Holy cow! What an absolute slog of a read. I’m not an AI model, but I’ll do my best to summarize that link:
[When writing code that works with non-trivial and null-able data types,] use a data structure that makes illegal states unrepresentable. Model your data using the most precise data structure you reasonably can[, …] as quickly as you can. [W]rite functions on the data representation you wish you had, not the data representation you are given. The design process then becomes an exercise in bridging the gap.
There. Hopefully someone out there learned something cool without having to read a 25-minute striptease before the author rushes through their main idea in the span of two bullet points found in the final 25% of the article.