It’s 8G of data in memory. You can raw dog that no problem.

(Indexes would be nice though)

This is about getting the planner to more effectively use your 'proper indexes', concerned with the edge cases where it currently does not.

They forgot to run a (same) model to optimize article for reading speed

should've made a TikTok video, instead of a write-up, amirite?

Think of it like a paper. You wouldn't ask why a paper was so long.

Also you can now ask AI to summarise it for you and even probe with questions pertaining to your specific interests.