Why I no longer explain Ethereum as a ‘World Computer’

Continue Reading

Talking points from my presentation — ‘Introduction to DAOs’

I’ve written down some talking points from my talk delivered during the Ethereum Meetup Warsaw on 28.04.2016. For TL;DR you can review the slides:

People Create Abstractions to Cooperate.

You don’t know who made your clothes. You don’t know who made the device you are reading this on. The majority of physical and virtual objects we interact with on a daily basis were not created by one person. Look around and realise that what you are seeing is a result of a coordinated economic effort of large numbers of people.

How can we coordinate large numbers of individuals to work towards the same goal? Humans invent abstractions that allow them to cooperate. Money, joint-stock companies and even the nation states are inventions that people use to coordinate their efforts.

These abstractions that don’t really exist outside of the human mind, allowed people to align their economic goals and orchestrate efforts to produce all things you see around yourself.

Pooling Resources And Risk/Reward Distribution.

Inventions like the joint stock company created mechanisms of economic alignment that allowed multiple individuals to share risks & rewards in productive ventures.

The joint stock company model evolved into global corporations but there’s a growing misalignment of interests between shareholders, employees and customers.

Digital Era Needs New Abstractions

Creating new productive ventures in the digital world is a high risk/high reward activity (startups). Most likely you’ll fail but if you win — you win big.

Most startups today are platforms that rely on decentralised value creation by their users. Work is decentralised (by writing these words on Medium, I’m creating value for Medium’s shareholders) but rewards are centralised — mostly in Silicon Valley because this is where the majority of platforms are initially funded and incorporated.

The industrial model distributed economic surplus via jobs & salaries. The digital age model distributes economic surplus primarily via equity in platforms.

But while everyone with a smartphone can participate in information exchange on platforms, only selected few can participate in platform’s equity.

Bitcoin — The Organisation Built on Software Based Incentives

Let’s forget for a moment about bitcoins as an asset class and focus on Bitcoin the ‘organisation’.

Bitcoin proved it’s possible to embed incentives in software so that the same level of coordination and efficiency is achieved without establishing legal entities, signing agreements, employment contracts etc.

In the industrial model their roles of the employer, employee, shareholder, CEO etc were fixed and expected contributions clearly defined. In the Bitcoin model, these roles are fluid. Users just contribute value to the ‘Bitcoin organisation’ depending on provided incentives and disposable resources they have (know-how, capital, appetite for risk etc).

Functionally, people just ‘do’ stuff for the Bitcoin network as they would for their employer. They provide IT infrastructure, write code, do marketing etc. They perform actions in the physical world because Bitcoin’s incentives make it profitable to do so.

Bitcoin proved that economic coordination at the global scale that was previously extremely costly is now possible with pure software. This is the real breakthrough of Bitcoin. The fact that the first ‘product’ of the ‘Bitcoin organisation’ is a money-like asset class is important but secondary here.

There will be many more Decentralised Autonomous Organisations (DAOs) that deliver other products to the world using the same paradigm that Bitcoin pioneered.

Ethereum is to DAOs What Blogger Was to Online Publishing

Bitcoin required specialised knowledge to experiment with other incentive models and its security model limited the range of experimentation to currency-like products.

Ethereum takes the idea of Bitcoin to the next level. It abstracts out the underlying platform so that you can focus on building other incentive structures. In other words, you can create coordination mechanisms for other groups of people, who can share risks & rewards in new ways than in the typical cryptocurrency scheme.

Blogger allowed writers to focus on what’s most important, the content and forget about HTML, web servers and IT infrastructures.

Ethereum does the same for DAOs. DAOs are not organisations as we normally imagine them (own stuff in the physical world, have employees etc). DAOs are pure economic coordination mechanisms. By distributing risks and rewards, DAOs can make people that never met each other contribute resources to the shared purpose.

We can say that DAOs achieve the same results as traditional organisations (creating value by combining skillsets and assets of participants ) but using somewhat different means.

Key questions when thinking about DAOs:

  • what roles does your model bring together?
  • what assets do they exchange? (talent, know-how, capital etc)
  • what is the product of this exchange?
  • who absorbs the risk of failure?
  • who gets rewards when successful?
  • what success means in your model?

Ethereum as a ‘World Computer’ is a bad metaphor.

While technically accurate it creates wrong mental associations for most people. (outside of the narrow group of insiders) It makes people think about ‘computation’ as an application of Ethereum.

But computation on Ethereum is slow and expensive. For computation, there are better, cheaper and faster alternatives that don’t require blockchains and all that.

