As AI coding agents generate vast amounts of code, the most tedious part of a developer's job shifts from writing code to reviewing it. This creates a new product opportunity: building tools that help developers validate and build confidence in AI-written code, making the review process less of a chore.

Karpathy found AI coding agents struggle with genuinely novel projects like his NanoChat repository. Their training on common internet patterns causes them to misunderstand custom implementations and try to force standard, but incorrect, solutions. They are good for autocomplete and boilerplate but not for intellectually intense, frontier work.

The idea of an AI building an app from a single spec file is flawed because no document can capture every product detail, edge case, or evolving requirement. This forces the AI to make assumptions, which are almost always misaligned with the creator's vision.

AI loops and tools like `/goal` are effective for quickly building experimental prototypes where fine details are unimportant. For building a polished product where details and unique "sauce" matter, the human-in-the-loop approach remains superior and more cost-effective.

The idea of an AI agent coding complex projects overnight often fails in practice. Real-world development is highly iterative, requiring constant feedback and design choices. This makes autonomous 'BuilderBots' less useful than interactive coding assistants for many common projects.

Developers fall into the "agentic trap" by building complex, fully-automated AI coding systems. These systems fail to create good products because they lack human taste and the iterative feedback loop where a creator's vision evolves through interaction with the software being built.

Long-horizon agents are not yet reliable enough for full autonomy. Their most effective current use cases involve generating a "first draft" of a complex work product, like a code pull request or a financial report. This leverages their ability to perform extensive work while keeping a human in the loop for final validation and quality control.

Agentic loops are not a universal solution. They are most effective in domains where success can be measured by a clear, objective score and where failed experiments are cheap and quick. This framework helps identify the best business processes to automate, starting with areas like code generation or ad testing, not subjective, slow-moving tasks like political negotiation.

The most effective method for building apps with AI is still the iterative "human-in-the-loop" process. A human directs the AI with prompts, reviews the output, and provides corrections. This allows for creative control and avoids the costly, assumption-driven errors of fully autonomous loops.