There are a lot of people who don’t think “blockchain without bitcoin” is possible.
Well, “Blockchain without Bitcoin” is very possible.
It’s a thing. Here, now, today.
So today I saw this video of Jeremy Allaire saying:
There’s a meme that’s going around… (that) the blockchain’s interesting but the “currency” or the store of value’s not very interesting. That’s just a fallacy. They’re not possible separately. There has to be an underlying value to the token that’s used to move value, and there has to be an incentive system for the creation of that token and the exchange of that token. That is the only way these systems work… that’s a thing I think a lot of people fail to understand.
Sorry to be the bearer of bad news for many of my readers (whom I like very much despite philosophical differences), but “Blockchain without Bitcoin” is not a meme. It’s working software.
Blockchain, together with smart contracts, has been let out of Pandora’s Box – never to return, never to be unlearned, no matter how badly some of you might want it to be.
There are perfectly good reasons for wanting to use a blockchain without Bitcoin. I have written this because every time I allude to these ideas on Twitter, the lack of space means presenting a nuanced view is impossible. Hopefully after reading this, the differences in approach between
- the “distributed application/ledger” blockchain set (database tool) on the one hand, and the
- “cryptocurrency” blockchain set on the other,
will be clear.
1) Further background.
Tim Swanson’s latest report, “Consensus-as-a-service: a brief report on the emergence of permissioned, distributed ledger systems,” (full report PDF) is now out.
The report features my company, Eris Industries, which was the first, and to date remains the only, open-source platform released to market in this field.
The report confirms what we at Eris Industries have been saying for months: that blockchains without cryptocurrency tokens are not merely possible, but also that they are incredibly useful as data verification tools.
Those of us who are in the “blockchain” set, while very fond of Bitcoin and cryptocurrency, took a hard commercial look at these “1.0” technologies. We determined that it wasn’t suited well for mainstream usage, both private-sector and public-sector, or as the entirety of a distributed global economic commons. As Bitcoiners often claim it can be.
So we invented our own stuff that we thought was a better fit.
Enter the permissioned blockchain.
So what’s the difference?
In cryptocurrency, people continually ask certain global blockchain designs to do things they were simply never designed for, an observation which could be made of practically any cryptocurrency which claims universal applicability for all of the world’s transactional data (as many of them do).
Cryptocurrency aims at creating systems that are global replacements for the financial system. Stateless, bankless, ownerless digital cash.
The permissioned blockchain/distributed cryptoledger set, on the other hand, proceeds from a different set of assumptions. We believe these data structures are designed be
- limited in scope and
- tailored to address highly specific needs for data verification and process execution.
“Blockchains” therefore are just another kind of distributed database. One which keeps itself in synch with very low supervision and without a central server to run it. Data infrastructure without physical infrastructure.
The world they live in – unlike Bitcoin – has legal rules and obligations, banks, states, and regulators. But they’re still really useful if you want to bootstrap a network at almost no expense, which gives you all the verifiability you would normally need a data centre to obtain.
“Blockchain” people are thus trying to solve a very different set of problems than the “Bitcoin” people.
2) Report highlights.
Tim’s conclusions, although stark and (presently) “controversial,” in my view are pretty obvious. Basically, he says that
- “Blockchains without Bitcoin” are possible.
- “Blockchains without Bitcoin” are commercially useful.
- Bitcoin does not do everything its most breathless proponents say it is capable of doing.
The following points from his report are instructive:
- “Decentralised” networks such as Bitcoin do not allow commercially necessary/legal-technical enforcement or other commercially necessary intervention.
2. Distributed ledgers are useful because they are automata that provide industrial-quality data verification while nonetheless running on nearly nil hardware.
Smart contract enabled protocols for value transfers between ledgers are one candidate for the Internet Protocol (IP) equivalent of money. The Internet Protocol serves to mediate the exchange of data between two geographically separate Local Area Networks. Similarly, operations described above serve to mediate the transfer or exchange of value between two distributed ledgers.
