# Stop Being Your Own Rejection Letter *In early 2024, a founding engineer at Cursor emailed me about being their ninth hire. I turned myself down without asking a single question.* - Author: Feitong Yang (https://www.feitong.phd/about) - Published: 2026-06-16 - Canonical: https://www.feitong.phd/essays/your-own-rejection-letter - Topics: career --- This week, SpaceX agreed to buy Cursor for $60 billion. When the news hit, the near-miss stories came with it — Alex Lieberman joking that ghosting Cursor's founder back in 2022 had cost him $600 million in advisory shares.The 2022 message wasn't even an offer. Michael Truell — then an MIT student building cursor.so — simply DMed Lieberman, the Morning Brew co-founder: "Do you use any tools to help your writing/creative process? I'm an MIT student working on cursor.so and looking for some feedback." Lieberman never saw it. The "$600 million" is his own joke — a mock checklist (don't respond, don't help with content, don't negotiate for 1% advisory shares) capped with "$600 million poorer" (his post). That's the version everyone tells: the jackpot you forgot to buy a ticket for. Mine is quieter, and it isn't about money. I probably would have failed the interview anyway. What stays with me is smaller: when the chance came, I turned myself down before anyone else could. Not a regret — a lesson, and one worth passing on. ## The email In early 2024 I was thinking about leaving Citadel when a founding engineer at Cursor — Andy — found my LinkedIn and emailed me cold. He was specific and generous: the company was funded by OpenAI, the product was an AI-powered IDE already used by tens of thousands of programmers, and he was looking for a ninth team member to join in SF — someone who would, "based on their area of interest, control a significant part of our product or ML research direction." Would I want to interview? I got a lot of cold emails back then and answered almost none of them. This one I answered — and then I talked myself out of it. My first reply deflected: cool project, but I don't think my stack fits your needs. (It actually had a typo that made it read like the opposite — like I might be a fit.) Andy didn't take the hint. He wrote back asking me to elaborate, offering to just chat, no interview required. So I made it official. I apologized for the "miscommunication," and the one thing I bothered to correct was to make my no unmistakable: I wasn't a full-stack engineer, I had no recent ML experience, and I'd pass. The other side held the door open, and I shut it anyway — in writing, and left no room for doubt. I never asked them a single question. Not what they were building, not what "founding engineer" required, not whether the gap I was so sure about even existed. I ran their screening for them, reached the rejection on their behalf, and mailed it to myself. ## A bar I never checked Look at what I disqualified myself on: not a full-stack engineer, no recent ML experience. Now look at what the role actually asked for — "superb technical ability," "comfort with lots of autonomy," and an area of interest they would build the job around. The email all but said *we'll shape this around you.* I answered a requirement they had never made. So where did my version of the bar come from? True: I was good at Python, and had barely touched frontend. But the deeper problem: I was measuring myself against a standard I had invented and never tested. Part of it was Citadel and Google: years of real depth in in-house stacks that narrowed my sense of what the rest of the industry was doing, so I over-weighted everything I didn't know. Part of it was a startup call I'd taken not long before, where a cofounder wanted Rust and FastAPI, and I'd walked away feeling unqualified — over FastAPI, of all things, which is a weekend of learning. I had taken one stranger's checklist and quietly turned it into the standard I held myself to. That is what self-rejection is. It wears the face of being realistic. It feels like humility, like knowing your limits. But it is really pre-emptive surrender: you give up the option before anyone has actually evaluated you. The other side never gets to say yes, because you have already said no for them. Not every no is a mistake — sometimes it's judgment, not fear. But I'd read the email, liked the project, and could picture using the product. That much interest earns one honest question before you say no for them. ## How it actually went The story has no clean "if only" ending, and I won't pretend it does. Cursor came back about a year later, while I was working at a startup. I let the conversation sit. When I finally interviewed, in 2026, I failed the system design round — partly because I wasn't prepared, partly because I was no longer sure I wanted it. The path and I were never going to cross. And Andy, the one who first emailed me, had moved on from Cursor within the year himself. Even the person on the "winning" side of my story didn't ride it to the end. There was no alternate universe where I joined, got rich, and lived happily ever after. The point was never the outcome. The point is the conversation I refused to have. ## What changed Since then I've made myself one rule: take the call. Talk to the team. Let them be the ones to decide whether I fit. What surprised me is how much was there once I stopped gatekeeping myself. The teams are more varied, and more generous, than they look from a distance. One founder offered me 1.5 million in equity to join as a founding engineer. It's paper money, and I know it — but that was never the part that mattered. What mattered was that the door had been open the whole time. I was just the one who kept it shut. So this is the lesson I keep coming back to. The failures worth worrying about are not the interviews you fail; those are honest no's, handed to you by someone who actually looked. The expensive ones are the no's you write yourself, in advance, and never send to anyone but you. Stop being your own rejection letter.