When Marrow identifies 10 possible concepts at the frontier, which 3 should it show first—and how does that decision logic differ from what ChatGPT would do?
Why now. You can build measurement and comparison frameworks, but without explicit prioritization logic, Marrow will either show random concepts or replicate ChatGPT's generic ranking. This is the missing foundation that determines whether recommendations feel personally relevant.
Capability it may unlock. Design a ranking algorithm that makes Marrow's output demonstrably different from and more useful than baseline chatbot responses
~45 min · confidence 85%
· grounded in your notes
How to start
List 5-7 recommendation scenarios from your notes (e.g., someone learning AI product development). For each, generate both a Marrow recommendation using note context and a ChatGPT curriculum. Identify what factors made one better—was it timing, specificity, connection to existing knowledge, or something else? Codify those factors as weighted ranking signals.
Supporting notes (1)
Great work emerges from the intersection of natural aptitude, deep interest, and noticing gaps at knowledge frontiers
Doing great work across fields follows a common recipe: choose work you're naturally good at and deeply interested in, learn enough to reach the frontier of knowledge, notice the gaps others overlook, and explore promising ones. Success requires working hard on excitingly ambitious projects driven by curiosity rather than rigid planning, while maintaining intellectual honesty and avoiding affectation.