Principle
Conviction counts when it costs you, not when the crowd agrees
Separate the content of a stance from the conditions under which it was held: the signal is in standing behind it at personal cost.
Stern found the specifics of Palmer Luckey's support for Trump uninteresting. What mattered was timing: Luckey held the position in 2016, when it led to Facebook firing him from the company he had built, rather than in 2024 when much of Silicon Valley aligned the same way. Stern read it as evidence about Luckey's integrity and belief in himself.
When evaluating a leader's integrity, ask what a position cost them at the time they held it. Ignore stances adopted once they became safe.
Principle
Trust the leader with the shortest gap between who they are and how they present
Consistency of self across contexts is a trust signal, and the absence of media training can be an asset.
Stern went looking for an AI figure who was a humanist people might trust with the technology, and the name that came up most was Scott Wu. He found Wu not stage-managed, with no layers of PR or media training between him and the interviewer, and the one subject most plausibly the same person in every setting.
Look for people who act the same across settings. As a leader, resist building a separate public persona; consistency is what makes people trust you.
Principle
Things eventually fall apart, so make yourself invulnerable while you can
Treat a strong run as a temporary window for building invulnerability, not as a permanent state.
Stern says what he appreciates about Kushner is his recognition that things eventually fall apart, so you have to make the most of it while you can and do everything possible to become invulnerable. He contrasts this with many in the space who act like they are on a rollercoaster having fun while it lasts.
In good years, prioritise moves that reduce how much can be taken from you: balance-sheet strength, control, independence. Assume the window will close.
Principle
The real story only emerges after the galaxy opens up
Hold your thesis loosely until immersion reveals what the work is actually about.
Stern cites the Talmudic idea that killing a person destroys a world. His Luckey piece was meant to be a 3,000-word interview on defense tech and became a 15,000-word cradle-to-grave profile; for Zuckerberg he spoke to him and nearly 40 others over five or six months. He spends six months to a year on each piece.
Budget for discovery: expect the initial question to change once you have talked to the people around your subject, and let the scope grow when it does.
Principle
Start with the question, then find the person who best explores it
Choose subjects (or deals, hires, guests) as vehicles for a question you care about, not for their profile.
Stern's Luckey piece began with the question of whether defense tech was real or bullshit after October 7 and Ukraine; the Wu piece began with whether any AI figure was a humanist people could trust; the Kushner piece came from a long interest in the family and the state of the American dream. In each case the question came first and the person second.
Before pursuing a high-profile person or opportunity, write down the question it answers. If you cannot, the pursuit is about access, not substance.
Principle
Do the one thing you are both good at and happy doing
Concentrate on the intersection of competence and enjoyment; it is usually a single thing.
Pressed by Senra on how the work changes him, Stern says he has been focused on the success of the craft, writing prose and narrative with literary or cinematic ambition, rather than on absorbing lessons from his subjects. He moved there from an army career only seven or eight years earlier.
List what you are good at and what makes you happy; commit your career to the overlap, even if it means leaving an established path.
Principle
Classical-mogul edge: decades of unglamorous infrastructure work, not splashy bets
A self-funding business run by a patient founder who iterates one product for decades is a durable advantage that rivals dependent on fundraising lack.
Stern says he came to see Zuckerberg as a classical American business mogul rather than a new-generation AI founder. Senra adds Zuckerberg's own framing: the structure of his business means he does not have to keep raising money or ask permission. Stern ties this to 20 years of boring work like negotiating data-center contracts and tax incentives with municipal and state governments.
Optimise for a structure that lets you keep improving the same product without asking anyone's permission, and treat unglamorous infrastructure work as moat-building.
Principle
If you have a gift, you have an obligation to it
Talent creates a duty to develop and share it; not doing so is a betrayal, not a neutral choice.
Stern's college Russian professor, Natalia, believed that people who do not do everything in their power to share and use a gift are sinning against themselves and against God. Stern says it changed his life, explains why he is drawn to the people he profiles, and is now the lens he brings to colleagues at Colossus. He named his daughter after her.
Apply the obligation framing to yourself and to talented people you manage: ask what they owe their gift, not just what they want.
Principle
Judge conviction by the instant no under maximum pressure
The most reliable character signal is a decision that is costly, made under pressure, and made instantly rather than after a tortured negotiation.
Stern points to Zuckerberg's 2006 Yahoo refusal: management and the board urged him to sell, it was not a two-sided deliberation, he said no immediately, and the management team left because he would not take the money. Stern treats the speed and the cost together as what makes it unfakeable, and notes the same pattern recurs throughout Facebook's history.
When assessing a founder, look for decisions where they paid a large, visible price and did not hesitate. Discount stated conviction that has never been tested by cost.
Principle
Ruthless competitiveness without cruelty is the rare combination
Extreme competitiveness and decency to the people around you are separable, and the leaders who separate them are unusually durable.
Asked what people still misunderstand about Zuckerberg, his family, colleagues, critics, rivals and enemies told Stern the same two things: you are probably underestimating how competitive and ruthless he is in business, and he is actually not an asshole. A former researcher now at a rival lab called him a Terminator who never stops but is not an asshole.
Direct competitive aggression at the market and rivals, not at your team. When assessing a hard-driving leader, check whether insiders and enemies both describe them as decent.