Articles

Atomflow: Decentralised Trust and Collboration

Published on May 11, 2019. Originally on Aug 6, 2018
Image by Joanna Kosinska

Published with permission. Please see the original article on Atomflow.

Centralised Systems

A centralised system is one in which a central controller exercises control over other components or stakeholders of the system either directly or through a power hierarchy (loosely referencing [1]). Graphically, a centralised system may be depicted as follows:

Interestingly, such systems surround us in everyday life. Most likely you had interacted with a bank, having a bank account, for transfer purposes, for crediting purposes, for brokering purposes or for other reasons. In such an interaction, you are the black small circle in the picture above and the bank is the reddish circle. Having a bank account is the simplest, you trust the bank for your money and the bank offers you its service to keep the money safe and available when required. This is a classical example of a centralised system. The bank controls the funds, sets the rules and different clients and customers interact with each other through the bank - the centralised authority. Banking throughout history would not be profitable without trade and exchange of money, goods and assets. Loaning is an integral part of banking as well, which is necessarily related to accumulation of monetary resources through different activities and in different forms.

This lengthy introduction is here to say that such a central authority exists for reasons. Profit generation, regulating and mediating interactions and keeping the process in check. For the good or bad, this is not the issue we are concerned with here. The point is that a bank's role is not intrinsically different from the server you access to read this very page, except for one difference. Both are centralised systems, however, a bank mediates and dictates the rules for interaction between multiple stakeholders and it is always the centre of authority, this server mediates the interaction only between you and itself, between a different individual reading this article and itself, etc. If you want to opt out and never read the page again, you are totally free to do so, the same goes for any other individual. Moreover, you can reach for any other individual or entity and set out to discuss and act upon information given on this page in your own. You do not need this website to mediate the interaction between you and other parties. This freedom or control you have and, everyone else has, is not present in the case of a bank. The bank operates by the rules of financial systems that are very tightly regulated and controlled. So whilst your choice to take your business to another bank is up to you in most cases, it is not possible to interact with another bank's client directly or opt out from the centrally nominated interest rate, for example. This is obvious, but there are other subtle things, your residential address is almost always a requirement for having a bank account, also a utility bill is most likely is. The system dictates subtle rules and in doing so these rules spread far and wide throughout different aspects of everyday life. The point here is that central systems impose rules and possibly limitations and make systems and services built on top of their infrastructure limited to and by those rules.

The centralisation model of this website and its users can be, roughly, depicted as follows:

The website does not act as a middleman, it is only authoritative on the information it produces and is not able to mandate how people interact with such information unless by mutual agreement that you can accept, reject or opt out from and have other choices. Still a centralised system in location and authority of information, nonetheless, if we exchange stakeholders in this system such that every stakeholder can produce and consume information with equal opportunity, rights and powers then it would be a decentralised system. We will talk about this later.

The ramifications of central systems are experienced by almost everyone everywhere. Google, Facebook, Amazon, Microsoft, etc. are central authorities and they centrally control how billions of people and businesses go about their daily business, accumulate and control an exceedingly huge amount of information and they are able to influence our choices and interactions with information and almost everything we do online. Being good or bad is not the concern here but centrality is. The choice and the framework, legal and social, to act in good and bad ways are dynamic issues and they differ according to different criteria at different times. We believe that most people would not be happy if one morning we find out that the legal code allows a company like Google to free to run a retail business tomorrow and use all information it has about individuals to affect their purchasing choices and dynamically set prices. In fact, such a scenario has happened when Internet service providers, ISPs, were given the right to use their customer data for monetisation purposes and it had happened and is happening everyday through making data available to a handful of companies to monetise alongside our individualistic behaviours. With advances in artificial intelligence, this is becoming more and more alarming.

In fact this is similar to the problems with the banking system and mainstream financial systems as a whole.

Some people think of banks, for example, as big safes, which is correct to a degree. Although it is only a valid assumption if there are guarantees that the perceived value of what is stored in the safe remains the same or increases over time and that the rules cannot be changed by one side only. Interestingly, modern times and crises prove that this is not always the case, or ever if parties controlling the system start to miscalculate or misbehave.

Middlemen

Mediating interactions between one party and another by a third party is something we got used to, however inefficient it makes most interactions are in terms of time, cost and authenticity or correctness of information. Retail and wholesale business division is largely instituted by this sort of mediation. The channels from farmers and producers and to consumers are usually very segmented because of the middlemen effect. This is also found in banking and financial services, auditing, advertising, legal and corporate services, and other business sectors. Middlemen exist largely because of system centralisation and lack of direct channels between parties either physically or as related to the information available to the parties.

