Anticipating holes in ChatGPT's knowledge
I finished my first long blog post whose creation heavily featured my Zettelkasten, and made use of notes stretching back a couple of years. Sharing it here since this post has relevance for anyone considering using LLMs for knowledge work.
https://amahabal.substack.com/p/anticipating-holes-in-chatgpts-knowledge
--Abhijit
Howdy, Stranger!
Comments
I enjoyed reading this post. You grabbed my attention from the first sentence, because I was shocked that anyone would use an LLM for travel itinerary planning, which strikes me as an extremely bad idea. It is something that I would never do, and would never have even thought of doing it before reading this post. A lot more could be said about analyzing methods that do work well for travel itinerary planning in contrast to this kind of stupid shortcut.
If that is your opinion about use in travel itinerary planning (and I am 100% with you), please sit down before reading this: https://x.com/yunyu_l/status/1946261211915468884
It's a good thing I have no interest in travelling. And as far as the x.com link goes, at least Claude rhymes with fraud.
GitHub. Erdős #2. Problems worthy of attack / prove their worth by hitting back. -- Piet Hein. Alter ego: Erel Dogg (not the first). CC BY-SA 4.0.