Once value transfer protocols are in place, more complex inter-nation value transfers, such as payments originating in Citibank USD and terminating in UBS Swiss Francs, can be realized by a chain of operations mentioned above. The most efficient chains or paths can be computed using automated services.
This idea leaves intact all innovative characteristics of cryptocurrencies. Any application pioneered by the cryptocurrency community can be implemented for the banking system by programs utilizing the value interchange mechanism.
3. “Mining” economics does not scale. Indeed the whole concept of “mining” is sub-commercial.
The entire threat model of Bitcoin was purposefully designed to make it purposefully expensive to attack and change votes – securing Bitcoin was “inefficient” on purpose… rational hashers will only destroy as much capital as an actual bitcoin is worth. So if a bitcoin is worth $300 they will only spend up to that point, otherwise it would be cheaper to just buy coins from the market and turn off the machines.
If a bitcoin reached $2,000 in value, the same behavior would take place: miners would destroy as much capital to reap the seigniorage (the spread between the marginal value and marginal cost) all the way up until they are expending the equivalent of $2,000 in exergy. And so on… For instance, if Bitcoin became a $100 billion network in the future, with a 14 million coin money supply this would equate to about $7,140 per coin.
By August 2016… this (would amount) to $89,000 per block or $6.42 million per day.
4. Despite the fact that distributed ledgers are more likely to be commercially useful, VC investment in the space has been allocatively extremely inefficient.
VCs have to date focused overwhelmingly on the “moonshot” Bitcoin play (least likelihood of success) while avoiding the more pragmatic distributed database play – which is more boring, but at the end of the day has the greatest likelihood of success and adoption.
5. “Permissionless” systems like Bitcoin do not fit the way that mainstream finance works. They are thus unlikely to be of direct usefulness in any commercially viable, mainstream application.
…the diminutive usefulness of permissionless systems for participants in the permissioned traditional financial system, on the part of Bitcoin, was not some kind of unanticipated shortcoming or design flaw, but a result of intentional choices by these systems’ designers who were quite clearly reacting to aspects of permissioned systems that they disliked…
Permissioned finance is different than permissionless, and each organization should look at which network best supports its requirements. Perhaps, as some Bitcoin enthusiasts suggest, this is all akin to Highlander or Lord of the Rings: there can only be “one chain” to rule them all and that chain is Bitcoin. While it cannot be known a priori, this narrative may not prove true for cryptocurrency systems and is most unlikely for distributed ledger networks as well.
3) A billion blockchains are more disruptive than one
Blockchains allow people to run their own network infrastructure without physical infrastructure. Permissioned/private blockchains allow them to do so at almost nil cost.
There’s nothing about a distributed ledger which changes anything about what Bitcoin does on a day-to-day basis. But this isn’t what seems to raise most people’s hackles. Most of the space:
- first, never thought the fully controllable blockchain/distributed ledger could be done – and indeed, as of yesterday there are still those who think that blockchain and cryptotokens remain inseparable; and
- second, reject out of hand any notion that such blockchains/smart contract machines would be commercially useful at all (a view I characterise thus: “blockchain is the best data structure ever, unless everyone can have their own, in which case it sucks”).
With our stack, blockchain applications which defy these assumptions can be deployed in a matter of seconds with a single command. We may conclude:
As to the first assumption (“it’s not possible to use a blockchain without a cryptocurrency”):
- this is now demonstrably false. Permissioned blockchains/ledgers can be administered, not competitively mined (although the attack vectors will change with particular design parameters – operators must decide the tradeoff, which will differ from application to application. We can use public/private key crypto to secure the chain.) or transmit value (as this value relates to the legal character of the relationship between the users and the database administrator, as with a regular database).
- A particular blockchain/ledger is a database tool in this view, not a standalone economy unto itself with tokens bearing a value in market exchange. Its administration is incentivised by its usefulness in reliably and verifiably executing particular, specialised functions, much in the same way as with other kinds of databases today.
- Because competitive mining (a process we call “committing”) is out of the equation, it can be set to be pretty inexpensive (e.g. with Bitcoin, once the obscene mining difficulty is removed one could capably process all of the transactions on the network on a Raspberry Pi).
As to the second assumption (“nobody will want to use a blockchain”):
- we, and others in commerce, government and governance, believe this to be false as well. Furthermore, scrapping the uncontrollable “decentralisation” piece isn’t just a convenience, but a necessity for such applications.
- Blockchains’ traditional inefficiency is a problem Bitcoin has, but not one a distributed ledger should. Yes, a central server might be faster and a blockchain will suck for an application like HFT (where the speed of light is your enemy), but (a) it’s been six years since 2009, so the tech has improved; (b) the way we use the blockchain keeps the processing burden light; and (c) blockchains are designed to be used in circumstances where you don’t want servers or where the need for verifiability/security in a public-facing (or interparty) data layer exceeds the need for speed.
- Even a relatively lethargic 500-ms tx verification with a 30-second blocktime would be 99.999% faster than existing processes for, e.g., FX clearing and settlement, keeping in mind that today this process is T+2-5.
- Thus the cost of transmitting information is roughly the same as transmitting other data, like e-mail. Other databases to hold, e.g., streaming video (we built a YouTube-type app as a tech demonstrator) such as a DHT are referenced by pointers in the blockchain rather than being stored on it in full.
As to definitions:
- there has been a bit of discussion over what a blockchain is or is not, with a number of more – shall we say, dedicated – crypto folks asserting that if it isn’t “fully decentralised” (a misnomer) and doesn’t have tokens (a type of data a blockchain can record), it isn’t a blockchain. I don’t think this holds water – a blockchain is a data structure, not a world-view. Equally, though, I don’t really care what it’s called. It’s just not that big a deal.
All told, the taming of a child of Bitcoin into a more commercially-palatable, and practical, permissioned “blockchain” design appears to have pissed a lot of people off. I can’t understand why: it’s a new set of tools to address a new set of problems. It’s a totally different way at looking at blockchain databases.
This being crypto, the tools are going to be open-source. Use them if you would like to. Nobody’s saying you have to.
I would add the important note that people *can* create decentralised applications – if they make sense – on our stack. The software permits the whole range – from fully-public, fully-open cryptocurrency driven on the one hand, to totally locked-down, one-user one-instantiation utility-driven on the other, and every gradation in-between (including using what we call the BBPCs – the Big Bad Public Chains – to checkpoint or otherwise provide security). How devs structure these applications is, as it should be, entirely up to them. And their chains are entirely theirs to control.
We merely provide the tools to help them achieve their objectives. If we’re right, the cryptocurrency technology which was once meant to revolutionise global finance on a fully peer-to-peer basis entirely on its own may no longer be the best tool to do so.
But then we have to ask ourselves if it ever was to begin with (not trying to be difficult – just calling it like I see it). I feel no need to explain my reasons for believing this here, as Tim’s report (both in terms of identifying relevant facts, and in not predicting success or failure of cryptocurrency but leaving it as a question mark) is a fairly adequate summation of my own views on the topic.
4) Whither Bitcoin?
Whither Bitcoin (and cryptocurrency more generally), then, as Tim asked?
For the time being, the simple act of even asking that question – at least in a business context – means betting against some $800 million of venture capital. I’ve never liked following the crowd, nor have any of my colleagues at Eris Industries. Alleged paradigm shifts are no exception.
Personally? I believe it will survive, and keep on keeping on – in the context of addressing a very specific set of needs for a very specific audience. Like any software application. This is a wholly reasonable position to take. People should have the ability to use whatever data structures they want, to interact with any people they want, for any purpose. Bitcoin is a form of freedom of association. That alone is valuable.
Nor is there a choice to be made; these are not mutually incompatible propositions. Utility-blockchain and cryptocurrency-blockchains, though they share common origins, are simply not comparable. They do different things and address different needs. Whether that means “cryptocurrency”-model protocols will become a dominant force in global finance, merely notable players, or something less, no-one can say. And it would be irresponsible to predict.
But I will say one thing: the game has changed.