Transitions

What I'm Working on Now and Why

I joined the blockchain industry full time in 2017 because of its potential to democratize finance and to create novel financial infrastructure and institutions that are fairer, more open, and more transparent than the ones I saw when I worked in traditional finance. And as I worked to build the infrastructure to make this possible, a second calling emerged. I’ve come to better understand how modern software is fundamentally broken and how it’s designed to work for big companies, not users. I began to feel a calling to create human-centric software: applications and services that put the user in the driver’s seat and ensure that they retain control of the application and its data.

I’ve written about both of these core motivations many times over the years, including a long manifesto explaining my motivation for joining Spacemesh in 2019. And over time I’ve come to understand that these seemingly distinct callings are in fact multiple facades of a single project. A blockchain is the ultimate manifestation of human-centric software since it allows anyone to participate in the network as a full peer, rather than as a vassal of a centralized server. Blockchains enable fair, open, transparent marketplaces that, in turn, enable further development of human-centric software.

While blockchain technology and the industry have changed enormously over these eight years, my motivation has not, and I still feel excited and privileged every morning to get out of bed and be able to work on this important project. In fact, at this point it’s clear to me that this will be my life’s work, and that the work has only just begun.

The other thing that’s changed over the years is the way I’ve brought to life these goals and the underlying values in my day to day work. Specifically, I’ve chosen to work on several different projects over the years for a variety of reasons including values alignment, my potential impact, and the desire to work with the very best people. Maximizing impact means being willing to contribute to different projects at different times, being willing to admit when you were wrong, and change course when necessary.

My first full-time role in the industry was at the Ethereum Foundation from 2017 to 2019. I got to work with and learn from a bunch of amazing people, and I also had the opportunity to travel around the world and meet the nascent, global Ethereum community. I’m extraordinarily grateful for this opportunity because it was an excellent way to learn the ropes of the industry, so to speak, and because it gave me the opportunity to learn by doing, both on the technical side as well as on the social side. I worked as a core developer on the Ewasm team, focused on EVM and its Wasm-based candidate replacement VM, and over time took on a number of responsibilities related to governance such as helping host the regular All Core Devs calls and cofounding the Ethereum Cat Herders.

My first big transition came in 2019 when I was given the opportunity to join Spacemesh, a fledgling layer one project with a contrarian vision to build a novel consensus mechanism that’s not based on proof of work nor proof of stake, and a de novo protocol stack to support it. Ethereum was already quite mature when I joined in 2017, and I was interested in building a different kind of blockchain, so I jumped at the opportunity to join Spacemesh, especially since the founders had already raised money and hired a world-class team. I was particularly excited by Spacemesh’s vision of building a blockchain that was fairer and more decentralized than any that came before through its novel use of proof of spacetime.

Over the following five years I worked on many different things at Spacemesh, from the core protocol and consensus to the API to the economic model. I spent a lot of time getting to know the community, virtually and in person at events around the world, where I also had the opportunity to deliver a number of talks about the project. I published dozens of articles explaining Spacemesh as I see it. I helped out with research, business development, and project management. Throughout this period I was proud to be part of a project that was working at the bleeding edge of the industry, pushing the technical limit, and most importantly, true to its values, including putting the little guy first and leading with code, not hype.

Finally, over the past year, I at last had the opportunity to do something I had wanted to do for years: help design and implement the Spacemesh virtual machine. Spacemesh initially launched without a VM, but over the course of 2024 I worked closely with the rest of the R&D staff to design and fully implement an initial prototype of the Athena virtual machine. After more than six months of intense, full time work, I was thrilled when the initial Athena testnet finally launched recently.

At that point, I handed the Athena project off to a team of more experienced developers to take it fully to production. Over decades working as a software developer, I’ve come to understand my strengths and weaknesses. I’m good at the early phases of a project, which require a lot of creativity, innovation, and unstructured design. I’m less good at productization and optimization, at taking a prototype and polishing it for production release. Today, Athena is in better hands than mine, and the team has continued to make great progress. With this handoff, I recently began my second big transition in the industry. I want to explain what’s changed and why.

While working on Athena and other Spacemesh projects the past couple of years, I noticed two exciting trends: the explosion of AI technology and tools, and the professionalization of governance. I was fascinated by AI, having first studied it with one of the earliest pioneers of the space, Stuart Russell, more than 20 years ago in college, and I very much wanted to spend more time with the technology to understand its strengths and limitations. And while I had been extremely active in the governance of a number of projects and protocols around 2018-2019, I had become disillusioned by what I saw as a lot of failed experiments with DAOs, and put this work on hold. But I’ve always remained fascinated by governance, and was very excited to see fresh ideas and more advanced tools begin to open the governance design space and allow for some important and successful experimentation.

