Things I’ve Learned from Spacemesh
Three Things #81: August 13, 2023
This should already be clear from my writing on the subject but, in case it isn’t: Spacemesh is a unique team building a unique project. I’ve written quite a bit about the technical reasons for this: that we’re building something quite different, that we’ve made some iconoclastic decisions, that we’ve chosen to do things the hard way, etc. But I haven’t written much about the social side of Spacemesh as a team and as a group of humans. I’ll attempt to do that this week.
I’ve been a part of maybe a dozen professional teams during my career, nearly all of them technical teams of engineers and researchers. Each was of course unique in its own way but Spacemesh is the rarest, most unique, most wonderful, and most fragile of all. We talk about the technology of Spacemesh all the time because it’s the thing we’re most proud of and because we’re builders who like to talk about the things we’ve built. We rarely if ever talk about the human side of the project, but it’s just as important as the technical side because we’re literally writing our values into our code and our designs.
Thing #1: Remote is Possible, But Really Hard
“Culture eats strategy for breakfast” - Peter Drucker
In the wake of the pandemic it seems like the world has divided into two camps. There are those who feel that remote work is the future, that tools and techniques have gotten good enough that you rarely if ever need to see your colleagues face to face, and anyway that the most talented people in the world have gotten used to the flexibility of remote work so, like it or not, you need to embrace it in order to attract and retain the best people. Then there are the stalwarts who feel that teams will never be as productive or cohesive, or be able to build trust and rapport well, while working remotely, and that everyone should be in the office at least a few times a week.
I haven’t totally made up my mind and I see both sides, but in general I put myself in the latter camp (and I think I’m showing my age by doing so). I think productive, effective teams are all about trust and it’s close to impossible to really build trust remotely or virtually. I decided a few years ago that the next time I build a team I’ll almost certainly do it in person for this reason.
Having said that, remote first can work for certain teams in certain stages of the company and product lifecycle but it almost never does work, so as a general statement I think it’s easier to say that it doesn’t. An effective, remote first team is something of a unicorn: a rare, mythical thing rumored to exist but one that no one has ever actually seen with their own eyes. All of the teams and projects that I’ve been a part of that tried remote first ran into these issues and either fell apart or just weren’t productive or effective.
Spacemesh is the exception. To be fair, it took a very long time to get here and a lot of mistakes were made along the way. The early team didn’t have a healthy culture or work ethic and almost everyone from that time has moved on. (Incidentally, this is something else I’ve learned from Spacemesh: you can’t blame culture on any one person. It’s a collective thing. And sometimes fixing a broken culture requires replacing almost the entire team.) We made a bunch of mistakes in trying to hire a new team, but we finally managed to do so and for the past year or so we’ve had a remarkable team in place that, somehow, has found a way to work productively and effectively in remote-first fashion.
The main thing I’ve learned is that it takes a certain person and a certain team culture to do this effectively. To paraphrase Tomer Afek, our cofounder and CEO, not all humans are capable of this—only the best ones are. Above all it requires trust and responsibility. Management has to trust that their team members are going to make the right decisions and do the right things even without micromanagement. And everyone must share a sense of responsibility towards something bigger than themselves. Everyone must share a commitment to the project and to the team, and to simply not letting the team down or violating the team’s trust.
Culture is the biggest determining factor in the extent to which the whole is greater than the sum of its parts. In other words, culture is the secret sauce that allows a team of talented individuals to come together, build something amazing, and change the world. There’s yet another difficult balancing act in constructing a team of autonomous agents, people who are competent, think and make decisions for themselves and don’t need micromanagement, who are nevertheless trustworthy and able to put the interest of the team before personal interest. Culture is a powerful thing but it must be built on top of a foundation of trust and can by no means replace trust. A team of autonomous agents working remotely is exceptionally vulnerable to abuses of trust, which is why it’s so important to establish integrity and trust above all else.
There’s a lot more to say about making remote work work. For instance, the times when it’s the most challenging (when for whatever reason you haven’t seen someone face to face in a very long time) and how we deal with that challenge, which tools we use and how we use them (Github, Slack, Zoom, and Google, nothing special), how we make the timezones work (our team is very distributed geographically but pretty tightly clustered around a couple of timezones), how often we do meet face to face and how we make that work (roughly quarterly), how to recruit (it’s complicated), etc. I could devote a lot more space to answering these questions and a bunch more, and I’m happy to do that if it would be interesting, but for now I’ll leave it here and reemphasize that, with the right people, people you trust and share values with, anything is possible—but finding the right people is really, really difficult and takes a lot of time.