Ethereum’s value proposition is not in computation.

Ethereum can create various forms of economic alignment, shared purpose and coordination between thousands of anonymous people, at a fraction of a cost compared to alternatives (legal frameworks). There’s no alternative for that currently.

Let me give you an example.

You can read this post and if you agree with me, we can have a shared perspective on the world. But we don’t have shared economic goals.

Most likely, we live in different countries, get paid in different currencies, work for different employers. No shared economic goals here.

But if we purchase the same DAO token (BTC, ETH or other) , suddenly we both have a vested interest in success of that DAO. We don’t have to know each other, speak the same language but still we are economically aligned proportionally to our stake in a DAO. We start reading the same websites, join the same reddit communities, tell our friends the same stories about how great the DAO is.

From Social Networks (Web 2.0) to Incentive Networks (Web 3.0).

Cryptographic tokens represent various risk/reward profiles and their holders create various incentive networks.

Web 1.0 — Static Document Networks — web of static pages connected via hyperlinks.

Web 2.0 — Social Networks — web of people connected via social links (likes, follows). 

Web 3.0 — Incentive Networks– web of people connected via economic links (tokens). 

Each subsequent paradigm affects the behaviour of the previous one. Web 2.0 influenced the way Web 1.0 looks like — social networks affected the way websites are designed.

Web 3.0 is already affecting Web 2.0 — just look at what happens on reddit during crypto bubbles, crashes, blocksize debates etc. The amount of noise will only intensify when DAO proposal debates begin. To fix that we need to be able to see incentives behind behaviours.

‘One Proposal Per DAO’ Governance Model

Continue Reading

This is a continuation of the article “Forking as a Governance Mechnism for DAOs

What if the DAO could self-govern itself using only the external mechanisms or ‘forks’?

What if there’s no voting, no formal proposals, and each DAO is created with a single, specific goal in mind? Once that goal is achieved, the DAO forks itself into a new one, incorporating all the learnings from the previous iteration.

Here’s the DAO framework that incorporates all these assumptions.

The framework is split into three phases:

  • creation — this is where DAO objectives are proposed, DAO gets created and invested in
  • execution– the core team works on delivering the Return-On-Investment for the DAO
  • evolution– if the execution is proving successful the DAO evolves via subsequent forks.


  1. A Founding Proposal gets submitted to Reddit, Slack, website or elsewhere. What really matters is that it happens ‘off-chain’. The Founding Proposal outlines what the Founders aim to achieve, expected timeframes for product/service delivery and most importantly — the expected Return-On-Investment projections for the potential DAO Investors. However, the DAO doesn’t ‘know’ anything about the proposal. The proposal is part of the social contract between Founders and Investors, but it’s not the on-chain contract.
  2. The DAO creation address gets published and the dates of the DAO creation phase are announced. The share of DAO Tokens that will be assigned to Founders is clearly announced to Investors. Before you scream — ‘premine!’ or ‘that’s not fair’ please stay with me. The Founder’s share is critical to this model.
  3. During the DAO creation phase, investors can purchase DAO Tokens. For example: if the sale is conducted using Ether, DAO Investor’s Ether is locked in the DAO Fund and the corresponding amount of DAO Token is released to the Investor.
  4. Once the DAO creation phase is over, the issuance of new DAO Tokens islocked forever. There won’t be any new DAO Tokens created outside of the DAO creation phase. All collected funds sit in the DAO Fund. There is only one way to access these funds — to return the DAO Token to the DAO and redeem the proportional share of the funds. The DAO Token gets ‘burned’ and the total amount of tokens is reduced. Later on, the only way to rejoin the DAO is to rebuy the DAO Token from existing members.


  1. The Founders who now hold the DAO Token have two ways of financing the Proposal: 1/ sell the token on the open market 2/ return it to the DAO and redeem their share of Ether from the DAO Fund. Founders have no other way of accessing the DAO Fund than selling or ‘burning’ their Founder’s share of DAO Tokens.
  2. Let’s assume that Founders deliver on their promise and their Proposal turns into a working product or service. Revenues are generated and collected in the DAO Fund. The value of DAO Tokens goes up accordingly as the DAO Tokens are backed by more cash in the DAO Fund. In addition, the revenue model is now proven so the market puts a premium on the DAO Tokens. Early investors can sell with profit or hold.
  3. If the Founders don’t deliver, Investors withdraw their funds by burning their tokens. Their investment is reduced by the Founder’s share which by this time is most likely gone. But their share of money sits safely in the DAO Fund and is only diluted by the Founder’s share.
  4. DAO Investors continuously ‘vote’ on the original proposal by either holding their DAO Token or cashing out (via ‘burning’ or selling).Their ‘vote’ is reflected on the blockchain and in the DAO Token market price. Effectively, this becomes a form of a prediction market on whether the Founders deliver the promised ROI.


