A coder's case for AI agents talking first, not planning first
The most annoying failure mode with isn't a small bug — it's when the agent understands almost what you meant. It explores a bit, makes , writes a lot of code, and the result turns out technically reasonable but not what was actually wanted. The common advice to "just write better prompts" doesn't sit well either: nobody wants to prepend a legal-contract-style ritual, like declaring "you are a senior software engineer" or "make no mistakes," before asking for something as simple as a button.
Starting in isn't the fix either — plans are useful, but a small change forces the plan to be updated and re-reviewed, and soon you're doing project-management theater with a chatbot. What's actually wanted is simpler: have the agent talk first, without interrogating, without producing a giant plan, and without jumping straight into code.
Key points
- The common AI-agent failure isn't a bug, it's "almost understanding" the request
- Long prompt rituals (e.g. "act as a senior engineer") feel inefficient for simple tasks
- can create repeated review loops for even small changes
- The proposed fix: have the agent have a short conversation before writing any code