Prompting AI with Natural Language Ultimately Reinvents Programming Languages
bendee983 · x · 2026-07-28
Responding to the hot take that "ordinary language is a terrible programming language," a developer outlines the typical evolutionary path of programming with LLMs:
- The Initial Magic: With simple natural language commands, AI generates code that would have taken hours or days.
- Intent Misalignment: As you push further, the ambiguity of natural language causes the LLM's interpretation to diverge from your actual intent.
- Restructuring: To keep the AI within specifications, developers start adding more structure, detailed specs, test cases, and guardrails.
- Full Circle: Looking back, you realize you have essentially reinvented programming languages.
The author concludes that despite this cycle, mastering these structural aspects is key to getting reliable output from AI coding agents and significantly boosting productivity.
More from coding & agent
- One user moved a valuable domain to a new registrar entirely with Codex — corbtt · 2026-07-28
- Raschka to discuss DeepSeek-V4, GLM-5.2, Kimi K3 and open-weight coding agents — hugobowne · 2026-07-28
- Codex steering gets praised as highly effective in long coding sessions — almmaasoglu · 2026-07-28
- MirrorCode benchmark shows Opus 4.7 solving a long coding task in 14 hours for $251 — Jsevillamol · 2026-07-28
- Reddit debates whether Omnigent should wrap Hermes Agent or stay simpler — domb_ela · 2026-07-28
- Ruff 0.16.0 Lets One Prompt Fix an Entire Codebase — xeophon · 2026-07-28