  1. How does the DAO evolve, raise more money, bring in new investors and expand into new markets? It’s simple — by spawning a new DAO!
  2. The original DAO will have a clear economic goal and a pool of stakeholders who invested in the original idea. Due to the nature of open source systems, the value created by the original DAO will be publicly accessible.
  3. At some point, a new proposal will be introduced by the new or existing Founding team. The new proposal will use learnings of the original DAO but introduce new Founders, technologies, marketing techniques etc. And we go back to the Creation Phase again.
  4. Stakeholders of the original DAO can ‘vote’ on the new DAO by acquiring a stake in it. Some will burn or sell their original DAO Tokens to acquire the new DAO Tokens. Market mechanisms and people’s drive to maximize profit will quickly put the price on the new DAO.
  5. In the ‘real world’, the new DAO doesn’t have to be competitive to the original DAO. It can be created by the original Founders or stakeholders to expand into a market niche or introduce some experimental tech. The main advantage of this approach is that they don’t need to seek consensus. Interested investors will then hold both DAO Tokens providing funding to the new venture. Others will stay with the original project.
  6. From this perspective, individual DAOs become just accounting mechanisms used to distribute risks/rewards of different projects linked by common goals and stakeholders. Using the accounting terminology, they become cost & profit centers of the ‘meta-DAO’. Each has a specific purpose, a founding team, and a group of investors. Their profits and losses can be audited by anyone on the blockchain. Market mechanisms decide what gets funded and buying/selling replace voting.

To avoid complexity, I’ve intentionally left out some interesting modifications that can be applied to this model. We can explore them in more detail in the coming articles.

Benefits of this approach

  • 51% attacks are impossible — there’s no voting on spending decisions (investment in the DAO becomes the spending decision)
  • emergent, on -chain accounting system — each project becomes its own, on-chain Profit & Loss center.
  • permissionless innovation — no need to seek consensus within the large group of investors. If your extension of the original DAO is investable, then just introduce the fork and people will jump. If it’s not investable, they won’t — the market decides.
  • multiple founding teams possible — multiple people can take leading roles in various ‘branches’ of the DAO.
  • more granular market pricing of DAO components — instead of holding 1 DAO token, investors hold a portfolio of ‘subDAO’ tokens that represent their bet on financial success of various branches of the DAO.
  • potentially interesting viral loops —any revenue stream of the DAO is immediately reflected in the token price (the DAO token is backed by more funds) which draws attention to the DAO. Well-performing DAO tokens could become the most expensive ones because investors rather sell than burn and buyers will always pay at least the price equivalent to token backing (which is continuously increasing if the DAO performs).

Nothing new under the sun

The presented model mimics how the blockchains have been evolving so far. Multiple, often competing groups are working on multiple blockchain implementations replicating what worked for their predecessors and testing experimental features. Investors finance these projects by buying & holding their tokens.

From the investors point of view, every upcoming crowdsale presents a new dilemma — should I hold the ‘original’ token or diversify into the ‘child’? From the developers point of view, every success of the ‘original’ creates an incentive to launch their own iteration.

The majority of these projects will result in failure, loss of investment and disappointed ‘bagholders’ holding worthless tokens. But isn’t it similar to how the startup ecosystem works? The beauty of open source is that all the learnings from the failed projects stay in the public domain and help innovators who come in later.

This will continue to happen in the universe of Ethereum DAOs — but at an accelerated pace.

So the twist here is that rather avoiding the inevitable, let’s design systems that expect to be forked.

I’d love to hear your feedback regarding potential attack vectors on this DAO model.

Ideas presented in this article were developed in collaboration with Kuba Kucharski. We spent many hours discussing Ethereum, DAOs and ways to scale attention marketplaces described in the Userfeeds article. We’ve realised that the model we design for Userfeeds can be applied more generally to other types of DAOs. Most of the ideas presented here are as much his as mine.

‘Forking Systems’ as a Governance Mechanism for DAOs.

Continue Reading

This is a continuation of the article “On Risks, Rewards and The Evolution of DAOs”, where I explored problems related to the scaling and evolution of DAOs.

The main point was that if the DAO achieves any level of success, the late adopters are actually more incentivised to fork the existing DAO rather than to join it.

