In Prompts We Trust
summary
Argument: careful prompt design reduces uncertainty and raises measurable trust by improving accuracy, reliability, personalization, ethics and alignment. Demonstrations use Prompt Flow, RAG with vector store (hybrid semantic + keyword retrieval), embeddings and system messages, plus chain‑of‑thought and tree‑of‑thought prompting, temperature/top‑p and frequency/presence tuning. An indexed 500‑page AI Index PDF is queried via RAG; the pipeline is built and generally grounds outputs, though the live demo shows failures (request failed) under low quota, illustrating both value and operational limits.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 4:10Language models explained
- 11:21Trust in AI
- 16:19Prompt engineering matters
- 20:12Key prompt considerations
- 25:52Prompting techniques
- 31:54Combining techniques
- 34:20Chain of Thought
- 39:57System messages power
- 43:13Prompt chaining
- 46:04Inference parameters
- 50:22RAG to reduce hallucinations
- 55:33Prompt patterns
- 57:19Demo: Japanese study plan
- 1:02:49Summary
related talks
01h 08mIn Prompts We Trust
02h 01mPrompt Engineering and AI Red Teaming
14mPrompt Engineering is Dead
16mPrompt Engineering for Testers: Types and Applications
44mSupercharge Your Productivity Mastering Prompt Engineering
47mThe Art of Prompt Engineering: Best Practices for Effective AI Interactions
13mPrompt Engineering Simplified
48mStop Fighting Your AI Engineering Prompts That Actually Work