[dead]
Ask HN: When has a "dumb" solution beaten a sophisticated one for you?
71–73 of 73 posts
Re: Ask HN: When has a "dumb" solution beaten a sophisticated one for you?
#72I recently needed AI memory and instead of setting up a vector db and RAG, I just used git as a history graph and a knowledge graph in one. https://github.com/michaelwhitford/mementum
I am surprised how terse this prompt is. > [phi fractal euler tao pi mu] | [Δ λ ∞/0 | ε/φ Σ/μ c/h] | OODA > Human ⊗ AI Is this some kind of priming incantation?
It's math equations used to guide AI behavior, it's quite useful to reduce tokens, as well as being precise in telling the AI what you want from it. I have it fully documented in it's github repository.
Re: Ask HN: When has a "dumb" solution beaten a sophisticated one for you?
#73years ago on an ACM programming contest we had a problem that was given a starting and ending position on a chess board, find the minimum # of moves it takes for a knight to go from one to the next.
the proper way to solve is this via breadth first search. while my team was working on other problems, I basically solved all possibilities by hand. (i.e. the it was basically just a delta-x/delta-y matrix with a few exceptions if the knight would have to go off the board).
As you had (have?) less computer than team members, when I finally got a chance to a computer, instead of programming out a breadth first search, I just inputted my hand calculated matrix with some scaffolding and submitted quickly and got a validation message quickly back.
I still wonder if that was the fastest turnaround time for a single problem they have seen :)