on AI coding

As in the software development loop:¹

1. Build a mental model of the requirements
2. Write code that (hopefully?!) does that
3. Build a mental model of what the code actually does
4. Identify the differences, and update the code (or the requirements).

AI can be very useful at step 2, but we still need to human-intervene on step 4 as there's several issues in the current generative models¹

Using re-work model² would be suitable for coding.

Footnotes

1. Why LLMs Can't Really Build Software - Zed Blog, Zotero ↩ ↩²

2. 场景化工作任务搭配不同AI工作模式 ↩