What we know we don’t know about blockchains

Originally published on the Ness Digital Engineering blog at https://www.ness.com/know-dont-know-blockchains/.

Bitcoin is already a household name. The blockchain, the data structure plus protocol that lies under Bitcoin and many other novel applications, gets mentioned in almost every list of hot technologies [1].

For every writing that extols the potential of blockchains, another pops up dismissing blockchain-based solutions as inferior [2]. Both extremes are of course “right,” such as you would expect from any projection of a complex, multidimensional issue.

Yes, blockchains have potential to impact our world. Like any technological innovation, blockchains have caused our collective imagination to run loose and create narratives of a future without banks, without government [3], etc. But, predicting the future is hard. In fact, if you’re in the business of telling the future, you can rightfully be called a visionary if you manage to get just a few predictions right and fail at all the others: read Jules Verne’s Extraordinary Voyages with a critical eye, or for a more recent example, Bill Gates’ The Road Ahead. Many other narratives about the future, that completely failed to happen, are now forgotten thanks to survivorship bias.

Now apply the same critical thinking to any narrative about the impact of blockchains to any particular industry. Will cryptocurrencies change finances and banking by a little or by a lot? There is no clear answer yet. One major unknown is what will credit look like in a world of cryptocurrencies with an inelastic, capped money supply. Bitcoin’s creator was wary of credit institutions and central banks and wasn’t necessarily focused on working within a credit-based financial system [4]. Yet, for a modern economy, credit is essential. There is good evidence of a link between the rise of credit institutions and the explosion of economic activity during the European Renaissance [5]. We don’t yet have a crypto monetary system that can replace our fiat currency system without major economic regress.

To tackle the other extreme: yes, blockchain products are the baby replicas of their existing 20th century equivalents. Bitcoin is slow and expensive compared to Visa. There is no court of law to judge the breach of an Ethereum smart contract, even when most agree there should be a difference between the spirit and the letter of the contract, such as in the case of the infamous Ethereum contract bug that brought down The DAO [6].

Remember that blockchains themselves are a solution for a very narrow problem, namely achieving consensus in a network of entities who don’t trust each other. The novelty is that blockchains don’t rely on a central authority. A protocol for decentralized consensus is a building block for distributed applications, but it doesn’t have to be the only one. Higher level protocols are already emerging on top of blockchains to deal with specific applications. For example, Bitcoin developers have created the payment channels protocol [7] and the Lightning network [8]. If adopted, the Bitcoin network will see payments happen “off-chain” quickly and cheaply, and only gross settlement transactions will be committed to the blockchain, an expensive and slow process.

A useful analogy is the history of the Internet itself. The necessary building block for the Internet is a packet-switched network protocol, such as IP, that can effectively route information between nodes on a huge network without the need for a central authority that decides which packets go where. IP alone, however, does not solve all our problems. In IP, the size of a payload of information is limited by design to a few KB [9]; there is no notion of a bi-directional stream of packets between two hosts; there is also no notion of a session. There are higher-level protocols on top of IP that needed to be invented, so that we can enjoy today’s sophisticated Internet products such as Netflix and Skype.

It’s safe to say blockchains are only layer one of a stack of protocols for distributed applications, and there are many more protocols to come which have not yet been invented. My prediction is that it’s too early to pick winners, just like in the ’80s, it would have been premature to pick SNA as the winner over TCP/IP in the battle for communication network protocols.

In conclusion, when you’re building your product or your company under the assumption that “blockchains are the future,” make sure you’re only taking the risks you can afford to take. If we go by past experience, the future will certainly be surprising.


[1] https://www.ness.com/technology-trends-2018-nesss-cto-associates-weigh/
[2] https://hackernoon.com/ten-years-in-nobody-has-come-up-with-a-use-case-for-blockchain-ee98c180100
[3] https://www.ribbonfarm.com/2017/10/10/the-blockchain-man
[4] http://p2pfoundation.ning.com/forum/topics/bitcoin-open-source
[5] https://www.ft.com/content/6851f286-288d-11de-8dbf-00144feabdc0
[6] https://www.coindesk.com/ethereum-classic-explained-blockchain/
[7] https://en.bitcoin.it/wiki/Payment_channels
[8] https://en.bitcoin.it/wiki/Lightning_Network
[9] https://en.wikipedia.org/wiki/Maximum_transmission_unit

Author: Iulian Dogariu, Ness Digital Engineering