Middlemen, especially, usually affect ecosystems in negative manners unless they add fundamental or intrinsic value. Think of individuals, small and medium businesses, society development, education and non-profit organisations gaining access and making use of readily available direct channels between the multiple stakeholder they serve and target without the additional cost and know-how factors and with direct reach to their stakeholders. Also, for large corporates and their clients or consumers, the efficiency of direct reach, available information and accountability would have a remarkable gain for the system as a whole. Such transparency, availability and accountability as well as the ability to have choices make for equal availability of economic opportunity and higher business competency.

Trust

So why do we trust banks, large corporations and middlemen? Reasons differ but usually three main reasons are:

  • legal and executive infrastructures that we have to abide by whether they make sense or not, if we want to change them, it almost always takes a political movement or party to do so, instituting the centralisation of authority in the hands of the few.
  • Need and lack of satisfactory alternatives; not approving a prominent search engine's or social network's or another sort of central corporation's policy does not mean it is readily possible to replace it as providing such a service would require massive investments.
  • Lack of direct channels which are sometimes a system design goal so that direct interactions are ineffective or impossible.

Such trust models more often than not, and sometimes purposefully, lack transparency and the ability to verify information and actions in easy or direct manners.

Decentralised Systems

The opposite of a centralised system is, not surprisingly, a decentralised system. In a decentralised system of multiple stakeholders having equal status, each stakeholder has the same degree of control and authority. It may be also the case that stakeholders have different roles or degree of control but there would always be multiples of such stakeholders. Information is processed without control or mandate of a central authority. In a self-organising decentralised system, stakeholder holders individually or collectively process information so that the system as whole reaches its goals without a central order. Without going into technical details, a deterministic self-organising system is a dynamic system that automatically evolves towards a state of equilibrium [2]. It is an autonomous system. A decentralised system may be graphically depicted as follows:

Applying the decentralised model to banking, each stakeholder, be it a service provide or a client, be it the current bank owner or you, has access to information without access limitations, has control over their funds without mandates. To put things in context in the case of monetary funds, other assets of value, such as services, each stakeholder has the ability and control to make them safe and readily available without central help and has direct channels to transact with any other stakeholder. Peer-to-peer systems are such systems that employ direct interactions between peers, or stakeholders, to attain the individual or collective goals of stakeholders.

Similarly, a decentralised social network, is a network that does not have its information stored, processed and controlled by a single central authority. Rather it has the information and their processing distributed amongst multiple nodes, or stakeholders.

A distributed system is a system which is decentralised and which has no single point of failure. In the decentralised financial system and social network mentioned above, it means that it does not affect the health, usability or the functioning of the system when multiple nodes fail and that other nodes or the failed nodes, after recovery, can rejoin the system and function correctly.

Blockchain

Blockchain is a peer-to-peer technology which uses cryptography - mathematical rules and formulas - to maintain integrity of information in the system in a decentralised manner without depending on a central authority. Simply put, a blockchain is a decentralised system in which peers collectively work to produce an irrefutable series of information called blocks that are agreed on to represent the sole and collective authority of information in the system. A blockchain may be depicted as follows:

Blocks - of information - may be produced, or mined, by anyone in the system but all in the system must agree on the authenticity of their content. Each block depends on the previous block for authenticity and correctness, hence, the name blockchain.

Niether decentralised nor peer-to-peer systems are new, cryptography is very established field, but what the blockchain brings to the wide computing and business communities is a trustless model that works. In other words, a consensus model for people to to agree about things without trusting each other that works and can be used to do business. Trustless models are also not a new idea as well. The Byzantine Generals' Problem [3] that finally greatly affected blockchain technology as it exists today is a consensus problem and solution described in 1982 by Leslie Lamport. However, the innovation and success of implementing this sort of technology has been driving waves of conflict between central systems as well established as they are today and the new technology with all sorts of peer-to-peer possibilities that can overturn established systems.

It takes 51% stakeholder, or those with computational power, in the system to break the chain and change "the truth", i.e. information that the other 49% have previously agreed on. In other words, it take 51% of the computational power of the system to make a fraudulent transaction for example. The design of the system aims to make it more costly than beneficial for attackers to carry out their attacks. As the system grows and as blocks continue to be produced it is more and more difficult and less probable to be able to change agreed upon information. If it sounds scary, let's consider the bank's case where it take one or a handful of stakeholders to carry out a huge amount of fraudulent transactions, especially when such stakeholders represent the interest of the central authority. In fact, in a traditional voting system, it takes as few as one to a handful of stakeholders to change the outcome, again the central authority.

