On our inability to define blockchain

August 21st, 2017 by Peter Evans-Greenwood & Ross Hancock | Centre for the Edge

Mid-way through most presentations on blockchain the speaker puts up a slide with a technology adoption curve, an s-curve. Then, pointing intensely somewhere early in the exponential bit, they declare “we’re here!”, before racing on to outline a future decentralised utopia.

We forget, though, that Bitcoin involved no new technology. It was an interesting agglomeration of OpenSSL and email anti-spam technologies. The key innovations were the use of triple entry accounting to prevent double spending (with transactions recorded on a shared ledger) and issuance as a side-effect of the consensus process (so that it's stateless, not backed by anyone or thing). It seems that Bitcoin did not emerge due to the invention of new technology, new maths. Rather, it was an environmental change – a shift in the price, performance and penetration of broadband – that was the trigger. This tipped the scale from centralisation towards decentralisation.

This has two implications.
First, we can’t expect blockchain to follow the usual technology adoption curve. Society’s reaction to an environmental change rarely follows an s-curve, so why should blockchain? It’s more likely to be a degenerate process; complex, redundant and evolutionary.
Almost a decade has passed since Satoshi’s whitepaper and despite a huge number of proof of concepts, there is no mainstream application of blockchain other than Bitcoin. No killer app has emerged. A huge volume of money is flowing into the sector, but we’re seeing very little for it.

It’s likely that this is because firms are working through a range of different organisational and consortia models to determine how to take advantage of the environmental change that blockchain signalled. What we’re seeing is communities getting their heads around building governance structures and operating models on more distributed topologies. Once these new models are established expect a rapid (though not necessarily disruptive) shift as industries and sectors reorganise.

Second, the environmental change suggests that other approaches to creating distributed systems are now also viable. Blockchain is just one way to realise many of the use cases proposed, from product provenance to financial settlement. Often a blockchain is not the most efficient or effective distributed approach possible. This might be why R3's CORDA doesn’t have a blockchain in its default configuration. The most common use cases that R3 investigated didn’t require the guarantee of global consistency that blockchain provides, and they realised that they could go with a cheaper and more efficient attestation model.

This presents us with both a problem and an opportunity.
The problem is that blockchain is a hammer and any (potentially) distributed problem is looking like a nail. Many (perhaps even most) of the use cases blockchain is being applied to may prove to be uneconomic. Sure, we can use blockchain to solve these problems, but other (decentralised and centralised) approaches will be more efficient and effective, and will win in the long run. It would be wise to pause and reflect a moment on what problem we’re trying to solve before rushing into proof of concept or implementation.

The opportunity is to rummage around in the back of the cupboard and review old distributed technologies. To reassess the good ideas who’s cost-benefit didn’t stack up at the time, as the bandwidth wasn’t sufficiently cheap enough or widely available. Contract Net springs to mind as a good example, though there are many others. There are opportunities there for the taking, we just need to think more broadly than blockchain.

blog comments powered by Disqus