Is there a Bitcoin in Bitcoin 2.0?

What does cryptocurrency look like in 5 years? Is it one dominant network (Bitcoin) interfacing with many applications built on top of it or is it multiple networks based on differing blockchain inspired design principles?

Of these next generation networks (collectively referred to as Bitcoin 2.0), Ethereum is the most interesting to me as it promises to be the most flexible. It is a platform for building any decentralized application that can be thought up and promises to enable developers build things as complicated as name servers, and currencies in under 10 lines of code. I wrote a high level primer on Ethereum in this post.

In the rest of this post, I’d frame the initial questions in light of Bitcoin and Ethereum and try to understand what scenarios are possible and which are likely.

Why isn’t Ethereum built on top of Bitcoin?

While I am a big fan of the Ethereum vision (I own some ethers from the sale), something about it not being built on top of Bitcoin bothers me both from a historical and practical perspective and I’m not yet sure in what direction I net out. The rest of this post is about exploring this question of whether next generation protocols should or will eventually just be absorbed into Bitcoin.

The reasons given by Ethereum for making an independent blockchain are sound. As I understand it, they are:

1. Bitcoin’s built-in scripting language is not Turing complete. Most noticeably it doesn’t have built in loops. The protocol also doesn’t keep track of states beyond “sent” and “received” thus limiting the kinds of apps we can easily build.
2. Verification of transactions that go beyond simple payments is not scalable with the current Bitcoin blockchain data structure

The first argument does not seem that strong to me. Loops were probably initially avoided due to potential computational disasters (caused by infinite loops for example) and I don’t see a fundamental reason as to why loops and other desirable properties can not be added to the Bitcoin scripting language. The same goes for adding state to the protocol. I’d love to hear what people working on the bitcoin core think.

The argument about scalability seems stronger to me. Since the blockchain will continue to grow in size, and everyone needs to look through a copy of this > 20 GB blockchain to verify transactions, we have to devise ways around this. We need ways in which some people can opt to be “lite” users who use bits of the chain that are relevant to them and together with users who have the full chain can still verify transactions. One good way is the “Simple Payment Verification” scheme. Ethereum’s main gripe is that this scheme doesn’t generalize to arbitrary contracts (beyond just simple payments) on the blockchain.

Fundamentally, both problems come down to what the Bitcoin Foundation chooses to prioritize in the near and medium term. I do not see a reason, in principle, why the Bitcoin Foundation could not modify its architecture to enable verification of arbitrary contracts. It would presumably be in line with its goals of ensuring that Bitcoin is scalable and can enable Visa level transaction rates. I understand that this sort of architectural change could require alot of effort and care in practice, but it’s obviously doable.

Besides practical concerns, there’s also the historical concern that networks tend to only grow bigger and engulf more territory in their wake making it really difficult for new networks to bootstrap growth.

On the flip side one could also argue that network effects imply that in order to succeed, an emergent network simply has to fill a sufficiently different need from that filled by the dominant network — this is the reason why in addition to Facebook, we have Twitter and LinkedIn. New networks could also emerge due to a fundamental platform shift — this is the story of Instagram, Whatsapp and Snapchat.

In this light, extrapolating from the present, one could make the following prediction:

Prediction X

In 5 years everyone views Bitcoin as “the network for sending and receiving payments and for managing digital money” while Ethereum is “the network for using decentralized applications like file sharing, file storage, decentralized organizations, machine to machine payments, smart property etc”.

In the above scenario which network is Facebook and which is Twitter? Does it actually play out this way or do Ethereum’s best ideas just get absorbed into Bitcoin which has the advantage of a very devoted following and, very much like Apple, of feeling like a true movement and not just a technology.

I ponder this not because I think Ethereum views itself as a competitor to Bitcoin but because it is intellectually interesting and as an investor (of my time) it’s somewhat important to make the right calls as to where to contribute. If the Bitcoin foundation sees itself as primarily focused on improving the core wallet and payments protocol and it turns out that Ethereum (or a similar protocol) is able to deliver on a technology that makes writing a distributed DNS a job doable in ~2 lines of code then it is likely that Prediction X holds.

