Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Most (all?) of the strategies described here also work with other language models. Prompt engineering fundamentals are useful (think: get more out of the model at hand) but also transferable.

The deeplearning.ai course by Andrew Ng in collaboration with OpenAI has similar content: https://learn.deeplearning.ai/courses/chatgpt-prompt-eng



Most of these strategies will get better results when working with other humans as well!


YES THIS!!! I always say they're human roleplaying machines. Pretend it's a human, do the same thing you would do with a human on your best day and you'll get better results.

Prompt engineering for me is about empathy in a way, learning to understand where the model's attention goes and leaning into that.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: