An japanese ad from 1989. I love that ads are such great time capsules. They're built to reflect what people want to hear right now, so they end up recording a society's values more honestly than a lot of deliberate history does.

What Chekhov was getting at is this: necessity is an independent concept. It has a different structure from logic, morals, or meaning. Its function lies entirely in the role it plays. What doesn't play a role shouldn't exist. What necessity requires does need to exist. That's what you call dramaturgy. Logic, morals, or meaning don't have anything to do with it. It's all a question of relationality. Chekhov understood dramaturgy very well.

"I think you're trying to make up for lost time."

She thinks about it for a while. "You may be right," she says. "But how do you know that?"

"Because I'm doing the same thing."

"Making up for lost time?"

"Yes," I say. "A lot of things were stolen from my childhood. Lots of important things. And now I have to get them back."

"In order to keep on living."

I nod. "I have to. People need a place they can go back to. There's still time to make it, I think. For me, and for you."

"You can't use that strength as a protective wall around you. There's always going to be something stronger that can overcome your fortress. At least in principle."

"Strength itself becomes your morality."

Miss Saeki smiles. "You catch on quickly."

"The strength I'm looking for isn't the kind where you win or lose. I'm not after a wall that'll repel power coming from outside. What I want is the kind of strength to be able to absorb that outside power, to stand up to it. The strength to quietly endure things—unfairness, misfortune, sadness, mistakes, misunderstandings."

"That's got to be the most difficult strength of all to make your own."

Jean-Jacques Rousseau defined civilization as when people build fences. A very perceptive observation. And it's true, all civilization is the product of a fenced-in lack of freedom. … So I want you to be careful. The people who build high, strong fences are the ones who survive the best. You deny that reality only at the risk of being driven into the wilderness yourself."

Listening to Fournier's flowing, dignified cello, Hoshino was drawn back to his childhood. He used to go to the river every day to catch fish. Nothing to worry about back then, he reminisced. Just live each day as it came. As long as I was alive, I was something. That was just how it was. But somewhere along the line it all changed. Living turned me into nothing. Weird . . . People are born in order to live, right? But the longer I've lived, the more I've lost what's inside me—and ended up empty. And I bet the longer I live, the emptier, the more worthless, I'll become. Something's wrong with this picture. Life isn't supposed to turn out like this! Isn't it possible to shift direction, to change where I'm headed?

it has a persistent, inward-moving spirit that's filled with a pliant, youthful sort of curiosity.

If a pistol appears in a story it's got to be fired. Necessity is an independent conveyor from logic morals or meaning. It's function lies entirely on the role it plays. What doesn't play a role shouldn't exist. What plays a role does need to exist

Memories warm you up from the inside. But they also tear you apart.

the boy named crow is gone

"Mr. Hoshino?"

"What's up?"

"I have a question I'd like to ask."

"Fire away."

"Can nothingness increase?"

Hoshino puzzled this one over for a while. "That's a tough one," he admitted. "If something returns to nothing it becomes zero, but even if you add zero to zero, it's still zero."

Agency. Exactly. You can just do things. Marry the outcome, not the process

And it also speaks to how I sort of see IC versus management, which is that it's not that management is going away. It's not that everyone's an IC, but everyone's kind of both now. If you're an IC, you're not typing code out character by character. You are managing something. You're managing agents, you're managing work that is happening that comes together to do a certain thing. If you're a manager of teams, you're doing the same thing, just at a different granularity.

Everybody's sort of defined less by the fence and the boundaries of where design stops and engineering starts, but more the average of where they're working. So if you average up all of the things that somebody on our design team does, there's plenty of code-writing things, there's plenty of things that are product work, but on average their dots are over here.

the most valuable person right now is someone who can take an idea from idea to done with the taste to know this is great. Just shepherding throughout, this obsession with making it awesome, this kind of high agency, high taste person

The same product ships six times before it works, and the shape never changes

• Product used to fail because of shape and communication. Now it often fails because the model underneath wasn't smart enough yet.
• You may need to release the same thing six times before it lands, with the shape unchanged each time.
• Whether a feature is good stopped being a question about its design and became a question about intelligence.
• Codex proves it. Andrew is confident the February Codex app would have failed in the market if it had shipped in November. Nothing about the product differed. The only change was the models between November and February.
• Operator, Atlas, Codex and ChatGPT all carry the same underlying feature. Operator didn't work out. Very cool idea, wrong moment.
• Re-releasing it with better intelligence changed the outcome completely.
• So don't be stubborn about calling something a bad feature. It might not be ready yet.

