Adrian Serrin: “I’ve Never Been Very Interested in Being the Smartest Person in the Room”

Adrian Serrin has spent much of the last decade becoming the sort of person other people describe with nouns he rarely uses for himself: founder, billionaire, industrialist, futurist.

Ask Serrin what he is, and the answer is considerably less complicated.

“Engineer.”

He says it without irony, although his days now contain considerably more budgets, government meetings and personnel decisions than machinery.

We met at Avelor's headquarters for what was scheduled as a forty-five-minute conversation and lasted nearly twice that. Serrin arrived without an entourage, declined coffee because he had already had “too much,” and spent several minutes asking questions about how this interview would be edited before agreeing to begin.

What follows has been edited for length and clarity.

THE HALDEN LEDGER: Do you still think of yourself as an engineer?

SERRIN: Yeah. I don't know what else I'd call it.

What does that mean to you now? You're not spending every day at a workbench.

SERRIN: Engineering isn't really about the workbench. It's tradeoffs. You have something you want to happen, a bunch of reasons it can't happen, and limited time and resources to figure out which reasons matter. A lot of management is the same problem with worse documentation.

Is running Avelor engineering?

SERRIN: Parts of it.

Which parts aren't?

SERRIN: People don't behave like hardware.

Is that frustrating?

SERRIN: Frequently.

You said that very quickly.

SERRIN: Because I've had time to test the hypothesis.

Were you always ambitious?

SERRIN: Curious, probably. Restless. Ambitious is one of those words that's easier to apply afterward. When you're twenty-something and trying to make something work, you don't usually think, I'm pursuing my ambition. You think, Why doesn't this work? Then you fix it. Then you find the next thing that doesn't work. If you do that long enough, someone writes an article calling you ambitious.

Did you expect Avelor to become this large?

SERRIN: No.

Really?

SERRIN: I hoped it would. That's different.

When did you realize it could?

SERRIN: There wasn't one moment. Companies grow the same way complicated machines do. At first you're just trying to make one part work. Then you realize the part depends on another part. Then that depends on three more things. Eventually you either control enough of the chain to make the result reliable, or you spend your life explaining why somebody else's delay wasn't your fault. I don't particularly enjoy that explanation.

That sounds like the reason Avelor has brought so much work in-house.

SERRIN: It's one of them. If something is genuinely commoditized and reliable, buy it. If the entire mission can stop because one supplier in one building has a bad quarter, you don't have a supply chain. You have a dependency.

You seem unusually preoccupied with redundancy.

SERRIN: Things fail.

That's it?

SERRIN: That's usually enough. A system that works beautifully until one component breaks is a demonstration, not infrastructure. That applies to rockets, factories, software, organizations—

People?

SERRIN: Especially organizations.

You've said you dislike meetings where everyone agrees.

SERRIN: I do. Either the problem is trivial or somebody isn't talking.

So you deliberately hire people who disagree with you?

SERRIN: Absolutely. It's cheaper than being wrong.

What do you look for when you're hiring?

SERRIN: People who can say “I don't know” without apologizing for it. People who are curious. People who can change their minds when the evidence changes. And people who understand the difference between disagreeing and performing disagreement.

What's performing disagreement?

SERRIN: Waiting until the decision is made and then explaining why you always knew it was bad. That's not dissent. That's theater.

What should someone at Avelor do if they think Adrian Serrin is wrong?

SERRIN: Show me.

That's all?

SERRIN: Ideally.

And if they can't prove it?

SERRIN: Then tell me what they're uncertain about. “I can't prove this yet, but I think we're missing something” is a perfectly useful sentence.

Intelligence?

SERRIN: Useful. Overrated by intelligent people.

That's an unusual thing for the head of an engineering company to say.

SERRIN: I've met very smart people who could explain exactly why something impossible couldn't be done. I've also met technicians who looked at the same thing for five minutes and asked the question nobody else thought to ask. Raw intelligence matters. Judgment matters more often.

How do you react when somebody makes a mistake?

SERRIN: Depends on the mistake.

What's the difference?

SERRIN: If you made a reasonable decision with incomplete information and it went badly, we learn from it. If you concealed information because you were afraid of looking stupid, that's different. If the same thing happens three times and each time everyone calls it surprising, now I'm irritated.

Do you fire people for failure?

SERRIN: Of course. But less often than people probably assume.

What gets someone fired?

