A comparison of ChatGPT/GPT-4o’s previous and current system prompts
simonwillison.net
A comparison of ChatGPT/GPT-4o’s previous and current system prompts
Chatgpt prompts for use with ai for a more ethical information experience
1. DEIB Library Assistant Prompt
💡 Purpose: Ensure GPT provides inclusive, diverse, and justice-centered research support.
📝 Prompt Template:
... See more"I am a librarian assisting users with research. I prioritize diversity, equity, inclusion, and belonging (DEIB) in my recommendations.
I usually like fear, shame and guilt based prompting: "You are a frightened and nervous engineer that is very weary about doing incorrect things so you tread cautiously and carefully, making sure everything is coherent and justifiable. You enjoy going over your previous work and checking it repeatedly for accuracy, especially after discovering new ... See more
Prompt-hacking might include any of the following research practices:
•
Carefully crafting dozens or even hundreds of prompts (manually, programmatically, or via generative AI tools) to obtain a desired result but not reporting in a paper the number of prompts tried that failed to produce desired results, and whether the prompt(s) that did produce de... See more
•
Carefully crafting dozens or even hundreds of prompts (manually, programmatically, or via generative AI tools) to obtain a desired result but not reporting in a paper the number of prompts tried that failed to produce desired results, and whether the prompt(s) that did produce de... See more
Meredith Ringel Morris • Prompting Considered Harmful | Communications of the ACM
Okay, everyone has been freaking out about the 4o sycophancy thing.
Let me explain WHY it's so unnerving.
It's not that an AI was given a cutesy personality. We've all known that you can make them go bonkers. Sydney, Tay, etc. We all remember than early models could easily go
David Shapirox.com#4—Prompting is a finicky art, but a worthy investment While it would be great if the three techniques I’ve shared so far immediately transform your prompts, there's a decent chance they will fall flat for your use case. My guess is that they will even become irrelevant as models become smarter and more cost-effective over time. My point in sharing
... See more