My Wife Is Usually Right. I Build It Anyway.
My wife tells me where the project is thin, and she's usually right. I build it anyway. The problem was never building alone. It was overruling the room.
My wife will sit through a 40-minute teardown of whatever I'm building and tell me, plainly, where it's thin. She's usually right.
I nod. I wait for her to leave the room. Then I keep building it the exact way she just told me wouldn't work.
For three years I told myself the problem was that I build alone. That was the third wrong answer in a row. I'd already called this same itch a hunger I couldn't explain, then a habit I was ashamed of. The truth was simpler, and worse.
I went in for a creativity lesson and got diagnosed
This week I listened to Ed Catmull talk to David Senra for two hours. Catmull co-founded Pixar and has spent forty years studying how creative work happens.
Most of it orbits one invention: the Braintrust. It's a standing meeting where Pixar's directors sit in a room, watch each other's unfinished, broken films, and say to your face exactly what isn't working. They have no authority to force a single change. The director keeps every decision. All they bring is candor, on a schedule.
And candor is a precise word here, not a soft one. Catmull chose it over "honesty" on purpose. Honesty carries moral weight; call someone dishonest and it lands as an accusation, so people go quiet to stay safe. Candor drops that baggage. In his words, it's "forthrightness or frankness… not just truth-telling but a lack of reserve." Saying the whole of what you actually see, holding nothing back. It's not a referendum on anyone's character. That's the thing my Braintrust offers and I keep declining.
Catmull tells a story about a guy who left Pixar to make a film on his own, to prove he could do it "without the safety net." Catmull's response is flat: "That's not a safety net. These were your colleagues trying to do things." Then the line that caught me. The wrong question, he said, is "how much can I do on my own." The right one: "how much can I do with others." He'd spent a year of his life stuck on the selfish version of it: how much of Pixar was actually me? He concluded: "Trying to answer it is an act of separation… it wasn't good for my soul."
Seventy-three repositories. Twenty-six products. Every one of them was me answering the question Catmull says is the wrong one.
The comfortable lie was that I build alone
That's the lie, and you already watched me tell it. I'm not alone. The wife with the 40-minute teardown is a Braintrust of one. So is the AI I've told to disagree with me, push back, ask for context before it agrees. The candor is in the room, on a schedule, exactly like Catmull's.
And I learned to veto all of it.
What I actually do is overrule it
Catmull describes what building feels like from the inside: "you're building this thing in your head, it's kind of fragile, and sometimes you need somebody to say 'that's not working' and jar you from that fragile thing." The fragile thing is the idea itself: half-formed, unproven, and so far seen by no one but you. The Braintrust exists to jar you loose from it before you fall in love with it. The goal of that room, he says, is the moment "ego has left the room… they're always back on the problem."
Mine never leaves. The disagreement arrives, I nod, and I keep building the fragile thing anyway.
Ray Dalio built the world's largest hedge fund and spent decades studying this exact failure. He names the mechanism: "The two biggest barriers to good decision making are your ego and your blind spots." The ego barrier is the part of you that feels a disagreement as an attack instead of a gift, so you defend instead of weigh. His fix was a single question he started asking himself: "Rather than thinking, 'I'm right,' I started to ask myself, 'How do I know I'm right?'"
I don't have a disagreement problem. I have a dismissal habit. My wife jars me loose from the fragile thing, and I reach right back and steady it before it can fall. I keep the disagreement and throw out the verdict.
I had the fear pointed at the wrong thing
A few posts ago I wrote that my Resistance points at selling. I had the object wrong twice over. It's not selling I flinch from, and it's not even people. It's letting someone else be right.
The proof clears about $900 a month in my garage. The washer-and-dryer rental is the one business I've built where I couldn't cast the deciding vote: customers either paid or they didn't, and no amount of me being sure overruled them. It worked because reality had veto power and I didn't. Every product I built alone is a monument to the version where I kept it.
What I'd actually do
If you're like me, the fix was never about finding better feedback. You already have it. The candor's in the room; the ego's at the door.
So stop trying to recruit a Braintrust; you almost certainly have one. A spouse. A friend. The model you keep over-steering until it agrees with you. The gap was never access. It's surrender. The next time one of them says "that's not working," don't answer. Wait a full day before you defend it, because the instant rebuttal is the ego barrier moving in real time, and the only way to starve it is to deny it the comeback. Then ask Dalio's question out loud: how do I know I'm right? If the only answer is "because it's mine," that isn't an answer. It's the thing Catmull spent forty years trying to get out of the room.
The AI in my terminal is the best production tool I've ever held and a useless judge. Not because it can't disagree, but because I can talk it out of disagreeing in one line. I made my Braintrust agreeable on purpose, then wondered why it never saved me.
Close
Catmull got the thing he'd chased for twenty years, the first computer-animated film, and then sat in the silence afterward asking "what's the point of my life now?" It took him a year to answer.
I'm nowhere near that problem. But I caught the cheaper lesson early.
The room was never empty. I just keep casting the only vote.