[josho]

And if this use case hit any kind of scale. We’d just have an llm generate a parser and be back to microseconds.

This was just a blog to generate traffic on the site. Not to showcase some new use case for an llm.