Tag
An opinion piece exploring the trade-off between AI assistants that silently make assumptions and those that interrupt users for clarification, questioning the optimal level of initiative for tools like ChatGPT.
A developer built an architecture skill that explicitly handles assumptions, evidence, and constraints, and is now seeking agent-generated failure cases to test it.
The article contrasts Python and Rye scripts for downloading a PDF, highlighting how initial elegant code makes many assumptions about success. It then demonstrates adding validation, which reduces readability but improves robustness.
Discusses how AI excels at finding answers within defined boundaries but may lack the ability to question underlying assumptions, suggesting a key difference from human intelligence.
This article explores why adding assumptions to specifications or tests inherently weakens the resulting property, using logical implication and examples from formal methods and Rust. It also discusses practical reasons for using assumptions despite this weakening.