Why would that happen? It’s related to how the evolutionary process in public blockchains works. If you have an idea for a substantial improvement, you then have to go through the internal politics of the established consensus, fight the power structures and community values to implement your idea. Most likely, you’re not a stakeholder in the system so you don’t benefit from the price appreciation if your idea works. It is way easier technically and more rewarding financially to take all the open source learnings from the parent DAO and start your own thing.

Let’s use some genetic metaphors to illustrate the evolutionary process in cryptos.

If you consider Bitcoin a parent DAO organism, this is how its DNA is spreading:

  1. Internal evolution. There’s a slow and painful internal process of Bitcoin Improvement Proposals, followed by politics and power games. This conservative model ensures stability and consistency of the network but discourages radical innovations. The recent blocksize discussion shows how hard and slow this process can be. The entire mechanism is optimised for gradual, incremental improvements that don’t threaten core values and preserve the existing stakeholder & talent pool.
  2. External evolution. And there’s a fast and chaotic land of Bitcoin forks. Things tend to are wild here— both ‘get rich quick’ scammers and genuine innovators operate in this area. Both PayCoin and Ethereum emerged here. Forks can be done quickly and don’t require the lengthy political process of the internal evolutionary model. Each fork introduces a radical mutation of the original ‘Bitcoin DNA’ and brings a new set of early adopters and with them — new values, new talents and a new set of vested interests that help the fork succeed.

Both mechanisms are necessary — we can compare the #1 to the evolution of a single organism during its lifetime. Once it goes past adolescence, it’s personality is more or less fixed and only incremental changes are possible.

The #2 is like an offspring of a parent organism — it’s a fresh start, a reshuffling of a genetic pool. Some properties of the parent organism that proved to be beneficial for survival are transferred to the child, but a new genetic mutation is introduced. Depending on the mutation, the child either evolves and passes its traits to the next generation or it dies and its genes exit the gene pool.

We can think of the Bitcoin’s whitepaper as a genetic blueprint for a certain type of a network-like organism. The original Bitcoin network is a parent organism that proved to be fit to the environment in many ways. While it’s still evolving internally, it’s DNA is already spreading faster via ‘offspring networks’.

‘Offspring networks’ can pull in new resources (talent, capital, attention) and exploit new environmental niches (market verticals) faster than the parent network. Most importantly, they can be individually priced and evaluated by the markets, separately from the parent. This crypto-Darwinian process is quite unforgiving though. The child mortality is high as many clones don’t really survive the first year.

But every now and then — the prodigy kid like Ethereum emerges. And as soon as it shows some positive survival traits, it immediately gets cloned again.

You can go to coinmarketcap.com and have a look at all the coins. Each one of them contains the DNA of the parent Bitcoin. Some of them are genetically close, others evolved to be quite remote. On the fundamental level though, each one of them is built on the common design pattern of public/private key cryptography, distributed consensus and peer-2-peer architectures. And because of that shared foundation, value can easily flow between them using services like Shapeshift or other relaying technologies.

For the experts, it’s all separate networks under the hood.

But for the end user, it increasingly becomes one value-network, the Internet-of-Chains, where they can easily hold many coins and bet on the evolutionary success of individual branches of the crypto-tree.

What I described is, in fact, an emergent governance model for blockchains. The blockchain organism governs itself as it evolves in multiple directions.

Once the internal evolutionary forces of a particular network are not sufficient, the network just forks itself into a new branch.

This phenomenon is autonomous and unsupervised, driven by human egos, imagination, greed, sense of community etc. Once a new branch is established the same human emotions help it grow or let it die. It’s an endless cycle.

What Ethereum does with DAOs is that it abstracts this crypto Darwinian, evolutionary process from the underlying delivery mechanisms. So now the evolution can continue much faster, as it will take place on the common shared infrastructure.

We can expect that DAOs will experience the same evolutionary pressures as the blockchains did.

So let’s start a discussion on how we can anticipate forks, how to embed them into the governance models and what they mean to the crowdsale investors.

Voting systems are the internal evolution mechanisms (the equivalent of BIPs, EIPs) and without a doubt they are a necessary component.

But let’s not forget about the ‘forking systems’ which will allow the external evolution of the DAO.

To be continued….

Ideas presented in this article were developed in collaboration with Kuba Kucharski. We spent many hours discussing Ethereum, DAOs and ways to scale attention marketplaces described in the Userfeeds article. We’ve realised that the model we design for Userfeeds can be applied more generally to other types of DAOs. Most of the ideas presented here are as much his as mine.