industry moves toward a model where AI is the primary builder of software infrastructure.