The Science of Effective Prompting
summary
Live demos use GPT-3.5 via OpenAI APIs with a bearer token and show code generation, iteration, and testing rather than polished slides. PowerShell functions are generated, converted between JSON YAML CSV TOML, annotated with comment-based help, and backed by Pester tests that get fixed iteratively. A custom loop runs code against unit tests, echoing the AlphaCodium idea that 100 iterations raises accuracy. The session uses the PowerShell AI module and shows failures (WMI on Mac, Wi‑Fi errors) and concrete fixes.
AI-written summary · about
chapters (8)
- 0:00Welcome to generative AI
- 3:58Generative AI evolution
- 7:15Effective prompting principles
- 14:48Prompt engineering strategies
- 18:55PowerShell and OpenAI demo
- 33:23PowerShell project build
- 40:08Prompting best practices
- 43:03Q&A
related talks
01h 08mIn Prompts We Trust
47mThe Art of Prompt Engineering: Best Practices for Effective AI Interactions
45mTest Driven Prompt Programming: How AI is Changing the Way We Code
22mAI, Make It So - A Dive into the Universe of Prompt Engineering
10mHow can you use prompt engineering to debug PowerShell?
24mBetter Prompting with Pseudocode
29mSupercharge Your Productivity: Mastering Prompt Engineering
16mPrompt Engineering for Testers: Types and Applications