Search for a command to run...
I am twenty. I dropped out of university. I live in New Zealand. I taught myself to build software by being obsessed with it, and I am now most of the way through building a thing called Panacea Index, which is a research interface for supplements. People ask me what it is like, doing this at this age, and the honest answer is more confusing than the version I used to tell.
I want to write down the real version, because the clean version is not true and the clean version is not useful to anyone.
I did not start a company because I wanted to start a company. I started because I could not stop thinking about a problem.
The supplement space is strange. Millions of people take these compounds, and the people they trust for guidance are influencers whose incentives are not aligned with the truth, and the actual research is sitting right there, in papers, unread. The information gap between what is known and what is believed is enormous, and almost nobody has the time or the training to close it for themselves. I kept noticing it, and then I kept not being able to stop noticing it, and at some point that turns into a project whether you meant it to or not.
I am interested in complex systems, in language, in biotech, in how we are going to interact with machines as they get smarter. Supplements turned out to be a place where all of that lands at once. You have a huge messy literature, a huge messy conversation, and a question — what does the research actually say — that should be answerable and mostly is not. Building the thing that answers it is the most interesting problem I can think of, so I am building it.
The part nobody tells you about building something on your own, at this scale, is how much of it is waiting for a feeling that may never come.
There is a version of the product in my head that is good. Actually good. The version where every search has a real nugget of value, where the words do not feel like machine slop, where you can dig as deep as you want and the depth holds. I am not there yet. I know I am not there yet. And the strange thing is that I cannot fully specify what "there" is — I just know I will feel it when it arrives, or I will not.
So a lot of the work is building toward a target I cannot quite draw. That sounds romantic and it is mostly agonizing. You ship something, you look at it, and you think, no, not yet. You change it, you look again, no. You are not sure if you are improving it or just moving it around. The fear is that you will spend years sanding a thing that was finished a year ago and you missed it. The other fear is that you will ship something half-finished and someone will try it once and tell you it is bad, and they will be right, and the reputation will set before the work is ready.
Both fears are real and they point in opposite directions and you just have to live in the middle of them and keep working.
I will tell you the numbers because they are part of the story and because they are the part people ask about. I posted about Panacea on X. Those posts were seen about six hundred thousand times. When I put the site up, a thousand people visited in the first day, and a hundred and thirty-three signed up. I kept it free, partly because I wanted people to be able to try it and partly because seeing what people actually ask about is itself the most valuable thing I can collect right now.
I also got an acqui-hire offer. A real one, for more money than a broke dropout has any right to expect, to go build someone else's marketplace. I turned it down. Not because it was a bad offer. Because I want to see how far this can go, and going to work on someone else's thing would have ended the question before I had an answer. I am aware this might be the wrong call. I am making it anyway. That is most of what being twenty and doing this consists of — making calls you cannot fully justify and committing to them.
The romantic version of this is that I am following a dream. The real version is that I am slowly getting less wrong, about a lot of things, one at a time.
I have learned that building a research tool is ninety percent data work nobody will ever see and ten percent the interface they judge it by, and that you cannot skip the ninety percent because the ten percent is meaningless without it. I have learned that taste is a real skill and that the only way to develop it is to make a lot of things, hate most of them, and figure out specifically what you hate. I have learned that being self-taught is not a disadvantage in the skills and is a real disadvantage in the network, and that the network has to be built the slow way. I have learned that the hardest thing about working alone is not the loneliness, it is the absence of anyone to tell you when you are being stupid, and that you have to learn to be that person for yourself, badly, and then better.
And I have learned, repeatedly, that the time disappears. You sit down to work on a feature and the day is gone and you are not sure what you have to show for it. That part is demoralizing in a way I was not prepared for. Where did the time go. It is okay. We keep moving. We are almost there.
If you are young and thinking about building something, I do not have advice about whether you should. I have one observation.
The thing that separates the people who do it from the people who do not is not talent and it is not resources, though both help. It is tolerance for the feeling of not being done. Most people cannot stand it. They need the project to be finished so they can stop feeling uncertain, and so they finish it early, badly, or they never start. The whole game is learning to stay in the uncomfortable middle longer than is comfortable, without giving up and without shipping too soon.
I am still in the middle. The site is free, the work continues, and I am waiting for it to feel there, which it probably never fully will. That is the job. I am doing the job.
