The Great Divide

Introduction

TL;DR: the smart contracts are here. Controllable, repeatable and free to use, they permit design that addresses many of the commercial problems Bitcoin doesn’t – and will upturn a number of sectors, including crypto, as they advance.

This blog post is about dissecting Richard Brown’s outstanding blog post, Bitcoin as a Smart Contract Platform, from earlier today; if you have not read it I recommend wholeheartedly that you do so.

Richard writes:  

For me, it is a mistake to think about Bitcoin solely as a currency. Because the Bitcoin currency system is a masterclass in mirage: underneath the hood, it’s a fascinating smart contract platform.

There are, of course, a number of smart contract platforms out there – Eris among them – which utilise blockchain technology to achieve various aims. On which Richard observes:

Different design goals, different implementations.  And the value of such systems to banks, corporations or individuals is, ultimately, an empirical question. I imagine 2015 will be the year where we discover many of the answers.

Richard names 4 protocols/projects to illustrate these different design goals:

  • Bitcoin: a cryptocurrency, controlled (notionally) by no-one. 
  • Ripple: a cryptocurrency and a distributed ledger platform, with one ledger hosted on Ripple Labs’ servers.
  • Hyperledger: uses a consensus mechanism and a PostgreSQL database to track asset ownership (tokenised, I think – Meher or Daniel welcome to chime in and I can throw some edits in here).
  • Eris: My shop. We built ErisDB/Thelonious, an open-source blockchain design (a template, not a single chain) which is fully programmable and controllable, designed to give rise to millions of instantiations like any other open-source database – and able to link into other data structures through an open-ended client called the decentralised application server or Decerver.

The significance of these statements did not escape Ian Grigg:

“as usual, (Richard’s) words are prophetic and dangerous.”

Why they are dangerous may have escaped the more casual reader, who may be unaware of the finer distinctions which are (for the most part) not express, but implied. I break this issue apart a little bit below, in the form of three questions that I think are relevant in determining the eventual direction of this tech:

  1. What’s the difference between cryptocurrency and the new class of ‘distributed ledgers’?
  2. Who’s called “blockchain” use-cases right?
  3. Is Bitcoin real?

1) First question: What’s the difference?

Q: What’s the relevant difference between these projects (Bitcoin/XRP on the one hand, and Eris/HL/Ripple-as-a-service on the other)?

A: Where the value is.

With Bitcoin, assets are said to be “on-chain” – a Bitcoin is valuable because it is a Bitcoin, and is a much-desired scarce resource, thus acquiring use- and exchange-value in its own right. Ripple also has a component of this in the form of its XRP tokens, which also have exchange-value in the cryptocurrency markets (at last count, 1 XRP was worth $0.008511). All other blockchain datastructures which incorporate the “cryptocurrency” mining-driven security model adopt this approach.

By contrast, with Eris, we figured out how to make a blockchain controllable. Accordingly, the value of the chain is not in the tokens (we have named our tokens “Junk” to further illustrate the point/for the lulz) but purely in the functions these blockchains perform. 

We have made an illustrated example of why this makes sense:

If, as I do (and, it would seem, as Richard now does as well), one ignores the rhetoric around Bitcoin and looks at it as

  • a database,
  • a protocol/application design, and
  • a network of people with a common objective

…one takes the view that blockchain’s primary benefit to its users lay

  • not in decentralisation per se, but
  • in process automation with limited supervision in accordance with a known and (reasonably) certain set of rules.

Because blockchains are databases – and therefore software – they will do exactly what we tell them to do. It follows logically that, if we write the appropriate coded rules, it is possible to still obtain the benefits of their use, while scrapping the “fully trustless” currency piece. Money is just glorified data, after all.

In brief, the value is

  • (where the chain records value) in the fact that data reflecting value (e.g. a cash deposit or an obligation) is enforceable against someone, and
  • (in any case) cost savings to the person/organisation deploying it. 

Looking at “blockchain” as an open-source DB rather than a cryptocurrency also means that blockchains can be structured to be

  • accountable
  • controllable
  • more or less infinitely repeatable, and
  • reversible

…in the hands of their operators, such that it suddenly becomes practical from a business perspective to allow anyone, anywhere to create millions of blockchains (much in the same way they create millions of SQL databases today).

