AI Dictionary › Prompting
Self-ask is a technique where the model, before giving its final answer, explicitly poses intermediate sub-questions and answers them one at a time. It breaks a complex query into verifiable steps. Example prompt: "Before answering, ask the follow-up questions you need. Question: who was younger when their first child was born, person A or B? Follow-up 1: when was A's first child born?... Final answer:". The model generates and resolves each sub-question, making reasoning traceable and more accurate on multi-hop questions.
Use it for compound or multi-hop questions where the answer depends on intermediate facts. It also helps debugging: if one sub-answer is wrong, you immediately see where the reasoning drifted.
Instruct the model to break a question into sub-questions, answer each explicitly, then compose the final answer from those pieces. Making the intermediate steps visible reduces skipped reasoning and makes errors easy to locate.
"Before answering, list the sub-questions this depends on, answer each in one line, then give the final answer. Question: should our 12-person agency switch from hourly billing to value pricing?"
Multi-part decisions, comparisons, anything where a hidden wrong assumption would poison the conclusion. The visible sub-answers are your audit trail: check them, not just the final verdict.
Multi-turn prompting · JSON mode · Prompt scaffolding · Contrastive prompting · Context stuffing · Retry with backoff · Chain-of-thought · Role prompting · Prompt injection · Temperature · System prompt
From our network
AGORÀ Intelligence: Enterprise AI Governance Platform
Govern AI at scale: policies, adoption and measurable results on your data. Built for boards and C-suite.
Visit agora-intelligence.com →From the Agora Intelligence blog
📱 Download the Android app (beta) iOS coming soon
Say what you mean. Get what you need.
Grace Certified, the AI coach that trains and certifies your prompt engineering, by Agora Intelligence.