Another (unlikely but possible) scenario is that a generalized protocol is redundant and only a few specialized applications are needed. This isn’t too strange of a possibility. While the internet is lauded for its variety, mobile and web applications are mostly dominated by news sites, social networks, games and simple applications that are all enabled by the ability to create, remove, update and delete records from a remote database.

In addition, the ability to innovate is limited by what either the proprietary (e.g. Apple) or non-proprietary (e.g. W3C) gatekeepers decide are the next generation of features to make available to developers. So it isn’t too crazy that we see a scenario where Bitcoin dominates as the core protocol and proceeds by adding new capabilities based on actual user demand or progress in R&D that makes those capabilities possible. This is the way the Web and Browsers currently work.

We could also see a scenario where Bitcoin dominates along with payments and a few specialized non-payments applications are popular.

Finally, there’s the PayPal-Ebay scenario where a network is built by piggy backing on a growing network. The same is true of Youtube-Myspace and Instagram-Twitter. In this scenario Ethereum (or something like it)[1] survives as an independent and thriving network which is intimately compatible with the Bitcoin network. Much like email, the Ethereum network might seamlessly interface with the Bitcoin network but still stay distinct and act as a necessary complement.

LAUNCH PHILOSOPHIES

Let’s leave aside the 10,000 ft question of what the constitution of crypto-currency looks like in 5 years and how that impacts current and proposed networks. I have a more philosophical concern about Ethereum as well as some of the newer crypto currency projects. This is nothing personal and I’d love for the contributors to these projects to take this concern constructively.

The recent crop of independent cryptocurrency 2.0 projects seem too large in scope with ambitions that vastly outstrip their “proof of work” i.e. the amount of actual work that has been done to realize their ambitions. I love everyone of these projects and I think some of these guys are absolutely amazing (see Vitalik Buterin’s writings on everything from Patricia trees to Schelling Coins if you doubt me).

However, i think a lot can be learned from how Bitcoin was launched. More importantly, because Bitcoin is now 5 years old, we can’t simply just write whitepapers and make fancy websites anymore. We have to go much further into the “deployment phase” of cryptocurrency.

In other words, someone who says that they will make an Uber for Food is not as interesting as someone who merely said they would make Uber in 2009 and explained how to make such an app. After Uber has been built, the only thing interesting about an Uber for Food is an actual working product that allows you get your food delivered to you in minutes from your mobile device and not a demo video of it. Satoshi started with a fundamental computer science and social engineering breakthrough (posted to a mailing list!), followed up with some quick and dirty proof of concept code (that worked!) and then passed on the reins of ownership and management to able and qualified people and as his anonymity proves — emphatically let his work speak for itself. We don’t all have to be as awesome or fame-averse as Satoshi but I think we need to scope these projects down. I am not sold on why we need to solve all the problems at once and why we need to raise millions of dollars to write important software.

Importantly, we need to do the hard work that for many projects is actually about interfacing with the real world and its labyrinth of large centralized organizations that make regulation an existential threat and/or control large amounts of online and offline commerce and will do so for the foreseeable future. I applaud the recent work by Coinbase and Bitpay in getting high profile merchants accepting bitcoin.

In the near term, the next set of breakthroughs in cryptocurrency will be less about the “crypto” and more about doing whatever is needed to accelerate the deployment of cryptocurrency to the real world — whether that means shipping working multi signature wallets to enhance security (technical work) or striking all the deals with slow corporations once so that everyone in the ecosystem doesn’t have to do it over (hard real world infrastructure work) or building working decentralized applications that actually deliver on their promise to dramatically cut down costs for the same level of performance.

To come full circle, I am excited about Ethereum and the more recent developments in the world of cryptocurrencies. I’m extremely curious to see how Prediction X fares and interested in learning about how the Bitcoin Foundation thinks about its role in Bitcoin 2.0.

Regardless of how things shake out, I think as a whole we should move towards important but achievable projects[2] that advance the state of the art and build considerable leverage into the ecosystem and shy away from a deluge of white-papers and sign up forms ☺

Happy Hacking!

I plan to write more frequently. If you enjoyed this, you can follow me on Twitter @levandreessen and send in thoughts about this write up ☺

References

1. This is closer to the Mastercoin approach.
2. Richard Feynman says important problems are the ones that we reasonably believe we can solve: http://genius.cat-v.org/richard-feynman/writtings/letters/problems