In particular, I got the strong sense that the intersection of these two ideas—AI and governance—was particularly important and would be particularly fruitful. AI tools seemed like the missing piece of the puzzle that could allow us to take governance to the next level, solving issues that had long plagued us such as apathy, ignorance about the issues, and lack of participation. What’s more, I didn’t see other people working on this, and it began to feel like an area where I was uniquely prepared to make a meaningful contribution.

At the same time, it wasn’t obvious what I should work on next at Spacemesh. Athena was in good hands, as was the rest of protocol R&D. I had become a “special projects” guy, but for the time being there were no remaining special projects: Spacemesh was focused on addressing its known issues, such as combating the centralizing force of mining pools, and I began to feel like I wasn’t adding much value.

I used to describe Spacemesh in two words as being about “maximal decentralization.” In other words, the Spacemesh project is completely dedicated to the proposition that, in order to be socially useful, blockchains need to be maximally decentralized. The only way to achieve this is to make it easy for ordinary people to run nodes at home, which is precisely what the Spacemesh tech stack is designed to achieve. Delivering on this proved much harder than we expected. I still believe in this proposition, and I believe in the mission, hard as it is, but I began to realize that I could add more value elsewhere, working towards a different and equally important vision: improving governance through AI.

In other words, the writing was on the wall: it was time for another transition. The opportunity soon presented itself when I was offered a research role with the NEAR Foundation. I had long been a fan of the NEAR team, community, and platform, and I had watched from the sidelines as the small but fierce team I met in the early days raised a substantial amount of money, built a team, launched a Foundation, shipped an amazing piece of technology, and fostered a thriving community. Since then, I’ve continued to be impressed as NEAR has shipped one technical innovation after another: sharding, chain abstraction, intents, and increasingly sophisticated, decentralized AI infrastructure.

In fact, NEAR was originally founded as an AI project. NEAR’s founder and the CEO of the Foundation, Illia Polosukhin, is one of the tiny number of people in the blockchain industry who actually deeply understands AI, having been an AI researcher at Google in a prior life. NEAR’s focus is different than that of Spacemesh: it’s not about maximal decentralization, but rather user-owned Internet, which dovetails perfectly with my strong belief in human-centric software. Today, NEAR is well on its way to establishing itself as the blockchain for AI. While its focus is different, I see the same strong determination and values alignment at NEAR that I saw at Spacemesh.

What’s more, I know how much Illia, the NEAR Foundation, and the entire NEAR ecosystem care about governance and how thoughtful NEAR folks are about the topic. The opportunity to work with the NEAR team at the intersection of governance and AI felt like a blessing and I couldn’t pass up the opportunity to join the NEAR Foundation full time as Head of Research.

My initial project in this new role is to expedite the establishment of the House of Stake, a decentralized, community-driven governance initiative that, over time, will take on much of the difficult work of governing the NEAR protocol including managing protocol economics. In addition to increasing political decentralization and governance participation, the project’s goals are to reduce issuance and increase the number and quality of validators. As part of House of Stake, we’re also building AI tools including an AI governance assistant for human delegates, as well as an AI delegate that will participate directly in governance. Expect to hear much more on these topics from me soon in all of the usual places.

Finally, it’s important to note that in spite of this transition I’m still contributing to Spacemesh, primarily by advising on the ongoing Athena design and implementation. It may take the project some time to find its footing, but Spacemesh has always been in it for the long-term and I can attest to the strength of will of the Spacemesh team, which has always been committed to a vision of a fairer, more participatory, more decentralized layer one.

I couldn’t be more excited about this transition and about the opportunity to work more closely with the amazing NEAR community, and I’m extremely grateful for this opportunity to work at the intersection of the most important ideas and technologies in the world with some of the smartest, hardest-working, best people in the world.

2025 is going to be a big year for our industry and for the world more generally, as abstract, pie-in-the-sky ideas that we’ve been ideating, discussing, and hacking on for years, like building a user-owned Internet and user-owned AI, finally begin to see the light of day and real user adoption.

I’ll continue to open source my thought process as much as possible, primarily here on my regular Substack, Three Things. My DMs remain open on X as well. If you’re excited about these ideas and if you want to contribute or even just to chat about them, reach out anytime.