5 ideas
What we call genius is partly a social construction shaped by the scarcity or abundance of practitioners and the maturity of a field — as more people enter a discipline, the threshold for being called a genius shifts. The same raw ability that stands out alone becomes ordinary when surrounded by thousands of skilled peers, meaning genius is relative to its competitive landscape, not an absolute trait.
Feynman refused to accept results he could not derive himself, reconstructing every theorem and law from scratch rather than trusting authority or memorized formulas. This habit made his understanding slower to acquire but far more flexible, because he owned the machinery underneath each result and could rebuild it when standard tools failed.
Rather than manipulating equations symbolically, Feynman translated abstract problems into pictures, diagrams, and physical intuition he could see and rotate in his mind. The famous diagrams worked because they replaced laborious algebraic bookkeeping with a visual notation that made the underlying physics directly inspectable, turning calculation into something closer to drawing.
Feynman treated serious problems with the curiosity of a game — investigating wobbling plates, lock-picking, or spinning tops simply because they puzzled him, with no concern for whether they mattered. This refusal to separate trivial amusement from important work was not a distraction but the actual mechanism of discovery, because freedom from obligation let him follow questions others dismissed as beneath them.
At the spring 1948 Pocono Manor Inn conference in Pennsylvania, Julian Schwinger won over the elite of physics with a day-long formal derivation of quantum electrodynamics. Richard Feynman then presented his own results using intuitive diagrams and path sketches, and Niels Bohr, Edward Teller and Paul Dirac objected that his pictures violated quantum principles, leaving him dismissed and dejected. His method was vindicated only after Freeman Dyson showed in 1949 that the Feynman, Schwinger and Tomonaga approaches were mathematically equivalent.