Read the original at Lenny’s Newsletter ↗lennysnewsletter.com · subscriber post
By Lenny Rachitsky · lennysnewsletter.com · @lennysan on X · YouTube · LinkedIn
An interview-style episode with Marily Nika, an AI product leader, on how product managers can use AI in their work and build AI-powered products. She covers practical ChatGPT prompts, when to build versus buy models, the data needed for AI features, and the unique challenges and career dynamics facing AI PMs.
Subscriber post — summary only01Key takeaways
- Prototype AI experiences in Figma before committing to an AI-backed MVP to win buy-in cheaply.
- Only add AI once a real problem is defined and enough data, including adjacent-product data, exists.
- Most startups should rely on off-the-shelf models rather than training their own from scratch.
- AI PMs should set clear expectations with hiring managers about how research progress will be measured.
- Pair leadership pitches for AI investment with examples of adjacent successes and clear rollback plans.
“AI is enhancing us, not stealing from us.”Marily Nika · Lenny’s Newsletter · 00:06:01
Summary and takeaways written by PM Atlas; quotes are short excerpts. © the original author.