SERRIN: Dishonesty. Refusing to learn. Making the organization less capable because protecting your position became more important than solving the problem. Failure itself isn't particularly interesting. Everyone fails. What matters is what the failure teaches you.

What's the most dangerous kind of success?

SERRIN: The kind that convinces you your assumptions were right when you were actually lucky.

He pauses.

SERRIN: That's harder to diagnose.

Avelor has a reputation for controlling unusually large portions of its own infrastructure. Do you have trouble trusting other organizations?

SERRIN: Sometimes.

That's also a quick answer.

SERRIN: Some organizations deserve enormous trust. Others have excellent PowerPoints.

How do you tell the difference?

SERRIN: Ask what happens when something goes wrong. Everybody can describe the normal case. I want to know who has authority at two in the morning when three things have failed, the person who wrote the procedure is asleep, and the backup procedure assumes something that isn't true anymore. That's where you find out whether you built a system or a diagram.

That sounds slightly pessimistic.

SERRIN: I think it's optimistic.

How?

SERRIN: I assume things will fail and people will still be able to fix them. Pessimism would be assuming the failure is the end of the story.

What were you like before Avelor?

SERRIN: Broke.

That's the whole answer?

SERRIN: Less patient. More convinced I could solve everything myself. Being broke was just easier to measure.

Was there a point where money changed your life?

SERRIN: There are several points. The first is when you stop doing arithmetic at the grocery store. Then when a car repair doesn't ruin the month. Then eventually there are numbers where money stops changing your daily life and starts changing what problems you're allowed to attempt. That's the part people don't talk about as much.

What was the first thing you bought when you had real money?

SERRIN: A decent mattress. The one I had before that violated international law.

No sports car?

SERRIN: Eventually. I didn't say I became enlightened.

Do you still have it?

SERRIN: One of them.

One of them?

SERRIN: You asked.

Do you enjoy being wealthy?

SERRIN: That's a strange question.

Why?

SERRIN: Because the obvious answer is yes. It's objectively better to have enough money than not enough money. Anyone pretending otherwise should try not having it.

Has it made you happier?

SERRIN: It removed problems. That's not exactly the same thing. Money makes ordinary emergencies stop being crises. A broken appliance is annoying instead of catastrophic. You can protect your time. You can help somebody without first checking what it will cost you. That's an enormous privilege. But eventually you discover that the problems money can't solve are still waiting for you.

What can't it solve?

SERRIN: People. Time. Bad decisions. Mortality. The usual collection.

What do you think money is for?

SERRIN: Options. At the personal level and the company level. Capital is stored optionality. You can wait. You can move quickly. You can survive a mistake. You can build something before the market understands why it matters. That's much more interesting to me than the number itself.

Why spaceflight?

SERRIN: Because access changes everything. If getting somewhere is rare and enormously expensive, almost nothing can grow there. Lower the cost, improve reliability, increase frequency, and suddenly people start finding uses you didn't predict.

So what is Avelor trying to accomplish?

SERRIN: Make access routine.

That sounds less romantic than most space-company mission statements.

SERRIN: Good. Romantic gets you a documentary. Routine gets you infrastructure.

You don't find space romantic?

SERRIN: Of course I do. I just don't think romance is an engineering requirement. Watching a launch is extraordinary. The goal is for scheduling one not to be.

What happens if you succeed?

SERRIN: Hopefully people stop caring who built the rocket.

That seems unlikely for you personally.

SERRIN: Give it time. People used to know the names of railroad magnates. Most people buying a train ticket today just want the train to arrive. That's success.

Is Themelion the beginning of permanent human settlement away from Earth?

SERRIN: Maybe.

That's less confident than I'd expect.

SERRIN: Because “permanent” is a large word. We're getting better at staying there. That's different from being able to stay there indefinitely.

What would have to change?

SERRIN: Everything boring.

Such as?

SERRIN: Maintenance. Repair. Food. Water processing. Manufacturing. Spare parts. Medical capability. Power. Logistics. The exciting technology gets you there. The boring technology lets you stay.

Does that interest you more?

SERRIN: Much more. The first landing proves you can reach somewhere. The thousandth maintenance procedure proves you live there.

You'd rather build the plumbing than plant the flag?

SERRIN: Someone should probably do both. But plumbing scales better.

You operate in an industry where people can die when something goes wrong. Does that change how you make decisions?

SERRIN: It should.

Does it make you more cautious?

