David Deutsch's case that knowledge can grow without bound, and that pessimism is a failure of imagination. Annoying because optimism turns out to be making the smarter argument. Hope with better logic than cynicism.

Hofstadter's masterwork on self-reference, consciousness, and what it means to mean anything. Made my brain feel like it had a secret staircase. I did not understand all of it, which is exactly why I trust it.

A biography of J.C.R. Licklider and the invention of personal computing as a project of the human mind. Computing felt romantic here, in a "people once believed tools could stretch us" way. Sexy, honestly.

Munger's mental models, lectures, and one-liners, collected. Reads like a brutally honest uncle who thinks most people are idiots. Annoying because he is usually right.

Andy Grove's playbook for running things, from a manager who actually had to. Wish I read it before confusing panic with productivity. Very founder, very unglamorous, very necessary.

Gibran's twenty-six prose poems on love, work, marriage, children, joy. The book for when ambition starts making you ugly. Soft, but with a spine.

Richard Hamming asking: are you wasting your best brain on a small problem? Rude question. Correct question.

Christopher Alexander's catalog of recurring patterns in how humans live in space. Design, but with a soul. People live inside patterns, not features.

Feynman's stories of safecracking, samba bands, and physics, told with the mind that earned the Nobel. Serious mind, unserious personality. The ideal combination, frankly.

A founder book disguised as scripture, with Easwaran as the gentle translator. Action without attachment is beautiful, impossible, and deeply annoying.

Krishnamurti on how much of our thinking we never chose. Not comforting. He is the harsh bathroom light of spirituality.

Jeff Hawkins on the neocortex as a fleet of small models running in parallel. Made memory feel alive. Also made me think companies are just many confused brains trying to agree on what happened.

James Scott on how governments simplify reality to control it, and what gets lost in the simplification. Every dashboard person should read this before thinking they understand reality. Clean systems can still be deeply bakwaas.

Michael Malone on the original Silicon Valley: chips, ego, risk, real work. Silicon Valley before it became a networking event. Dust under the nails.

J. Storrs Hall on why energy and technology stalled, and what unstalling them would look like. Civilization disappointment, with receipts. We got better apps instead of better futures and somehow clapped.

Kleppmann on storage, replication, and the unsexy guts of real software. The "I refuse to be technically decorative" book. Hard, useful, humbling.

Nadia Eghbal on the strange economics of open source maintainership. Open source is not kumbaya. It is invisible labor with GitHub stars sprinkled on top.

Tyler Cowen's case for caring about long-run growth and the welfare of future people. Made ambition feel moral, which I loved. Wanting a bigger future is not automatically spiritually bankrupt.

Ben Horowitz on the parts of running a company the other startup books leave out. Finally, a startup book that does not smell like laminated optimism. The dread is the point.

Claire Hughes Johnson on building the actual org around the actual product. Culture is not what you put on the careers page. It is what you repeatedly let happen.

Elad Gil's collected interviews and tactics for scaling a company past the early stage. Not poetic, but useful. The book equivalent of "here is where the floor usually gives out."

Hamilton Helmer's framework for what an actual durable advantage looks like. For everyone who says "moat" and means "nice UI." Cold little book. Good.

Clayton Christensen on how successful companies do all the right things and still get killed. The scary part is that the losers are competent. They do the right thing until the right thing kills them.

Sebastian Mallaby on venture capital from inside its own mythology. VC is finance with a religious belief in monsters. One huge win forgives a cemetery of mistakes.

William Thorndike on the eight CEOs whose capital allocation outperformed every market darling. Quiet CEOs doing precise, ruthless things. Much hotter than founder charisma once you grow up emotionally.

Andy Grove on strategic inflection points and what it costs to miss one. AI makes this feel less like history and more like a warning label. The old instincts work just long enough to betray you.

Martin Gurri's pre-2016 explanation of what was about to happen to every institution. Reads like someone saw the storm before everyone started posting about the weather. Institutions got see-through, and then everyone acted surprised.

Peter Bevelin's collected mental models, mostly from Munger and Charlie. Mental hygiene. Basically, your brain lies to you with excellent posture.

Fred Brooks on the eternal truth that adding people to a late project makes it later. Somehow executives keep needing this written down. More people can make software slower.

Dwarkesh Patel's interviews with the people actually building the frontier of AI. Read it for the ideas and the cadence. Some people think in ways you want to steal, respectfully.

Chris Miller on semiconductors as geopolitics, history, and the substrate of everything else. AI is not cloud magic. It is fabs, borders, politics, and machines that cost more than your entire delusion.

Nick Bostrom's careful framework for thinking about AI before everyone else started. Dry in a way I respect. AI risk does not need more theater.

Paul Graham's essays on tech, taste, and craft. This is why tech people secretly want to be artists. Honestly, fair.

Don Norman on why everyday objects fail and how to design ones that don't. Ruins doors forever. Also makes you blame bad design before blaming users, as one should.

Jordan Mechner's journals from building the original game, mostly alone. The work before the mythology. Lonely, slow, stubborn, beautiful.

Eliyahu Goldratt's parable about finding and fixing the bottleneck in a factory and in life. Find the bottleneck. Fix the bottleneck. Stop making productivity your personality.

Donella Meadows on stocks, flows, feedback, and why the world behaves the way it does. Gave names to things I already felt. The world is less random once you can see the loops.

Krishnamurti asking whether your thoughts are even yours. Worst question. Best question.

Yogananda's life across India and America, told with stories that read like a friend's. I would read this generously. Some books matter partly because of who handed them to you.

Harari on information networks across all of history, from clay tablets to feeds. The Harari I'd keep. Every team is an information network pretending to be an org chart.

Rob Fitzpatrick on how to ask customers questions that don't just flatter your idea. Embarrassing because it shows how often "customer discovery" is just asking people to compliment your delusion. Painful. Helpful.

Peter Thiel on building something genuinely new instead of competing on existing ground. Everyone reads this and suddenly thinks they are contrarian. Most are not, but the "secrets" idea still hits.

Geoffrey Moore on the gap between early adopters and an actual market. Early users are not the market. Sometimes they are just nerds with high pain tolerance.

Bruce Greenwald on where competitive advantage actually comes from once you do the math. Very unsexy, therefore probably useful. Asks where the actual advantage is, which most decks do not survive.

James Burnham on the unfashionable truth that power exists and politics is about who gets it. Cold book. Useful because power exists even when tasteful people refuse to look at it.

Eric Raymond on the social architecture of open source and why it works at all. Public work is messy, social, reputation-heavy, and somehow works. Very internet, very human.

Max Tegmark on the futures available to us depending on how AI lands. The more readable AI futures book. Good for when you want to think about the end of humanity but still make dinner.

Andy Hertzfeld's insider story of building the original Macintosh. The Mac before it became inevitable. Great products are built by brilliant, impossible people.

Will Larson's playbook for engineering management at scale. Engineering orgs are tiny civilizations with Jira. This helps explain why they become weird.

Atul Gawande on why even surgeons need lists, and what that says about excellence. Humbling because intelligence is not enough. Sometimes excellence is a boring list someone actually follows.