John Riccardi @arcanegrain.dev · 17/06/2026Me to devs: DO NOT trust what the LLM wrote Me to LLM reviewer: DO NOT trust what the human wrote in PR description / docs Me to LLM reviewer 2: DO NOT trust the code the LLM wrote 000
John Riccardi @arcanegrain.dev · 18/02/2026OpenAI seems to be using a dark pattern where you're unable to delete your account, and then when you talk to the chat bot (bc of course its a chat bot) and you confirm that you've tried the listed steps, it just stops responding. This has happened twice now. #openai #chatgpt 000
John Riccardi @arcanegrain.dev · 04/02/2026Welp, now that Claude is having issues, can we talk about how having all companies use a single (ok, maybe 3?) product is a bad idea? 010
John Riccardi @arcanegrain.dev · 28/01/2026I created a little RAG chatbot connected to some old school D&D rules (Old School Essentials / BX). It runs completely offline using llama3. I ran into some interesting issues, so I've been trying to document as I go in case I turn it into a tutorial. 140
John Riccardi @arcanegrain.dev · 26/01/2026I'm starting to feel more and more like half the AI terms thrown around are just the same thing re-marketed 😐 010
John Riccardi @arcanegrain.dev · 08/04/2025AI is great at filling in code scaffolds, but it struggles with vague, broad goals. So when I write test descriptions like this, it has context. AI can infer function names, required params, and even edge cases. 100