Prompt engineering is a term coined in the 2020s to describe the practice of creating and crafting effective prompts to guide the behaviour and output of large language models (LLMs). Most people ask simple questions and can get simple answers. Somewhat like in the real world, the art of asking questions will bring the best results.
Effective prompt engineering involves understanding the model’s capabilities - as well as its limitations - to achieve the most optimal results.
AI prompt engineering: A deep dive - Anthropic (1h16m42s)
Dan Koe introduction to Prompt engineering
Dan Koe has released a short 30 minute video to introduce how he has learned prompt engineering, which he terms an essential meta-skill -> some timestamps below. He also walks through his own product Kortex.
System - assign a role and description of the task.
Context - reference information or an expectation of what you want to do.
Instructions - detailed instructions to accomplish the task.
Examples (Optional) - if you have specific examples, like social post templates, you can add them like we will below.
Constraints - what to avoid or include that may not be taken into account.
Output - how you want the output to be formatted, different from examples.
Jump to: Creating a good system or meta prompt (16m27s)