As blockchain technology allows trustless collaboration, cryptocurrencies were born. Cryptocurrencies are virtual currencies which are produced according to the blockchain rules and to everyone's verification and agreement. There is no central authority to issue, value or control the currency, rather they are a peer-to-peer instruments. As financial things make headlines, cryptocurrencies got most of the fame but the truth is, it is the blockchain technology itself is what of more importance.

Nonetheless, blockchain brings many problems as well. Energy efficiency is of utmost importance because it matters to us all. Bitcoin mining - block production - is estimated to use more energy a year than Ireland and could match Austria by the end of 2018 [4]. Whilst the market capitalisation of Bitcoin is currently estimated at USD 130.257 billions and Ireland's nominal GDP is USD 325.831 billions. Whilst the importance of this number may be debated, the combined utility and consumption of the numerous cryptocurrencies is more important for certain.

Moreover, there are sensitive privacy and confidentiality problems with the technology. In order to maintain integrity of information, mainstream blockchains make every bit of information public. The counter measure to this is what is called pseudonymity: information is publicly associated with opaque identifier but these identifiers are not associated with a personal identity. This is alarming for, in fact, any leak of information that can link a person to an opaque identifier gives out all information available about that person. Imagine a marketing company having access to your transactions or a crediting service or imagine that all of your transactions are published online. This is dangerous, even if you are careful, any agency with access to the start of the rope, reaches the end. Furthermore, maybe, artificial intelligence will be targeted to try inferring the association. This limits the use of blockchain technology for privacy-sensitive applications and services.

Blockchains are also not general-purpose computing platforms but rather either specialised for one or more purposes or are suitable to specific clusters of applications. Current landscape of blockchain technologies looks like isolated islands although they are much related and in fact they affect each other but it is difficult to build a system on top of a blockchain to talk to another system on another blockchain.

Further, whilst privacy is an intrinsic problem, the amount of data to replicate by almost every stakeholder to preserve the integrity of all information, relevant or irrelevant to a stakeholder it may be, is a problem as well. Another intrinsic problem is that blockchain technology, somehow, asserts what it was built to avoid: having choices. One is again left with a single choice to do things in the way they are done on the blockchain implementation (system) as much as one is obliged to go by the bank's rules. The alternative is to start your own system, which is a huge investment with high entry barriers, for example, know-how.

Most of the well known cryptocurrencies are in fact assets not currencies in the usual sense. They are not, usually, connected to real-life activities and they do not serve as payment systems. Moreover, they usually acutely fluctuate.

Yet a major problem with blockchain is in fact centralisation which current blockchain network were supposed to eliminate. Both Bitcoin and Ethereum, the two major blockchains, are not decentralised systems in the true sense of the word. This is evidenced by the pattern of mining pattern and where mining originate from, their cryptocurrency pricing dynamics, etc. Few individuals and entities have controlling computational power in the networks, which somewhat equates to authority. Also a small percentage of stakeholders are in possession of large amounts of the networks' cryptocurrencies which makes it easy to affect their prices. The following material go into details of such centralisation from different perspectives:

  • A paper [5] studies centralisation aspects of both Bitcoin and Ethereum.
  • A article [6] wraps up the previous paper.
  • A more technical article [7] about the centralisation of Bitcoin.
  • An article [8] summarises the state of Bitcoin mining.
  • An article [9] that discusses effect of centralisation on cryptocurrency (Bitcoin) pricing; 1,000 own 40% of the market.

What Atomflow Brings

Technology

Atomflow takes a very different approach to the problems stated above. Atomflow is an infrastructure for decentralised ecosystems and is a financial system. Its main values are providing equal opportunity and providing the ability to have the choice.

Atomflow is a general-purpose distributed computing cloud that is intelligent and consistent. It is a peer-to-peer decentralised system which makes it possible for individuals and entities to have their systems and networks without the cost or know-how barriers and without limitations of blockchain.

In this sense, Atomflow is an infrastructure to run and deploy chains, or networks or systems, as a service. Think of a cloud computing platform that makes it possible for people to have their cloud up and running from a laptop. The difference is that Atomflow is decentralised and is not owned or controlled by a central authority. Another difference is that Atomflow allows anyone to run a chain of their design according to their requirements using a mixed mode of remote resources and their own.

