6 ideas
Programming is not a branch of mathematics or science but a making discipline akin to painting: hackers learn by doing and by studying others' work, not by proving theorems. Good software, like good art, comes from iterative sketching and revision rather than fully specifying the design in advance.
Smart teenagers are unpopular not by accident but because popularity is a full-time skill they decline to invest in, spending their attention on substantive interests instead. The brutal social hierarchy of high school is an artifact of confining people in an institution with nothing real to do.
Money is merely a medium for moving wealth, while wealth is the actual stuff people want; you create wealth by making things, not by taking a slice of a fixed pie. This reframing means inequality from creating new value is not zero-sum theft but net addition to what exists.
A startup lets you do many years' worth of work in a few years by working at extremely high intensity, with the payoff being a measurable lump sum proportional to the value created. The mechanism that makes this possible is measurement: small teams whose output can be individually assessed and rewarded.
Programming languages differ in power, and a more powerful language lets you express the same program in less code, giving its users a real competitive advantage. The reason most programmers can't see this is the Blub Paradox: they judge other languages from the perspective of their own, unable to perceive features they lack the concepts to want.
Good design comes from making something for an audience you understand intimately, and the audience you understand best is yourself. Designers should solve their own problems with taste and conviction rather than designing by committee or chasing the imagined needs of an abstract average user.