Rhodri Davies, Programme Leader, Giving Thought

Rhodri Davies

Head of Policy

Charities Aid Foundation

The role of giving


09 May 2017

(NB: This is blog about our Giving Thought discussion paper Losing the Middle but Keeping the Heart: Blockchain, DAOs and the decentralisation of charity .)

When it comes to the potential impact of blockchain technology on philanthropy and civil society, the focus has so far been largely on three areas:

  1. Reduction of transaction costs through disintermediation (e.g. Disberse)
  2. Automating processes through the use of smart contracts (e.g. Alice.si)
  3. Boosting transparency through asset tracking (E.g. both of the above, plus the BitGive Foundation’s GiveTrack)

We have explored many of the implications of these (and other) features of blockchain in the context of philanthropy in a series of blogs and papers.

But it struck me recently that where there has been less focus up to this point is on the broader question of decentralisation, and what this might mean for civil society. We flirted with this issue in our paper Block and Tackle, where we looked at the possibility of creating charitable entities on the blockchain; whether these could be structured as Distributed Autonomous Organisations (DAOs); and what the implications might be for charity regulation.

However, developments such as the emergence of a new wave of DAOs following the hack of the original DAO last year, the creation of blockchain prediction markets such as Augur and Gnosis, and the massive rise in the number of Initial Coin Offerings in place of traditional IPOs have led me to think that we didn’t develop these ideas far enough. (I won’t get into the current debate over whether the current ICO gold rush represents a bubble, except to say that even if it does blow up that doesn’t mean that the DAO model is inherently flawed. After all, the South Sea bubble of the early 18th century went a long way towards establishing the joint stock corporation model that paved the way for modern corporate business).

That is why we have just published a fourth paper in our series on blockchain and philanthropy, Losing the Middle, Keeping the Heart: Blockchain, DAOs and the decentralised future of charity. In this paper, we look at various ways in which decentralisation will impact on the nature of civil society and the organisations that work within it, including the following:


One way in which blockchain can be put to work in a philanthropic context is by disintermediating the transfer of money. Currently, to get money from donors to end beneficiaries (particularly in an international context) involves a long chain of banks, NGOs, government agencies, law firms and so on. Each of these adds transaction cost, and thus reduces the overall amount available for the intended social purpose at the point of dispersal.

Blockchain offers the possibility of disrupting this chain, because it reduces or even removes the need for third party authorities. There are already a handful of platforms looking to take advantage of this potential (e.g. Disberse as already mentioned, and in a broader sense, Humaniq). But in the future, we might not even need this degree of centralisation: we could take the models of direct giving exemplified by GiveDirectly.org or direct social lending exemplified by Kiva, and disintermediate them. If you assume widespread adoption of cryptocurreny or blockchain in some from, then it would be possible to operate such a direct giving or lending model at scale without the need for centralisation in order to manage transactions. Which is not to say that there is no value to having a centralised infrastructure: it may carry significant benefits in terms of advice on where to give, how to maximise and measure impact etc, and therefore be entirely worthwhile. What is different is that we now have to be much clearer about what that value is given that there is no longer an assumption in favour of centralisation.

In the future we will have to be much clearer about what added value there is to centralisation, given that it will no longer be necessary.


Some of the most exciting avenues of development for blockchain technology, to my mind, are the attempts to use it to effect radical overhauls of governance and decision-making structures. If you are creating organisations as entities on the blockchain (as detailed in our previous paper Block & Tackle), why do you need them to adopt structures that reflect the practicalities and limitations of the physical world? Why not just go the whole hog, and just decentralise the organisation itself?

That is where the idea of a Distributed (or Decentralised) Autonomous Organisation (DAO) comes in. A DAO is essentially a collection of smart contracts (self-executing compute protocols) that govern the interactions of individuals who choose to become members of the DAO (usually through the purchase of ‘tokens’ or ‘coins’). The basic idea is to incentivise those behaviours deemed to be desirable (e.g. developing stock picking algorithms in the case of Numerai, or accurately predicting and reporting real world events in the case of Gnosis or Augur), whilst disincentivising undesirable behaviours (e.g. dishonesty, self-interest etc.). A mechanism for joint decision making can also be established based on whatever form of consensus is deemed to be most desirable (i.e. simple majority, weighted etc.)

Once you have this framework in place, you no longer need a centralised corporate structure. Decision making among peers is now practical, even at a large scale and with a wide geographic spread. In the paper we explore whether this opens up scope for creating DAO cooperatives or DAO giving circles, and the various ways in which this could work.

