Unlike traditional software development, where engineers provide precise step-by-step instructions to compilers and interpreters, Generative AI workflow typically starts with prompting. That is, the models are programmed through prompts - the instruction to the models. Prompts are often dynamically constructed from multiple components and inputs, and the final instruction is often in natural languages, not byte code.
For Large Language Models (LLMs), constructing good prompts, often called prompt design, is the most critical skill when developing Gen AI applications. The difference between a good prompt and a bad one is often the deciding factor of the success of an LLM-powered application.
A new engineering discipline has risen from programming Gen AI models: Prompt Engineering.
Prompt Engineering Resources
Generic Resources (for every level)
- How to Research and Write Using Generative AI Tools
- Prompt Engineering Guide
- Anthropic Prompts Library