Decentralization only works with accountability attached

• The goal is giving people a unit they own, with real flexibility to pursue goals and real goals to hit.
• This gets underthought as companies grow.
• At 100 people it's a non-issue. The only place you need it is the management team.
• At a few thousand you need units that move quickly, make their own decisions and hold real autonomy. It's very natural to lose this as you grow.
• In a small company, people with decision-making freedom sit close enough to everything that they just do the thing that makes sense. That's what you lose at scale.
• It took Confluent a few years to go from fully centralized to product areas that act like 45% of a company.
• G&A is handled for them. They carry a revenue target, dedicated marketing, dedicated sales support, a plan for success and a plan for growth.
• The message becomes real: you own this, figure it out, don't wait for someone to come to you.
• The failure mode it kills is narrow ownership. I wrote the specs, engineering shipped it, design did their part, hopefully it works.
• Owners think end to end about what's working, why customers use it or don't, and what's holding back growth.

For engineering decisions, they're mostly knowable and there's mostly kind of right and wrong answers. But usually, especially the early phase of the company, you're making a lot of very big critical decisions with a lot of unknowable aspects. It will certainly impact how the company turns out and you know that, but you don't know what the right answer is and you won't find out until later. The CEO job generally, you operate much more in a kind of fog of partial understanding. I don't know that everybody understands that, but very quickly as the organization gets bigger, it's impossible to know everything about everything. And so you have to kind of be roughly directionally right.

in practice you can't understand everything about everything. You have to understand a lot about the most important things and enough about some of the other things, and try and make that judgment.

There's always two lenses in a company. One is what can we do? And then the other is what do we have to do? And so, what can we do? That's an opinion from the team. What do we have to do is kind of imposed by the world, the market.

Teams, because they spend all their time thinking about how to do something, they become very fixated on what can be done. And it's actually very important to step back and be like, hey, to be successful in this product area, what do we have to do? And then, it's weird. Once you know you have to do it, then you find a way to do it.

Diagnosing what's broken is the hardest work in a company

• You get a top-level result you don't want. Figuring out what caused it is shockingly difficult.
• A disproportionate share of the time it's people. It's also product philosophy, timing, and being a year early. It's genuinely all of those.
• Sometimes a team gets trapped in a failing mindset or methodology. Changing the person is really about opening the aperture on what that part of the org is willing to try.
• The default diagnosis is always sales. Sales are down, so the sales lead is bad.
• That's often right and commonly true. It could also be anything else in the chain of value leading up to the sale. That's the maddening part.

Nobody inside a function can diagnose a cross-functional problem

• Anything not working at a high level is inherently cross-functional.

• People in each function usually lack the global context to diagnose it, and personality decides how they get it wrong.

• Someone on the product team assumes the answer is more product features. Maybe. It could also be that nobody knows how to sell the thing, and they're not thinking about that at all.

• Other people go the other way and blame a different part of the org out of defensiveness.

• The CEO has to think flexibly about changes across every part of the org, which is exactly what diagnosis requires.

• The method is simple. Go ask everybody why this is happening, collect all the answers, work out which makes the most sense, and ask what you'd do if that were the cause.

• You're not doing it in isolation. You're aggregating views that are each too narrow on their own.

• You're doing all this inside a fog of partial understanding. Past a certain size it's impossible to know everything about everything.

• Aim to be roughly directionally right. Know a lot about the most important things and enough about the rest.

• Jay's target is knowing about 80% of what each executive knows about their function. Not enough to do the job, enough to know what good looks like and whether it's going well.

Tenaciousness, pain tolerance, I think that's actually probably the most important ingredient for the willingness to go after the things that you have to do.