The concept of a DAO opens up the possibility of taking existing collaborative structures such as co-operatives or giving circles and scaling them up indefinitely without the need for centralisation.

We also consider the challenge of maintaining points of contact with the real world: i.e. how would you know that any social outcomes recorded on the blockchain, and triggering payments via smart contracts, had actually been achieved? We propose that there are at least three different possible solutions to this problem, all with their own advantages and disadvantages which we explore further in the paper:

  1. Rely on a network of trusted individuals or organisations (often referred to as “oracles”) to verify that the semantic link between the blockchain representation and reality is correct (i.e. get existing known charities or other organisations to validate outcomes).
  2. Integrate with other technologies (e.g. Internet of Things, drone tech, wearables etc) to automate the recording of data that can directly validate the achievement of outcomes.
  3. Incentivise members of the DAO to record information truthfully using a system of punishment and reward based on tokens/coins, along the lines of the systems used by prediction market DAOs like Augur and Gnosis.


The paper also considers the convergence of AI technology and blockchain, and the fascinating and radical possibility that you could have DAOs in which some or all of the members are AIs (so-called AIDOs). These AIs may be embedded in smart objects or they may be simply algorithms that exist only in digital space.

Why, you might ask, would you want to allow an AIDAO? Well, one reason is that the additional convergence with the Internet of Things means there are going to be vast numbers of smart objects and appliances operating in the world in the near future, and these will often need to coordinate. Using the current paradigm, they would have to be controlled by a centralised human bureaucracy, but in this convergent future there is no real need for that to be the case. Why not simply let machines and smart objects interact and transact directly with one another? Assuming that AIs aren’t precious about job titles, it seems unlikely that a hierarchical structure makes any sense: instead they would all operate as autonomous agents linked together in a DAO.

There are clear examples of industries that could benefit enormously from the introduction of AIDOs. For example, car sharing. Currently, Uber is on the verge of effecting a major shift towards the use of driverless cars; which has some major advantages, but also one massive disadvantage – capital cost. At the moment they don’t need to own any cars, because the drivers own their own vehicles, so Uber essentially just operates a network layer on top of that.

But if those cars are driverless, who owns them? The current model won’t work, as it is likely most people will not see the point of owning a driverless car; choosing instead just to have access to one when required. So Uber either has to rent them from another company, or purchase a vast fleet of vehicles itself. Unless… What if those vehicles were autonomous agents, able to generate and own their own revenue, and connected by an AIDAO? Uber (or any other car access company of the future) would simply deal direct with the DAO.

All of the benefits of this scenario apply just as much to charities. A big development NGO, for instance, might currently have to operate a huge fleet of vehicles to get supplies about and warehouses in which to store those supplies until needed. But those could be replaced by an AIDAO-owned warehouse of smart objects, or an AIDAO-owned network of vehicles, and the NGO could simply interact with the AIDAO to direct the distribution of required goods.

The future possibilities presented by decentralisation and the convergence of blockchain, AI, IoT etc are mind-blowing. And if those of a charitable mindset can harness even a fraction of this potential it could transform the ways in which we address social and environmental issues.

Of course, we have already seen that the NGO itself could be restructured as a DAO too.So we can take our DAO giving circle to its logical conclusion. We focussed before on the movement of money to those in need, because that is traditionally how charity has operated: you give money to organisations or individuals to enable them to buy things that do social good. But what if the purchase and distribution of goods themselves is decentralised and integrated with this system too? So it is just as easy to give physical objects (or access to them) directly. Hence, our philanthropic AIDAO consists of a combination of humans who make decisions about where to give based on a mixture of rational and emotional considerations and AIs with whom they interact to effect the distribution of the necessary goods and services to achieve the desired social outcomes.


The future possibilities presented by decentralisation and the convergence of blockchain, AI, IoT etc are mind-blowing. And if those of a charitable mindset can harness even a fraction of this potential it could transform the ways in which we address social and environmental issues. But equally, charities and NGOs (and perhaps more importantly, the people who work in them) need to make sure they are in a position to play the part of human overseers or ‘ethical handbrakes’ on the operations of technologies that otherwise might generate unintended consequences at a pace and scale that is truly frightening.

For now, the main thing is that we start thinking through some of these issues. So why not download our discussion paper and get in touch if it sparks any thoughts?