# Mission: Think Like a Researcher

## Why
Develop the researcher's core meta-skill the way Karpathy/Hamming embody it: the ability to ask the right question about a problem, design a cheap way to test it, and recognize when an answer is actually settled versus merely plausible. Domain-agnostic, but applying it to AI/ML work is a welcome bonus.

## Success looks like
- Given any fuzzy problem, can reformulate it into 2-3 sharp, testable questions and rank them by (importance × tractability)
- Can state, before running an experiment, what result would change their mind — and what result would mean nothing
- Can spot when they're fooling themselves: distinguishes "consistent with my idea" from "discriminates between my idea and the alternatives"
- Habitually asks Hamming's questions: "What are the important problems in my field? Why am I not working on them?"

## Constraints
- Self-paced, lesson-based learning in this workspace
- Prefers concrete examples over abstract epistemology

## Out of scope
- Academic logistics (publishing, PhD admissions, grant writing)
- Deep statistics coursework (only what's needed for "is this answer real?")