OpenAI’s founding team member Andrej Karpathy

Former Tesla AI director and OpenAI cofounder Andrej Karpathy, who popularised the term vibe-coding has introduced a new phrase for the next stage of AI-driven software development: agentic engineering. While vibe-coding refers to humans prompting AI to write code, agentic engineering describes a process where AI agents themselves generate and refine code autonomously. In a post shared on social media platform X (formerly known as Twitter) Karpathy explained that the tern reflects not only the role of agents in writing code but also the ‘art & science and expertise’ required to guide them effectively.Karpathy who now believes that the next phase of AI-assisted programming deserves its own name: agentic engineering. He feels that unlike vibe-coding, where humans prompt AI directly, agentic engineering includes AI agents autonomously wiring and refining code while humans act as orchestrators and supervisors. Karpathy explained:– “Agentic” because the new default is that you are not writing the code directly 99% of the time, you are orchestrating agents who do and acting as oversight.– “Engineering” to emphasize that there is an art & science and expertise to it. It’s something you can learn and become better at, with its own depth of a different kind.Karpathy predicts that 2026 will bring continued improvements in both the model layer and the agent layer, making agentic engineering a central part of software development.

Read Tesla former AI director Andrej Karpathy’s complete post here

A lot of people quote tweeted this as 1 year anniversary of vibe coding. Some retrospective –I’ve had a Twitter account for 17 years now (omg) and I still can’t predict my tweet engagement basically at all. This was a shower of thoughts throwaway tweet that I just fired off without thinking but somehow it minted a fitting name at the right moment for something that a lot of people were feeling at the same time, so here we are: vibe coding is now mentioned on my Wikipedia as a major memetic “contribution” and even its article is longer. lolThe one thing I’d add is that at the time, LLM capability was low enough that you’d mostly use vibe coding for fun throwaway projects, demos and explorations. It was good fun and it almost worked. Today (1 year later), programming via LLM agents is increasingly becoming a default workflow for professionals, except with more oversight and scrutiny. The goal is to claim the leverage from the use of agents but without any compromise on the quality of the software. Many people have tried to come up with a better name for this to differentiate it from vibe coding, personally my current favorite “agentic engineering”:– “agentic” because the new default is that you are not writing the code directly 99% of the time, you are orchestrating agents who do and acting as oversight.– “engineering” to emphasize that there is an art & science and expertise to it. It’s something you can learn and become better at, with its own depth of a different kind.In 2026, we’re likely to see continued improvements on both the model layer and the new agent layer. I feel excited about the product of the two and another year of progress.