How Models Learn to Behave: SFT, RLHF, and What Alignment Means for Product
The model was accurate. Users called it "robotic" and "like talking to a FAQ page." The team had solved accuracy. Nobody had solved alignment.
Written by
Shipping AI products. Working smarter with AI. Sharing what I learn.
Senior Product Manager writing about two sides of AI: building AI products that work at scale, and using AI to work more effectively as a PM. I share frameworks for Applied AI product management—economics, evaluation, agent design, responsible deployment—alongside practical guides for AI-powered productivity, workflows, and decision-making. If you're building AI products or figuring out how to leverage AI in your PM workflow (or both), this is for you. Currently based in Seattle.
Responses
Markdown supported · ⌘↵ to post
No responses yet.