Disclaimer: No one in the world wants what I said in this post to be false more than I do.


ConsenSys is not all rainbows, sunshine, and unicorns. (to the surprise of no one)

The flat governance structure is something that definitely makes ConsenSys unique and enjoyable to work at: everyone treats you as equal, snobby dicks are rare, and no one tells you what to do at work everyday or when you go to work. It would be hard to go back to the traditional hierarchical world.

However, no one here seems to have read The Tyranny of Structureless, because if they did ConsenSys wouldn’t be so radically decentralized. The lack of a formal governance structure and clear rules gives rise to informal governance structures and rules, which are born out of social relations and rules of precedence. This is bad. Incredibly bad.

Why? Because social structures and rules are super fuzzy and cannot be deposed and reformed through clearly defined processes, which means the system would be extremely inefficient and confusing, and improvements can only occur by luck and happenstance. Just think: instead of filling out a shit ton of useless forms, would you prefer it if you had to become the DMV lady’s best friend in order to get your driver’s license renewed?

This isn’t decentralization, this is fucking feudalism. Yeah, that thing that’s existed for more than three millennia.

Read the tyranny paper to learn more. Anyone in crypto should read it.

This has manifested itself in many ways, the most obvious of which is the lack of communication between spokes. Unless someone has a “good buddy” in another spoke, collaboration is rare and nearly impossible: I mean, you have to fucking cold email/slack someone to start a partnership with another team in the same fucking company. No wonder ghosting is so common.

I love the autonomy each spoke has, since everyone can just focus on their own stuff, but does the mesh and ConsenSys have any reason for existing, other than giving out Joe’s money for free? (nothing against Joe, love that guy)

IMHO, the mesh should build a clear and formal process for establishing partnerships between two spokes, which would be the first step to having a formalized governance process.


No one, I repeat no one, not even Vitalik, knows what the fuck is going on with governance. One of the main reasons, other than the field’s inherent complexity, is the lack of proper theoretical framework and nomenclature.

The word “governance” refers to so much shit right now it’s lost its purpose as a word: it includes blockchain software upgrade/forking governance, blockchain protocol development governance, dApp upgrade governance, DAOs, TCRs, digitizing real-world elections, using voting systems (like QV, LD) for DAOs, using voting systems to reform political elections, token-based incentive systems, prediction markets…the list goes on. And please for the love of fuck please don’t use the term “on-chain governance”, it’s making me puke.

And the thing that annoys the living fuck out of me is that most people, even Nick Szabo, talks as if the narrow subfield of governance that they’re talking about/building encompasses all of fucking governance and shares the exact same problems as the rest of the subfields.

Seriously, what the fuck. Everyone’s talking about governance, but they are in most cases referring to completely disparate things.

I’ve never realized how important it is to have proper nomenclature and shared lingo.

The crypto community has some of the brightest minds in the entire human history. Get your fucking shit together.


I don’t know how it happened or why it is, but almost no one in the crypto space seems to understand and hold dear the scientific method.

(Vitalik, Vlad, and all the people working on Casper, consensus and stuff don’t count, since math and logic are outside of the scientific discipline)

Almost no one is building their product or framework on reality. They start from the idea and then head towards users and empirical data, rather than let users and empirical data guide their work. This may be because you need an MVP to start doing empirical tests and experiments, and most projects don’t even have that.

I myself is also guilty of this with Betoken and WikiGit, so from experience I understand why people do this: there’s just so much possibility, and you can’t help but dream big and want to pull that vision down from the clouds. But now I have Betoken’s Testnet Beta, and I’m dealing with real users with real needs, rather than analyzing spherical humans in vacuum.

Most projects have 0 users, and have 0 market & user research. This is why many VCs ask for user research once they’ve been pitched a cool blockchain product. No matter how cool your tech is, the market comes first, users come first.

Another closely related observation: many projects are not going on the MVP->launch->improve development path; instead, they’re hopelessly idealistic, in that they’re straight up building their ultimate vision, without regard for the current market or reality in general. Aragon, Democracy.earth, Colony are some of the prime examples. Aragon is trying to build the infrastructure for the future of crypto and DAOs today, without regard for what dApp developers actually need today. Democracy.earth is literally trying to overthrow fucking governments and abolish the idea of nation states. Like what the fuck, give us a product first.

I’d love to be proved wrong here, but having the tech before UX has always been a recipe for disaster.