Read the original at Lenny’s Newsletter ↗lennysnewsletter.com · subscriber post
By Lenny Rachitsky · lennysnewsletter.com · @lennysan on X · YouTube · LinkedIn
Lenny Rachitsky shares a guest post by prompt engineer Mike Taylor covering eight prompting techniques, from role-playing and few-shot examples to chain-of-thought, retrieval, and using an LLM as a judge. Each tactic comes with a template and product-management examples. The core argument is that clearer, more specific guidance and structured multi-step systems yield far better AI outputs.
Subscriber post — summary only01Key takeaways
- Models cannot read your mind, so state what you want as specifically as possible.
- Specifying a famous style or persona can noticeably sharpen AI output.
- Few-shot examples from your own artifacts help match team formats like user stories.
- Splitting complex tasks into steps, such as planning before executing, improves reliability.
- Using AI to rate outputs against clear criteria can stand in for slow human feedback.
“Much like how becoming a better communicator leads to better results from the people you work with, writing better prompts improves the responses you can…”Lenny Rachitsky · Lenny’s Newsletter
“neither can AI, so you have to tell it what you want, as specifically as possible.”Lenny Rachitsky · Lenny’s Newsletter
02Frameworks mentioned
Summary and takeaways written by PM Atlas; quotes are short excerpts. © the original author.