A blockchain’s a tool, not a money-printing press, in this view. We have built Eris to permit the full range – from a fully-open cryptocurrency clone to a totally locked-down single instantiation with one user – and every gradation in-between. It should be up to application developers, not us, to decide how they want their Thelonious blockchains to work.  

2) Second question: who’s called “blockchain” use-cases right?

That’s the $64,000 question. Who’s right about “the blockchain” –

  • the cryptocurrency camp, or
  • the (at the moment, vanishingly small) utility camp?

I don’t think these positions are mutually exclusive (nothing about a controlled chain does anything to stop Bitcoin from carrying on what it does) – or that the dominance of one or the other is a zero-sum game. For the time being, what we do know is that there is a substantial philosophical – or as Dave Birch put it, religious – division between the (supermajority) Bitcoiners and the (far and away minority) blockchainers in whether a blockchain data structure should be used at all in circumstances where a cryptocurrency hasn’t been attached.

If you talk to the Bitcoin set, the blockchain is all about “decentralisation” – the fact that third parties are cut out of the equation (even though the entire Bitcoin industry as we know it seeks to monetise Bitcoin by reinserting third parties into said equation – which is silly, really). Centralisation of administration is pointless, it’s argued, so anyone who would want to use a blockchain DB should just use a SQL database instead.

Of course I disagree with this view.

At EI we take a somewhat different (and in our view novel) approach: a blockchain is a database technology like any other (MongoDB, MySQL, etc.) and possesses certain characteristics which are commercially valuable. These are that

  • its security properties,
  • its efficiency in propagating communications in a verifiable way to a global network without the use of a central server, and
  • the manner in which the transaction history constrains a user’s ability to perform actions beyond the scope of their permissions

will result in a significant cost savings in the form of

  • predictable and secure automation of data management,
  • with reduced supervision,
  • without needing a central server or fixed hardware,
  • in a way which is easily repeatable, again without the need for dedicated hardware.

My personal view – as I think Richard’s is – is that to say “decentralisation is valuable but automation is not” is fallacious, and confuses cause with effect: 

  • Cause: reliable, secure process automation on certain terms + transparent audit log and consensus mechanism
  • Effect:  Ease of onboarding new users + ease of distributing the application.

The Bitcoin hardliners’ objections also fail to acknowledge that most people possess an abundance of computing power and hard drive space which lies fallow most of the time and could be put to better use. We may expect this spare capacity to grow substantially in line with technological improvements in the near future.

Our first tech demonstrator, a distributed, serverless YouTube-type thing that I’m not allowed to call “Youtube”, is often derided as being commercially useless. Right now, it is, so usually we brush it off. But seeing as I have your attention, consider what such an application in a more complete form could do for Google in reducing their server costs, or for artists in capturing more advertising revenue by providing this data processing service themselves – as a guild – instead of relying on “the stacks” (Google, Amazon, etc.).

Let that sink in. Then apply that thinking to Uber, AirBnB, Spotify. I could go on.

Yes. This runs on a blockchain. Today. 

3) The third question: is Bitcoin real?

Put differently: if the “utility blockchain” view is right, whither Bitcoin and cryptocurrency?

Finding an answer is not easy as either of Bitcoin’s promoters or detractors would have us believe.

On the one hand, there can be little doubt that it is being used as a value transfer mechanism and thus – at least in the minds of its users – it is a value transfer system.

On the other, there can be little doubt that – in terms of being a valuable asset/commodity/currency/whatever – Bitcoin has been subject to the vagaries of market manipulation on a vast scale, is not as “trustless” as we would like to believe, and possesses characteristics which make it suboptimal for a wide range of commercial applications. 

Where the truth lies is not presently known. What the truth is will have profound consequences for the space, if not for the Internet itself.

That said, I’ve placed my bets. (You can probably guess on which number.) If, as Wences Casares put it, one primary benefit of Bitcoin over a traditional payment system is that: 

In a closed ecosystem, it only processes them when they have enough to make the transaction fees justified. They will still take a few days to receive it, and you still pay 3.5% to Visa. It creates that sense that it’s paying immediately, but it takes three days to clear. It costs a fortune.

