In an interview with Y Combinator’s Diana Hu, Boris Cherny, the creator and head of Claude Code at Anthropic, stated that prompt engineering is largely not important. Cherny suggested Claude Code users would be better off unfollowing AI influencers on Twitter and LinkedIn. He gave an example of an internal experiment where an engineer gave Claude access to OpenCV (Open Source Computer Vision Library), an open-source library of several thousand real-time vision algorithms, and asked it to draw images, which it did—a capability that was completely unexpected. Cherny said the skill nowadays is less about prompt engineering and more about figuring out how to give Claude a hard task that seems a little bit too hard, and then making it possible for Claude to verify its work along the way, calling verification "probably the single most important thing that people do not get right, largely." He described giving Claude Tag access to a Mac virtual machine and using a simple, direct prompt to ask it to rewrite the Claude desktop app from the Electron framework to the native Apple programming language Swift.