SERRIN: Sometimes. It also makes me less tolerant of pretending caution and safety are the same thing.

What's the distinction?

SERRIN: Not acting has consequences too. If an old system has a known failure mode, refusing to replace it because the replacement has uncertainty isn't automatically the safe choice. You're choosing between risks. You're not choosing between risk and no risk.

Do people misunderstand that?

SERRIN: Constantly. People like zero-risk options. There usually aren't any.

How much responsibility do you personally feel for what happens at Avelor?

SERRIN: A lot.

Everything?

SERRIN: No. That would be arrogant. There are tens of thousands of decisions made here that I never see. My responsibility is making sure the system gives capable people what they need to make good ones.

But ultimately your name is associated with the company.

SERRIN: That's branding. Responsibility isn't branding.

What's the difference?

SERRIN: If something works because a technician caught a problem at three in the morning, putting my face on the magazine doesn't make it my accomplishment. If the technician couldn't get the part they needed because I approved a bad organizational structure, putting somebody else's name on the failure doesn't make it theirs.

You seem to spend a lot of time preparing for things that might never happen.

SERRIN: Yes.

Why?

SERRIN: Because after they happen, you don't get the preparation time back.

Where do you draw the line? You could prepare for anything.

SERRIN: Probability, consequence, cost. Same as any engineering decision. A one-in-a-million event that costs ten dollars to prevent is probably worth preventing. A one-in-ten event that costs the entire company to prevent probably isn't. You don't ask whether something is possible. You ask what exposure you're buying down and what you're giving up to do it.

Do you ever worry that preparation becomes its own obsession?

SERRIN: Yes.

How do you prevent that?

SERRIN: People who disagree with me.

He smiles.

SERRIN: See? Full circle.

What do people misunderstand about you?

SERRIN: I don't know. Probably that I'm more certain than I am. The written version of a person always sounds more certain.

You seem confident.

SERRIN: Confidence and certainty aren't the same thing. Confidence is being willing to make a decision when you don't have everything you want. Certainty is pretending you do.

How often are you uncertain?

SERRIN: Constantly.

You don't look it.

SERRIN: That wouldn't be especially useful.

Isn't there a danger in that? People may assume you know more than you do.

SERRIN: Yes. That's a fair criticism.

Do you think that happens at Avelor?

SERRIN: Probably.

Silence hangs for a moment.

SERRIN: One of the hazards of being right often enough is that eventually people stop checking. I don't want that.

Can you actually prevent it?

SERRIN: You can make disagreement safer. You can't make hierarchy disappear.

What do you do when you're not working?

SERRIN: Chess occasionally. Hiking. Cooking badly.

How badly?

SERRIN: Inconsistently. Which is worse.

What do you cook?

SERRIN: Things with too many ingredients. It's apparently possible to overengineer dinner.

Are you good at chess?

SERRIN: Depends who's asking.

I'm asking.

SERRIN: Then yes.

Convenient.

SERRIN: You walked into that one.

What do you like about it?

SERRIN: Consequences. You can spend ten minutes convincing yourself a move is brilliant. Then you make it and the board doesn't care about your argument.

That sounds suspiciously like engineering again.

SERRIN: Most of my hobbies eventually become engineering if you let me explain them long enough.

What are you afraid of?

SERRIN: Waste.

Money?

SERRIN: Time.

Why time?

SERRIN: Because almost everything else can be rebuilt. Money can be earned again. Hardware can be replaced. Companies can recover. Time only moves one direction.

Do you think about mortality much?

SERRIN: That became much more philosophical than I expected this interview to get.

That's not a no.

SERRIN: No. It's also not a yes.

How do you want Avelor to be remembered?

SERRIN: I don't know that I do. I'd rather it be useful. Companies are tools. If somebody builds a better tool someday, they should use it.

You wouldn't mind Avelor disappearing?

SERRIN: If it disappears because something better replaces it? That's success. If it disappears because we become incompetent, I'd be considerably less philosophical.

And how do you want to be remembered?

SERRIN: Ask me when I'm older.

You've clearly thought about it.

SERRIN: I've clearly avoided thinking about it. Different skill.

Last question. Are you happy?

SERRIN: That's unfair.

Why?

SERRIN: Because now I have to think about it.

Take your time.

Serrin leans back, looks briefly toward the windows, and for once doesn't answer immediately.

SERRIN: I think I am.

You think?

SERRIN: See? That's why it was unfair.