Thing #2: Hierarchy vs. Meritocracy
“Bureaucracy destroys initiative. There is little that bureaucrats hate more than innovation, especially innovation that produces better results than the old routines. Improvements always make those at the top of the heap look inept. Who enjoys appearing inept?” - Frank Herbert, God Emperor of Dune
Hierarchy serves several purposes at work. The most obvious are structure and chain of command, but recently I realized that hierarchy (which is synonymous with bureaucracy) serves another more sinister purpose: job security, especially for those on top. Hierarchies and bureaucracies are almost universally good at causing an organization or project to stagnate and socially ossify, and they’re almost universally bad at innovation, which requires the opposite, namely allowing good ideas to percolate from anywhere.
In this respect a hierarchy is almost definitionally anti-meritocratic. Even if we assume that the people on top of the hierarchy are there for meritocratic reasons—they have a good track record—once the hierarchy has been formed and people have climbed aboard it’s nigh on impossible to dislodge them. One definition of a bureaucracy is a mechanism that turns an arbitrary, ostensibly meritocratic hierarchy into a permanent structure.
To the extent that people do advance and move up in a hierarchy it typically has a lot less to do with merit and a lot more to do with things like charisma, connections, and how well they play the corporate ladder game. I remember this clearly from my corporate days, it still gives me PTSD years later, and it’s the main reason I’ll never go back to working at a corporation again.
By contrast, Spacemesh is anti-bureaucracy, anti-hierarchy, and very meritocratic. I know every modern tech organization and every project that calls itself a startup (no matter how big) says it has a “flat hierarchy,” but in my experience this is a meaningless euphemism that companies throw around the way countries throw around the term democracy.
Spacemesh is really flat. In fact it’s so flat that we often have trouble figuring out the right person to make a decision or approve something. The rule of thumb is, we’re a do-ocracy: we encourage team members to take initiative and do the right thing, we trust our colleagues to do the same, we communicate early and often about these decisions and actions, and when all else fails we have a CEO who can make a final decision or approve something (but this is extraordinarily rare).
In this respect Spacemesh has somehow managed to strike a unique balance between meritocracy and democracy on the one hand, where everyone has equal voice, and the ability to execute fast on hard decisions when necessary on the other. I’ve seen too many projects slide too far into one direction or the other: either a hippie commune love-fest where nothing ever actually gets done and no one has any authority to make a decision, or else an authoritarian hellhole. I’m not sure I’ve ever seen a project or a team strike such a healthy balance between these two poles.
As I mentioned in the last thing, however, this also isn’t easy to achieve. In particular it requires a certain kind of team and a certain kind of person. It isn’t for everyone and most people won’t thrive in this sort of environment. Like successfully doing remote first it requires a great deal of trust in one another. In general I think it’s true that if you hire the right people—smart, good people who are there for the right reason—then you can trust their judgment to make good decisions without needing to micromanage them. But finding those people and building a team of those people is wicked difficult.
One thing that helps is the emphasis on raw meritocracy. Without a bureaucracy to disappear into, everyone’s swimming naked, so to speak. In other words, no one has a title or a fancy office to hide behind. We’re all evaluated and we’re all evaluating one another every single day on the basis of our current performance. We trust one another—once someone has made it through a trial period and created value and earned the trust and respect of the team, they’re a member of the family and we’ll stand by them—but that trust has to be renewed on a daily and a weekly basis. No one on the team gets to ride on the coattails of past performance (which, after all, in my experience is not necessarily indicative of future performance).
If someone has a particular title or responsibility and someone else comes along later who’s better suited, that person assumes both. This is in the best interest of the project and thus it’s expected behavior. I’m surprised so few organizations do this, although I can understand why it’s difficult: egos get in the way. But this actually works surprisingly well when you all trust one another and put little stock in titles to begin with. (It does make it a bit tricky when third parties ask you for your title, though, since we don’t really use them and this confuses people.)
I’ve never been part of such a meritocratic organization before. It’s exciting and can be a bit nerve-wracking, but it’s also liberating. It’s that rare low ego workplace where I can be myself and do good work and not worry about selling it to the team, or strategizing around roles and promotion opportunities.
