Medium rare

A large part of my job is to come up with and evaluate ideas for building new medicines and molecules. These could take years to build out, and tens or hundreds of millions of dollars of someone's capital. I'm trying to get better at this, and have some notes on friends who do an excellent job of building up versions of ideas that earn conviction.

My last year of high school, Soham emailed me off a Hacker News post I made about a side project, trying to recruit me for his manufacturing automation company. I didn't have the bandwidth to take up the gig, but we stayed in touch. Every so often he'd call me and just go on monologues about his ideas — a new medical school in the Caribbean, a research institute in India, a new mRNA vaccine platform with thermostable lipid nanoparticles — almost all of which I had nothing interesting to say about. I'd spend those hours on the phone in awe of how generative he was. The ideas didn't stop with this guy.

Over the years I started to have more to say. At some point I suggested he shut down the manufacturing automation company and full send the mRNA platform — not for any technical reason, I just thought he seemed 10x more excited about it. He did it. I was in high school and knew almost nothing about biology, and he took it seriously anyway.

That's what he did for me. He showed me it wasn't magical to come up with ideas, that I could do it too. He showed me ideas didn't come fully cooked, and I got to see him build them in real time, inch by inch. Eventually, I got to college, met people working on robotics and biology, and started having ideas of my own that I wanted to run past him. He gave them a shot and built them up with me.

I later realised how much time and effort that must have taken him. But I also realised that it didn't really cost him much — and that doing that was one of the most fun ways for him to spend his time. It's maybe his favorite thing in the world to build up the ideas people bring him, and it became my favorite thing as well. That's most of the reason why I moved back to India after graduating to join him at PopVax — most of my job is shooting the shit with him and others about new medicines and molecules.

My third semester in college was a weird one. The first two I'd spent locked in on experiencing SF and America — random events, rushing a fraternity, almost paying for college off crypto. But somewhere in there I stopped finding joy in learning, and in the projects I was building. I felt lost.

I told Seyone about it, and the same night he pulled an all-nighter with me. It alternated: whiteboard in the BAIR office, walk, Nash Cafe, repeat. Every option went up on the board. Drop out, work on side projects, make enough money to effectively retire. Switch majors to cognitive science because maybe that'd be more interesting. Make NFTs (it was early 2022!).

I was panicking. He understood that, and he spent the first five hours not saying what he thought. Seyone has opinions and he isn't shy about them, so this wasn't him figuring out what he thought — he was trying to get inside my head, to figure out what I actually wanted, because clearly I didn't know. Then he articulated what I was going through when I couldn't: I didn't feel challenged.

Only after that did he start giving advice. And what we landed on wasn't on the board. Not one of the things I'd walked in with, but a gap semester doing internships at a robotics startup and a bio startup — two fields I knew nothing about. Just thinking about doing it made me feel more alive than I had in a long time. That semester changed my life, and there's no chance I do it without him.

This is just what Seyone does. He's done it for basically everyone in his life. I watched him shape the careers of at least 7 friends in college. It takes hours, it takes real attention, it takes constraint, and it takes the best of intentions. Seyone is generous enough to give it all.

Will is the best person I know at getting me out of my own ideas.

A while ago I was planning to start a company doing droplet microfluidics for co-encapsulation-based screening of secreted proteins. I was telling Will about it one day, and he just kept asking questions. He doesn't know much biology, so the questions weren't clever.

I explained what we'd be doing. He asked if you needed to do it this way to screen secreted proteins at high throughput. I said yeah. He said neat, that makes sense — why do you want to screen secreted proteins at high throughput? I gave some long-winded answer about having a differentiated drug platform. He said oh okay, why do you want a differentiated drug platform? I said because I want to make therapeutics. He said hmm, does that need a differentiated drug platform?

I guess not.

He pushes my reasoning when I'm attached to bad ideas, and I trust him to do it. I've come to almost be relieved when he does it because it allows me to let an idea go. I'm good at convincing myself of things and building elaborate pitches internally… but Will helps me see through them.