Prompt Engineering refers to an area of Artificial Intelligence (AI) for ensuring quality and purpose of AI Models. Prompt Engineers are typically Quality Assurance skilled content writers with deep domain knowledge who are able to generate the right prompts to use as input to any Natural Language Processing (NLP) / multimodal text to image/video AI Models such as ChatGPT, Dall-e2, GPT-4, pi.ai, PaLM, Bard, Midjourney, Character.ai and Chatbots. Frameworks include LangChain and Eval. Prompt generation approaches include Zero Shot, One Shot and Few Shot prompts (aka Chain of Thought Prompting – CoT) on one hand and Self consistency, Active Prompt, Generated Knowledge Prompting, Directional Stimulus Prompting, ReAct, Multimodal CoT, Graph Prompting and Prompt Injection on the other. Use cases span testing AI Models for topic/Q&A coverage, bias, hallucination, misinformation, answer conflicts, logical reasoning, ability to handle context and security. Coverage strategies include using tools like Answer the Public to generate real prompts.
Basic Elements: <INITIATION> <VERBS> <DATA INPUT>
Specifiers: <OUTPUT FORMAT> <OUTPUT TONE> <TARGET AUDIENCE> <TEMPORAL SPECIFIER><SCOPE> <CONSTRAINTS>
