🤖 AI Summary
Andrej Karpathy, who popularized "vibe coding" in early 2025, has declared it outdated just a year later, introducing the concept of "agentic engineering" as the next evolution in AI-powered coding. Vibe coding allowed users to input plain English instructions for AI models to generate functional code, streamlining rapid development. However, Karpathy argues this was merely a transitional phase, now replaced by agentic engineering, where humans set goals while AI agents execute tasks autonomously. Major tech companies already report leveraging this new approach, where engineers create tools and features far more efficiently than previously possible.
The significance of this shift lies in the transformative impact it has on development timelines and workflows. With agentic engineering, features that once took weeks to build can now be completed in a single day, effectively doubling the capability of teams without increasing headcount. This paradigm not only enables founders to deliver working products to customers more rapidly but also lowers the costs and risks associated with experimentation. As firms integrate this approach, they will need to adapt their strategies in expectation of a future where AI agents handle increasingly complex tasks, raising the bar for innovation and efficiency in the AI/ML community.
Loading comments...
login to comment
loading comments...
no comments yet