Challenging assumptions in AI

Coding agents such as Claude Code are very good at producing a version of what you want. When coming up with a design for a solution the underlying large language model (LLM) is going to make a bunch of assumptions derived from the collective opinion of all the data it has been trained on. These assumptions can radically alter the design.

When designing a solution with your LLM, I think it is important that these assumptions are challenged. Once a first draft of a design is produced, try challenging it by doing something like the following:

List all of the assumptions in your design.

For each assumption list the decisions you have made without my input explaining your reasoning for making that decision.

Once all of these assumptions have been listed as concrete decisions, you alter any you don't agree with.