Interview with Jeremy Millar, Consensys Chief of Staff
Last month (June 11th) I had the chance to interview Jeremy Millar, the Chief of Staff of Consensys, before the launch party for the Ethereum-based group’s Dublin ‘Innovation studio’. Here are the answers Jeremy gave me on topics around non-hierarchical decision making, the building and adoption of new technology and whether there is room for multiple ‘infrastructural’ ecosystems such as Ethereum.
You gave an interview in Ethereal [Ethereum conference in U.S.] where you spoke about decision-making in Ethereum and Consensys. How was the decision to come to Dublin actually made?
It was somewhat opportunistic in that we were able to solicit a team that had already worked together [former Deloitte directors and consultants]. The decision to do it was made out of a group of 20 key stakeholders. There are some decisions at Consensys that every individual should participate in — but if every individual was required to participate it would be hard to make decisions with velocity and hard for everybody to stay informed on every decision. Subsequently many more people have been involved in reviewing the office lease and the list of projects that should be worked on. The decision to come to Dublin was the tip of the iceberg.
How about the hiring process?
People hired at Consensys tend to go through more interviews than normal –because of an emphasis on ‘consensus’ with a ‘U’. We also highly value autonomy. If a team in Dublin wanted to hire a person it would be highly unlikely that someone else would have the authority to veto that. On the other hand the Irish budget has probably been reviewed by eighty people.
You have spoken about management tools, about tokenizing and incentivizing work within your model. Could you outline that more?
Traditional enterprise management software or ERP is traditionally hierarchical. There is no ERP for a decentralized company that I’m aware of. In a flat or non-hierarchical organization that are some features that won’t exist in WorkDay or NetSuite, like voting. It would be hard for NetSuite to add a group voting function to its capabilities. There is an interesting opportunity to create tools for managing a decentralized company leveraging blockchain technology and tokens. So that you can have greater fluidity and more transparent and decentralized governance.
Consensys is a successful ecosystem. There are other ecosystems around the world that are growing, such as EOS…
Is EOS growing? How many block producers are there on EOS? I think there are 21 nodes on the network. I think that EOS has raised a lot of capital and made some aggressive public statements, but it remains unproven as a technology.
Some would say that Ethereum is relatively unproven…
It’s been in production for over 3 years, it settles $10 billion in financial transactions a day, it has gone through various governance challenges and emerged out the other side intact. Its founders do not have a track record of project abandonment… I think the two are not comparable at this time.
Could you have more than one comprehensive ecosystem like Ethereum in the world?
Absolutely. You can have many protocols out there. In technology diffusion theory there’s a concept called a ‘dominant design’. Proof of Work is a dominant design. Satoshi’s whitepaper created a dominant design of how blockchains work. I think we are past the point of a dominant design. Could there be multiple versions of a design which integrate? Yes. So the next question is why do we see standardization in tech platforms? There is a series of reasons why. Most have to do with positive and negative externalities. It is not helpful if you have to rewrite your code lots of times if you want it to access lots of devices. It is much more helpful if you have a standard and a common interface. Which is why technology markets tend to have one or two common operating systems. It makes it a lot easier to write products. If you are writing a product you are incentivized to deploy it on the largest network. If you are writing a mobile game it is unlikely that you would target the blackberry platform first today from an ROI perspective. There are other costs associated with that — for example, people know how to run Ethereum nodes today, it remains unclear that people know how to run say, an EOS node. If you want to be sure that the application your organization is making will perform and create value, you are going to pick the most proven technology. Historically there are economic reasons that innovation has led to small number of platforms. It is also fair to observe that it has not tended to lead to a single platform. Interoperability is a key area we are working on.
In one year what do you predict for Consensys Dublin?
As Lory [Kehoe, Managing Director of Dublin Studio] said, the ambition is to be about 60 people in a year. That’s something to shoot for, it could be faster or more challenging than we thought. What’s most important is that our operating models are cohesive and that the individuals who join us are of high technical quality and are strongly culturally aligned. I think achieving that is more important than the volume. It is important we continue to operate in a ‘meshy’ way, that people participate in projects as appropriate and as they desire to, and are free from the burdens of traditional hierarchical organizations.
