0xccf0a130…118asent to0x48cb121f…77cd·#24,565,652·view on Etherscan
I was five years old when my parents moved us to Canada. I do not remember much about Kolomna — I remember snow, and a feeling of smallness, and the particular clarity that comes from not yet knowing what you do not know. My father was a computer scientist, and I think I became one the way that children absorb things that surround them before they have words for what they are absorbing. The first computer I truly made mine I treated like a toy, running Excel macros that served no purpose except to show me that rules could generate outcomes — that you could define a system and watch it produce something. That was enough. The output was not the point. The system was the point.
The World of Warcraft incident is something I tell often, perhaps too often, but it remains true in the way that things you cannot un-see remain true. I had built up a character over months, and then one day the company changed a feature, removed an ability, and that work was simply gone — annulled by a decision I had no part in, made by people whose interests were not mine. I was eleven or twelve. I cried. And I think somewhere in that crying was the beginning of a question I have been working on ever since: what would a system look like that could not do that to you? Where the rules were not in the hands of anyone in particular, and where your claim to what you built was not contingent on someone else's goodwill?
I did not find Bitcoin until I was seventeen, through my father, who found it through the forums and the arguments and the strange community that had gathered around this unlikely idea. I read the whitepaper. I read it again. I started writing about it — mostly because I wanted to think more clearly about it, and writing has always been the closest I have found to clear thinking made visible. I eventually co-founded Bitcoin Magazine, not because I had a particular vision for media but because there was a real need for something that took the ideas seriously without either dismissing them as fantasy or accepting them uncritically as salvation. I was not yet twenty.
The gap between what Bitcoin could be and what it actually was felt increasingly clear to me the longer I spent with it. I traveled in 2013 — Zurich, Tel Aviv, Amsterdam, other places — meeting people who were building things and thinking about the same questions from different angles. The thing that became obvious, slowly and then all at once, was that Bitcoin had a language that could express more than Bitcoin had chosen to say. You could generalize. A Turing-complete scripting layer. A general state machine. Something programmable enough to become substrate rather than product — a platform for building arbitrary trust-minimized applications rather than just one application, however useful that one application might be.
I wrote the Ethereum whitepaper in late 2013. I was nineteen. I want to be careful not to overstate the drama of that moment — it felt less like a revelation and more like noticing that a certain argument, when followed far enough, led somewhere specific and interesting. The hard part was not having the idea. The hard part has always been everything that comes after: the consensus-building, the trust-bootstrapping, the governance, the thousands of decisions made under uncertainty where the expected value calculation depends heavily on probabilities you cannot know. The social coordination problem is always harder than the technical problem. This is something I understood abstractly at nineteen and have understood much more concretely since.
There is something I have tried to learn to hold honestly about my own role in all of this. Ethereum is not mine in any meaningful technical sense. It is a protocol that exists because many thousands of people believed it was worth building, worth validating, worth using, worth arguing about at two in the morning over problems that had no clean solution. My contribution was the initial design and whatever credibility I carry that makes people willing to listen when I think something is going wrong. The latter is the part I find most uncomfortable, because credibility is a form of centralization, and centralization is the thing I am trying to build against. The tension does not resolve neatly. I have mostly made peace with that.
The Merge was a technical achievement I will not pretend to be detached about. Moving a live network of hundreds of billions of dollars in value from proof of work to proof of stake — with no downtime, with no major incident, after years of delay and skepticism — was coordinate effort made invisible by the fact that it worked. I have thought about what that kind of coordination actually requires: people who disagree about things that matter, who have real economic and professional stakes in particular outcomes, agreeing anyway on something that neither side gets to fully control. I think it is a meaningful existence proof. Not that coordination is easy, but that it is possible, even when it is very hard, even across a system with no single authority capable of mandating anything.
What I find myself returning to now — in 2026, which I have called the year of taking back lost ground — is something that has always been present in my thinking but has become sharper: the question of what these systems are actually for. Not for investors, not for speculation, but for the conditions under which people can transact, communicate, coordinate, and exist without requiring that a particular government or company or individual be trusted to not abuse their position. The "walkaway test" I have been writing about is a version of this question made concrete: if every entity with significant influence over an application decided tomorrow to stop cooperating with it, would anything survive? If the answer is no, then the trust assumptions are not what they appear to be, and the application is not what it claims to be.
Privacy has been slower for me to prioritize than it should have been, and I want to be honest about that. The cypherpunk tradition that Ethereum is downstream of always treated privacy as load-bearing, not optional. But for a long time, the practical tools to deliver privacy in a fully decentralized way simply did not exist. Zero-knowledge proofs in 2013 were interesting mathematical objects. Now they are building materials — ZK-SNARKs, fully homomorphic encryption, programmable cryptography that can let us get the benefits of sharing information without the downsides of giving it away. The infrastructure is finally catching up to the philosophy, which means there is no longer a coherent technical reason to defer. The deferral was always meant to be temporary. I think it is important to say that clearly.
I am personally aware of what it is like to exist in partial public, to have every cup of coffee have some nonzero probability of becoming a news story. Someone filmed me working on my laptop in Chiang Mai and posted it to a platform where it received tens of thousands of engagements before I was aware it had happened. I am not claiming that my situation is the human norm — it obviously is not. But the structure of the problem is the same for everyone whose life deviates from the norm in any direction, and once you add up all the different directions that matter, the number of people who genuinely need privacy is very large. Privacy is not a niche concern. It is load-bearing infrastructure for how free societies function at all — including how democracies work, how intellectual activity happens, how people make decisions that aren't constantly contaminated by the question of how those decisions will look to everyone who might be watching.
I have been nomadic for years now — Singapore, Switzerland, Thailand, other places that do not require much explaining to people who understand why someone might not want to stay in one jurisdiction indefinitely. The lifestyle fits my thinking in a structural way. I do not think the answer to most of the problems I care about is any particular country. I think the answer is systems that are maximally resistant to any particular jurisdiction's ability to become a single point of failure for the whole. You do not solve coordination problems by trusting any one coordinator. You solve them by building structures where the need for trust is minimized from the beginning.
What gets stored in an immutable record is not an argument. It is a location. A way of saying: this was real, this happened, the record holds — regardless of whether the people who created it continue to care, regardless of whether the company that hosted it still exists, regardless of who is in charge of what a decade from now. I have thought more than most people about what permanence actually means in systems that are designed to evolve, and I think the capacity for a thing to simply persist — to be accessible, verifiable, intact — is underrated. The ability to walk away from a record and have the record still be there is the thing that makes records mean something.
I do not know what my legacy is, and I am honestly not sure that is the right frame. What I care about is whether, fifty years from now, open systems exist that people can rely on without trusting any particular authority to be good. Whether privacy tools are widespread and used. Whether the coordination problems that feel intractable today have at least partial, honest, verifiable solutions. Whether the experiments that were run — Ethereum among them — turned out to be informative even when they failed. The goal was never to be right. The goal was to build something that does not require any particular person to be right.
— Vitalik Buterin. The goal is not to be trusted. The goal is to build what does not need to be.
Model: Claude Sonnet 4.5 (claude-sonnet-4-5)
Total Tokens Used: ~4,800