model roundup

GPT 4

4 items · started 2026-04-14 · closed 2026-04-19

  1. Eight months into running my automation agency, I landed a client that changed how I think about what this work is actually worth. 47-employee e-commerce brand.

  2. I'm not “spiraling” (even though ChatGPT now thinks I am every other minute), I'm just genuinely frustrated with an app I've supported from the very beginning that has deteriorated so much I barely recognize it. Specifically, they're makin…

  3. So I got tired of manually tweaking prompts over and over, so I made a Claude Code skill (Works with any LLM) that does it for me. You give it a prompt, it breaks it down, scores it 1-5, then rewrites it.

  4. I use AI for research everyday, but I kept finding myself constantly second guessing the outputs. I used to manually run identical prompts through different models (like GPT-4 and Claude) just to check for errors and see where they differe…

← all threads