searching for your life's work is a multi-turn endeavor
on pivots + an update: blog is changing, because I have changed
Here’s a pattern I’ve witnessed in countless conversations with people whose career trajectories I admire. Most discovered their life’s work, their techne — or true craft and path on this earth — through adjacencies. Their first hypothesis of what kind of life would fit was wrong. Searching for their life’s work was a multi-turn endeavor1. Every turn was partly right and partly wrong. Rightness revealed itself gently over time, and it brought them closer to finding a life that fit them.
In other words, slight pivots compounded over time alter the course of your life quite dramatically. Each decision itself need not be large. Of course, there is a time and place for radical change. Of throwing out your old life and ushering in the new one without sentimentality or longing. But whenever people bemoan stagnancy or ill-fitting circumstances, it seems to me they are actually one to two vector changes away from them being happy and fulfilled. The raw material is there, but the orientation or method needs alteration.
I find it comforting reading the old blogposts of early technologists. So many legendary companies’ core products that created immense economic value were once side projects, orthogonal or behind the curtain to the main thing. Slack appeared after the founders wound-down their venture backed gaming company Glitch. To solve their own developer communication frictions, they built custom internal communication piping and realized it was general purpose for any type of online team communication. Olympus sold their entire beloved consumer camera division after 84 years and became one of the world leaders in pure-play medical instruments. In interviews, leaders of both companies said they had the intuition that if they didn’t pivot it would be too late, and failure would be locked in, illiquid.
Instead, they viewed the world and their movements within it as inherently liquid. Importantly, you were restarting with key primitives. You just stripped away the hulking mass of distraction from the clean and pure real thing.
This requires a clear-eyed internal honesty. The ability to course correct when fundamental attributes of your initial decision have changed. Often the decision to alter a particular path comes with sunk cost. Are you willing to admit being wrong about a particular theory of the world? Well, maybe you were right about the theory of the world. Then, the world changed.
Pivots, and deciding on the right degree of pivot, requires acute levels of attention. It’s easier to say that you dislike the whole thing and give up on it, rather than getting more granular into exact clusters of feeling. You feel bad about X, but you simultaneously like Y. It means noticing internal comms. In flinches and an internal squeezing of the heart, or conversely recognizing the types of problems you enjoy solving. Even if no one cared, would you still play the same game? It means collecting data and market feedback. What fails, and have you tried 10 instances before marking it a failure? What do your closest friends think you’re really good at?
Like tuning an instrument string by string, I’d notice one note was off and I’d play it again. Turn the peg. Listen again. Turn the peg. It strikes me searching for your life’s work is just that. Eventually, the music rings clear and sweet.
Reality has a surprising amount of detail, after all. When I started getting more curious, I began to see more granularly across dimensions. I felt euphoric in certain environments, and stagnant in others. I had skills that were oddly transferable from one domain to another, and felt the pull from the market for this set of illegible skills. I could see the bits of my life that were impervious to change and the bits I could bend and warp to a wonderful new shape.
This is how you get intellectually unstuck. We put so much pressure on ourselves to get things right on first attempt. What we seek is some wild step change that plunges us into clarity. Instead, we should embrace the little pivots that get us closer. It might be a series of smaller changes over a long period of time that feel less than radical. But in five years, you’ll look back and hardly recognize yourself. All we can ask for is a sharpening of certain senses.
An update on the blog: documenting new media experiments in 2026 ˚⟡˖
I wrote in the subtitle of this piece that the blog is changing because I have changed. The truth is that I’ve struggled for a long time to merge some of my multiple identities into writing. The blog represents only a sliver of my overall interests which include: an analysis of startups and new technology, relationships, health/fitness, and creative introspective reflections.
In the spirit of small pivots, this year I’m documenting the process of running mini media experiments around my broader interests. Two areas I’m excited about writing and bringing onto the blog going forward:
Interviews and collaborative writing with founders and builders I admire in SF.
The “weekly discourse in the groupchats".
Here’s an example. Vedika Jain and I are running a set of collaborative experiments together. Our first podcast (mini clip below) is a snippet of Vedika discussing human data markets, an area she and Sean Cai have written about and invested in extensively. See their excellent piece “The TAM is All of Human Labor,” a look inside the real gold rush in AI: RL environments where models learn to work.
Here are some quotes from our longer conversation. It was very fun, and I learned a lot.
I don’t think we have an access problem in AI. I think it’s a steerability problem. Friends at the labs say, we’re constrained by our own imagination. And I think that’s very true because AI is essentially made like execution cheap. Like trending towards free, but it’s like, but it doesn’t really solve what is worth working on
The bottleneck for models now is actually reasoning based data, which is not really about the state of the world. It’s how do you make decisions in the world especially along these, longer horizon tasks? These are things that models are not really good at single-shotting. I think we’re in this entire world of these subjective domains. And, I think for the models to be able to hill climb on these subjective domains, they need reasoning based data.
When you think about illegibility, it’s good to think about what does the local culture over reward and under reward like relative to outcomes? San Francisco for example, over rewards younger founders, technical cleverness and ingenuity, grand ambition. And I think you can find pockets of mispricing along every single one of those dimensions.
(Also, it’s probably the first time you’re seeing my face on here as I’ve been pseudonymous for 3 years. Hi!)
I hope you’ll enjoy joining me on this new direction of the blog. Thank you for reading as always,
Nix.
Some other essays you might enjoy on craft, work, and purpose:
This term was inspired by a conversation Henrik Karlsson and I had 2 years ago in 2024. Henrik: “I always thought, I kept taking the wrong turns and then they magically add up to the right place? Isn’t that too good to be true? How could that be? And now I realize that it’s because the turns were only partly wrong. They all revealed something.”