In short the Spacemesh culture is a fantastic manifestation of the builder ethos: where you come from, who you know, who your parents are, what you did before, how much money you made, etc., don’t matter compared to what you’ve built and how good of a builder you are. It’s the purest form of meritocracy and it inspires people to do great things.
Thing #3: Leadership Takes Many Forms
“A team is not a group of people that work together. A team is a group of people that trust each other.” - Simon Sinek
I went to business school and have a degree in management. As part of the curriculum we studied leadership quite a bit, including leadership of many forms and in many contexts: ancient, modern, corporate, government, military, popular, unpopular, strong, weak, good, bad, you name it. We studied different leadership traits and styles and bases of power and authority, and we read case studies from dozens of companies in dozens of industries around the world. We participated in team-building exercises and leadership exercises such as technical mountain climbing in Ecuador. Of course you can only learn so much from case studies and “in the classroom” (even when your classroom is, say, an active volcano), but on top of business school I’ve also been a part of roughly a dozen teams at a dozen different organizations throughout my career. I’ve run a small startup.
In short, I thought I knew a thing or two about leadership. I thought I had been exposed to just about every kind of leadership out there, good and bad, and that you couldn’t surprise me. Then I joined Spacemesh. It should be fairly obvious from what I wrote above that Spacemesh is iconoclastic in more ways than one. One of those ways is leadership: both the type of leaders we have in the project as well as the type of leadership that’s respected and encouraged.
Spacemesh leadership is surprisingly successful but it’s unlike any other form of successful leadership I’ve seen before. It’s not authoritarian, i.e., leaders don’t simply expect everyone else to respect them and listen to them just because they’re in charge (this is almost never good leadership). But nor is it hands off and out of touch. It’s light touch leadership that emerges among peers who love and trust one another and have a long history of having worked together and done hard things together. It’s based on the sort of camaraderie that exists among soldiers who have gone to war together. It’s flexible and patient and compassionate, but strong in non-obvious ways.
Leaders garner respect and followership in different ways. For some it’s charisma or force of personality. For others it’s love and compassion. For yet others it’s fear and cold, hard respect. For a rare few it’s leading from the front, setting an example, and for even fewer, sheer competence. I’ve encountered all of these sorts of leaders before. Spacemesh leaders don’t embody any one of these things, but they rather embody a little bit of each of them.
I respect the Spacemesh leadership team because they are genuinely good people. They’re people who genuinely care about what we’re building, but also about how it’s built. They believe in living our values and expressing them in code, but also in excellence and integrity and in not cutting corners. They genuinely care about the team: the team’s personal and professional needs are paramount. We work hard, and occasionally we work long hours, but personal health and family always come first.
Leaders in Spacemesh are people who have a rock solid track record of having built good things before, which results in a form of quiet confidence (as in, “I don’t need to tell you how good I am, my track record speaks for itself”). They’re leaders that don’t assert themselves simply because they can; on the contrary, they step back and let the team take charge, and only step in where and when necessary. In a nutshell, they’re low ego leaders in the same way that Spacemesh is a low ego project. I cannot overstate how rare and special that is.
I think the biggest thing I’ve learned from my time at Spacemesh is that strong leadership takes many forms, including some forms that aren’t what you’re used to or what you see on TV and in movies. What’s more, strong leadership doesn’t always reveal itself immediately. Sometimes it takes time to understand how and why someone is a strong leader, or why a company’s iconoclastic culture is so resilient. There are many ways to be a strong leader, to inspire a team of humans to do something remarkable.
Leaders in Spacemesh are also people I trust and respect because they’ve continually demonstrated sound judgment. I’ve learned to trust the judgment of my peers even when I strongly disagree with them and never to judge a book by its cover because I’ve been proven wrong many times before. I like being proven wrong by high integrity people because it provides new information and literally lets me rewire my brain.
Finally, Spacemesh leaders are “first among peers.” We all have an equal voice, but our leaders are the ones who are willing to take that extra step, to go the extra mile, to put in the extra hour to make sure things get done, get done well, and get done the right way. I have boundless respect for such leadership and now, after four years, I can see how it’s inspired the Spacemesh team to do such amazing things. It’s working well for us today: after overcoming a lot of challenges the past few years, today I can safely say that Spacemesh is the most functional, talented team I’ve ever been part of. At the same time it’s fragile and I’m not sure whether or how we’ll be able to scale it. I’ll leave that as a topic for a future issue.
Special thank you to Tomer Afek for invaluable feedback in the process of writing this issue. All mistakes and omissions remain my own.