…and furthermore, it is also the case that

The entire transaction volume of the Bitcoin network could be handled by a Raspberry Pi in a shoebox. Instead, we have a transaction volume of a few hundred thousand database entries a day being handled by a compute cluster comparable to those used to simulate atomic bomb blasts…

… it stands to reason that the availability of free-to-use blockchain technology without the competitive mining mechanism and with legal control in its place will certainly act as a constraint on Bitcoin’s ambitions – simply because these solutions are cheaper to deploy and can be designed to better address specific needs of specific users of a given network.

Whether Bitcoin will be able to withstand this competition will depend on the strength of its network and its relative value proposition – how “real” it is. For which the data structure is arguably the less important part of the story – since “blockchains are (also) made of people.”

4) The conclusion I reached a year ago, and reiterate now

The existence of cheaper, application-specific alternatives to any product or platform is, in a free market, something which cannot be ignored. I do not, however, think this is such a bad thing: Bitcoin is an application, designed for a specific purpose, and it’s good at doing what it does for people who want it to be what it is.

Trying to make a single “decentralised” network do everything is a fool’s errand. It will never behave quite how you want it to behave, never fully adapt to or address the problem you want it to solve. And there is the everpresent possibility, lurking beneath the surface, that it will turn against you – leaving all your hopes and ambitions in tatters.

Maybe it’s time we let Bitcoin be to do what it was originally designed to do – and learn from its example: by using the innovations it pioneered to design new applications, with new tools, which are actually fit for purpose.

POSTSCRIPT: the above isn’t a radical view. It’s a nuanced one.

4 comments

  1. Johnston’s Law ~ Everything that can be decentralized, will be decentralized.

    Steve’s Law ~ Whatever could be, should be

    Byrne’s Law ~ Whatever should be, should be

    Doris Law ~ Whatever will be, will be…♪ Que Sera, Sera ♪… the future’s not ours to see…. que sera, sera. https://www.youtube.com/watch?v=azxoVRTwlNg

    JP Morgan’s Law ~ A man always has two reasons for doing anything: a good reason and the real reason.

    (Edit: correcting the order of Byrne & Steve.)

    Liked by 1 person

  2. I think you’re focusing on “serverless” consensus- but I think a “decentralized blockchain” is another level up from it. Decentralized ledger is the true innovation, no?

    Liked by 1 person

    1. By “serverless” I mean that the protocol is doing the work of managing all the data without a single point of failure – which is as true as it is in Bitcoin as it would be in a more restricted blockchain-governed application.

      The “decentralised”/”distributed” argument gets tricky as Bitcoin is arguably both and neither at once (depending on whether one is a proponent or critic – whereas Eris can be structured in such a way, if desired, that the answer is more definitive). Although admin keys can intervene in any function of the network if that’s what a developer specifies, as with Bitcoin those admin keys can be broadcast from anywhere where the admin can connect to a peer, not necessarily from some central point of operations. How the system looks and works will be a question of what’s most appropriate for a given data-driven problem and the application which has been built to solve it.

      Semantics aside, “decentralized ledger” really has nothing to do with money – I prefer the term “decentralised rulebook.” A rule-book for a data-driven process. Where Bitcoin solves the double-spending problem, a smart contract ledger uses its “ledger” to solve it in respect of whatever user permissions a dev can describe in code. Just as the ability to spend a Bitcoin is a write permission, so the ability to change the state of a particular smart contract-driven application (or even a particular smart contract) is also a write permission – one for which a blockchain is ideal at controlling.

      In my view, the key commercial takeaway – to paraphrase Satoshi Nakamoto – is this: the blockchain is a secure “network (with) minimal structure” which automates process. Given the fact that very little processing power is required to validate txs even on Bitcoin, my hunch is that blockchains will serve – going forward – as cheap, secure, and repeatable data infrastructure.

      Like

  3. […] Check out a couple great blog posts from today: “Bitcoin as a Smart Contract Platform” and a related post, “The Great Divide.” […]

    Like

%d bloggers like this: