The real value of Bitcoin is in the P2P stack

This is one of a series called Explorations down the BItcoin rabbit hole.

Will developers use Ripple, Ethereum, Maidsafe or Open Transactions or some combination to build the killer apps of the Bitcoin era?

As I started down the road to understanding Bitcoin, one of the most confusing things was distinguishing between a payment network (“Bitcoin” upper case) and a currency unit (“bitcoin” lower case). That is now the first thing I explain to people who ask me about bitcoin.

It is lower case bitcoin – the currency – that has the media attention. It is simple enough to understand; it could be like gold or tulips or a reserve currency that replaces the US$ for international trade. It could be a currency or a commodity. Whatever it turns out to be (probably none of the above because the future always surprises) everybody understands the concept of a currency or a commodity.

It is possible that lower case bitcoin – the currency – may not change the world. For the bear case, read Felix Salmon’s post and for the bull case read Ben Horowitz’s rebuttal annotated to his post. They set up a bet:

Five years from now, in January, 2019, we’ll poll a representative sample of Americans. If 10 percent or more say they have used Bitcoin to buy something in the past month, Ben wins. If it’s fewer than 10 percent, Felix wins.

Even if bitcoin the currency becomes a footnote in history, it is possible that Bitcoin the payment network based on Blockchain technology may change the world. This is what has VCs excited. A new payment network could not only disrupt the global financial services business (for good or bad depending on your point of view). It could also return the Internet to its roots as a decentralized P2P system (aka “re-decentralization”). A decade from now, centralized cloud servers may be seen as a footnote in the history of the Internet.

For this to happen, we will need to see platforms that make it really simple for developers to create applications that reside on the unused cycles of all of our machines in order to deliver value to us in the way that Skype or Bit Torrent does.

Update from 28 July 2020, on 6 year anniversary. It is clear that Felix Salmon won that bet.

The emerging P2P Blockchain technology stack

This got me exploring technologies that are sometimes tagged Crypto 2.0 or Bitcoin 2.0, such as Ripple, Ethereum, Maidsafe and Open Transactions. I prefer to think of them as an emerging Bitcoin stack (capital B, used for more than bitcoin the currency). If Bitcoin is as revolutionary as many believe, this stack will be at least as important as the Wintel stack that ushered in the modern digital age.

Before diving into these platforms and the sometimes-heated debates between the adherents, it is worth reading the original Satoshi White Paper. He envisaged:

“a solution to the double-spending problem using a peer-to-peer network”

“Double-spending” is the problem created by the fact that anything digital can be copied (for almost no cost). Many ventures have used this perfect copy machine capability; it’s great for communications and media. However, for anything involving financial assets, this is a problem; if I own this asset, you do not own it. If you can simply copy the record that says that I own it, then I regard that as stealing. As anybody involved in cybersecurity will tell you, anything digital that is connected to the Internet can be copied i.e. stolen.

Satoshi’s solution was to have a cryptographically verified record of each transaction stored on every computer in a P2P network, which he called the “Blockchain”. The Blockchain is fully distributed; it sits on every machine in the network. That is how the double-spending problem is solved and trust is enabled. You can “see” all the transactions. The “mining” concept is simply a way to financially motivate people to use their compute cycles to verify transactions.

Many consumers have a strange image of peer-to-peer networks. They either see something illegal and piratical like Napster or they use something every day like Skype without realizing that it is peer-to-peer. That is probably the way that the Internet will return to its peer-to-peer roots; consumers will trust the Cloud and the Cloud will move from centralized servers to peer-to-peer networks. What is a seamless transition for consumers – the same product but just cheaper – will be a revolutionary change in the IT industry. Bitcoin and the Blockchain will play a key role in this as people start to grasp the strange notion that it is trustworthy precisely because it is peer to peer. This goes against all our 20th century faith in centralized institutions.

Who will be the Red Hat of the Blockchain era?

Any platform will have to be open source. Thus the question is who will be the Red Hat of the Blockchain era?

Nobody wants “one Blockchain to rule them all”. Nobody wants to see this critical layer of the new financial services stack dominated by one company. Yet the logic of peer to peer will tend towards network effects and a winner takes all market (just like it did in the Wintel, Google and Facebook eras). Consumers will have to trust something enough to accept a download of code that will have control over their machine; this is a scary proposition and a level of trust that people won’t give to many companies.

So the prize at this platform layer is huge.

Building Internet scale decentralized P2P systems is technically really, really hard. Ask the guys who built Skype. Building a payment system is far harder than a VOIP system because the risk of loss is so much higher. Some noise on the line that forces you to ask your buddy to repeat something is OK and a small price to pay for getting something free; losing some money through a technical glitch is not OK. It is a hard technical problem because you have to deploy to millions of machines of varying power and type that are only intermittently connected to the Internet and deliver a service that is as fast and reliable as the centralized server based competition. This is not something that your average Ruby on Rails or Javascript developer can do. Yet, for the Internet to return to its roots as a decentralized P2P system, we will need the platforms and tools that make it as easy to build and deploy to a decentralized P2P network as it is to build and deploy to AWS or the machine in your closet.

To understand this emerging stack, I started by interviewed people from two decentralized development platforms that use Blockchain concepts: Ethereum and Maidsafe. Warning, bleeding edge alert, these platforms are not yet ready for live applications; despite this they have many developers spending time on them because the prize, if they can deliver on the promise, is very big.

First, I wanted to know if Ethereum and Maidsafe are competitors. It’s an obvious question that is being asked in Google searches and Bitcoin related forums, because they are both positioned as Blockchain related tools. Ethereum pitches itself as a full stack platform with a “logic layer” and a “storage layer”. However, as the Ethereum storage layer is based on Bit Torrent, Ethereum see their core competency in the logic layer and so they make friendly noises about Maidsafe which positions itself at the storage layer. Maidsafe concurred on this point about being complementary, not competitive. This sounds like a classic “stack” emerging and that’s OK, as long as layers in a stack don’t create programming complexity or latency.

This does not mean that all is happy talk in this Blockchain platform space. Asked to comment about Ripple (funded by big VC money), Stephan Tual of Ethereum put it down as “14 lines of code in Ethereum”.

I did not have time to interview Ripple before publication and I am sure that they would dispute Ethereum’s view. My perspective from online research is that Ripple is positioned higher up the stack than Ethereum. You might not use Ripple to build a whole new system but you might use it to power a consumer-facing mobile app. It seems that Ripple positions itself more to serve existing banks than it does to serve new ventures. To be fair in any discussion of Ripple vs Ethereum one should point out that Ripple has already got a product in the market, while Ethereum is hoping to release at the end of 2014.

The “sharing economy” is built on trust and that should be distributed

The most obvious competitor to Maidsafe technically is Cleversafe. Both “shred” your data and put it on computers controlled by users in the network; the data is reassembled when you need it. This enables lower cost services, because the provider does not need to pay for servers; in fact Maidsafe promises to pay users for the privilege of using their computer’s storage. Imagine a Dropbox type service that paid you!

Seen in this way, Maidsafe and Ethereum are really a part of the sharing economy, like AirBnB or Uber. You share your compute cycles like you share your spare bedroom or your car. When pressed to come up with the sort of applications that people are building on top of Ethereum, Stephan Tual talked about these kind of sharing economy services. This makes sense because the Ethereum logic layer is all about Smart Contracts (something that knows about a financial asset as well as changes related to that financial asset, such as a change of ownership) and the sharing economy is all about lots of contracts that need to be managed efficiently.

The lower cost of storing data in P2P networks has not been compelling, because Moore’s Law has ensured that costs are falling anyway; there is not the kind of massive cost arbitrage that enabled Skype to thrive. The other driver is privacy/security, because there is no centralized server to attack and sniffing the network will only get the shredded pieces of data. However, while many consumers feel like they should be concerned with privacy/security, few are willing to pay in money or time to ensure greater privacy/security. The compelling use-case for P2P decentralized storage has yet to emerge; both Maidsafe and Cleversafe have been working on this for nearly 10 years.

However, Blockchain related applications must have strong privacy/security because they relate to financial assets. A pragmatic mainstream consumer reaction goes along these lines:

“I am really not too worried if somebody wants to snoop on my chatter with friends and family or to understand my shopping habits; but I am concerned if they can steal my financial assets”.

Blockchain applications are all about financial assets, something that you own that has value. Banks currently store these assets and spend a lot of money on privacy/security. The promise of the Blockchain is that anybody can build systems that store financial assets; for that to come to reality they must have strong privacy/security controls.

There are many financial processes we go through as consumers that are klunky, paper-driven, time-consuming and expensive. Which of these could be replaced by systems using Smart Contracts? Will we see more efficient (aka lower cost) versions of existing services such as AirBnB or lending applications? Or will totally new services emerge that are only possible because of the Blockchain? The history of innovation indicates the latter scenario, but as always only time will tell. In short, the killer app for the Blockchain is out there but if anybody knows what it is, they are not telling.

There is so much at stake that developers looking to commit to one of these platforms spend a lot of time figuring out whether the platform is truly open

This led me to look at Open Transactions. I did not have time to interview them. I suspect that I would need more technical chops to understand it. I come at this as a Fintech entrepreneur and adviser who is comfortable working at the intersection of bleeding edge and leading edge, but I am not a hands-on developer. It is possible that a technical developer would opt for Open Transactions rather than Ethereum and it is possible that they serve different needs. There are usually trade off decisions between time to market and flexibility and from my perspective Ethereum looks like they have the trade off about right.

But, it is still really early days in this game and we are all learning every day.

This is one of a series called Explorations down the BItcoin rabbit hole.


  1. […] There are many companies competing with Ethereum in the re-decentralization space. The same was true for PCs in 1975. The trend – towards re-decentralization or to a PC on every desk – is more important historically than which specific technology wins out. I know I will annoy all those who prefer other re-decentralization technologies. I am not able to say which one is better or which one will win out in the end, although my gut tells me Ethereum has a good shot and I have been fascinated by Ethereum since I wrote this post in July. […]

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.