Atomflow does not impose a system structure on services and ecosystems developed on top of it but it abstracts such systems' requirements and technicalities as an internal network design aspect. In other words, a system developed on top of Atomflow can be designed freely without being limited by how Atomflow works. It is possible to develop trustless as well as trust-based services, for example. It is possible to use cryptographic methods suitable for the service or not use any at all. To develop a network as public or private in which case the network is only accessed by those who are given access. In both cases every service can choose to let other services communicate and interact with it. Nonetheless, Atomflow always provides the infrastructure for computing, communication and collaboration.

Atomflow eliminates privacy problems associated with centralised systems and blockchains by making it possible to choose different levels of anonymity, public presence and confidentiality of information on the main chain, the atomchain, and inside different networks. Information is not publicly known in the system unless it is chosen to be so, nonetheless, it is always integral and verifiable. Every stakeholder is able to choose how they interact with and share information with others and the requirements for that interaction to happen. Atomflow makes it possible to develop regulated and unregulated services on top of it whilst making it a choice to join such a network and to opt out from it. At the same time Atomflow gives the choice to service providers to provide their services within constraints that people who want to use their services have to agree to or reject. Hence, whilst Atomflow promotes new models of doing businesses and establish systems to employ such models, it enables already established system to integrate with the new systems and to transform themselves into new systems.

The system as a whole is a self-organising system that autonomously deters demand and supply manipulation and centralisation by means of intelligent agents that autonomously learn, act and adapt to preserve equilibrium in the system. Moreover, the system limits replication of information whilst maintaining integrity and accountability. Furthermore, the system employs adaptive energy consumption strategies and consensus models for different tasks to be a sustainable energy efficient system.

Financial System

Atomflow is a self-regulating financial system which uses a cryptocurrency, named Ely, and a higher-valued crypto-asset named Okean. The system makes it possible to transact, trade, do business, make payments in peer-to-peer manners which may be regulated or unregulated. Transactions can be made public or private, anonymous, pseudonymous or public. Ely is aimed to be an efficient global payment system suitable for everyday life activities and business transactions.

Transaction data, sender, recipient and worth are all, by default, confidential to parties between which the transaction takes place. At the same time, by means of cryptographic methods, a transaction is always verifiable, can be audited and accountability is preserved. Access to a transaction data is controlled by the stakeholders having the transaction and the context they are transacting within. Furthermore, policies can be attached to transactions by peers to control sending, receiving and characteristics of a transaction.

Atomflow makes it possible to act as a guarantor for transactions still to take place in the future or guarantee resources that exist outside of the system in an integral and transparent manners. This makes it possible for Atomflow to manage transactions inside the system and off-the-system transactions.

Ely and okeans are autonomously priced according to activity levels, transactions worth and volume and business performances. The system, continuously, tries to reach equilibrium point for ely so that it is sustained as a currency that facilitates fair transactions and preserves expectations for transacted resources' value. Ely and okean are designed to be connected to real-life economic activities.



References

[1] Bekey, G. A. (2005). Autonomous Robots: From Biological Inspiration to Implementation and Control. Cambridge, MA: MIT Press.
[2] Ashby, W. R. (1962). "Principles of the self-organizing system", pp. 255–78 in Principles of Self-Organization. Heinz von Foerster and George W. Zopf, Jr. (eds.) U.S. Office of Naval Research.
[3] Lamport, L.; Shostak, R.; Pease, M. (1982). "The Byzantine Generals Problem". ACM Transactions on Programming
[4] John Leonard; "Bitcoin now uses as much energy as Ireland", 21st May, 2018. Fetched on 25th May, 2018.
[5] Adem Efe Gencer; Soumya Basu; Ittay Eyal; Robbert van Renesse; Emin Gün Sirer; (2018). "Decentralization in Bitcoin and Ethereum Networks"
[6] JP Buntinx; (17th January, 2018) "Study Shows Bitcoin and Ethereum Are Far Too Centralized Right Now". Fetched on 25th May, 2018.
[7] Egor Homakov; (17th December, 2017) "Stop. Calling. Bitcoin. Decentralized.". Fetched on 25th May, 2018.
[8] Steamit user nikul567; (June 2017) "Which Countries Mine the Most Bitcons". Fetched on 25th May, 2018.
[9] Olga Kharif; (8th December, 2017) "The Bitcoin Whales: 1,000 People Who Own 40 Percent of the Market". Fetched on 25th May, 2018.