Blockchain: The ability to scale trust

27^th March 2015

Anthony Macey asks: What wider applications might digital currency protocols have?

Most people I speak to regarding the Bitcoin protocol have never heard of it. Of those that have, the majority have heard of only bitcoin the ‘currency’.

I put that in pretty quotation marks and italics because I am firmly in the “moonshot it will not go to zero” camp. Most of the reasons I believe this are covered quite nicely in Sidney Zhang’s blog post – 3 Reasons Why Bitcoin Won’t be the New Internet, but this does not mean the technology does not have wider applications.

Part of Bitcoin’s fallacy is the aim to be completely decentralised, when in reality it has become increasingly centralised:

• Although arguing that bitcoin is a better way to distribute wealth, it has been estimated somewhere between 78%-96% of all bitcoin are held by 1% of users
• Mining has become increasingly centralised as a result of the profitability – individual mining is no longer cost effective
• The core bitcoin development is maintained by very few individuals. – the present bitcoin protocol, although open source and open adoption, is at the control of a few individuals.

Is centralisation necessarily a bad thing? I would argue no (even Vitalik Buterin of cryptocurrency fame agrees) and it actually makes a great deal of sense to centralise certain aspects of the implementation of this kind of technology. Not necessarily for commercial reasons (although that is certainly one) but also for the policing of bad actors looking to subvert the original application for nefarious reasons.

As a data person, the technology is nothing more than a decentralised database protocol. What makes this useful and disruptive is that on top of this you can authenticate the data.

This means that those entering the data can be held to account by the others also entering data as, if something does not add up, you can see;

• Who added it,
• When they added it,
• Whether that adds up.

However, in order for this to be useful you might need to move the needle as to the amount of component parts you have in the system. You might need certain actors to have more control (e.g. in a decentralised social network where explicit or damaging material is added and needs to be removed), or you might need a little less transparency (you may not want all of your personal transactions broadcast to the entire world but like the ability to make them accessible to trusted third parties). You may want people in the network to ‘pay’ for their involvement, not with their data but instead with a bit of their hardware (storage and processing), using cryptography and private key pairs they should only be able to read and edit the bits they have authority over (more control), but it means the information is secure without the need for an army of firewalled, ‘completely secure’ data centres (anything with an IP address is arguably not secure).

This is why I have slowly moved further and further from the currency camps of the likes of Bitcoin, Ripple and Ethereum, increasingly favouring those that focus on the database value and operability such as Stellar or Eris.

So what could you apply this to?

Let’s take the failed experiment that was the NHS “Connecting for Health” programme. The idea was a good one; make everyone’s medical records available to over 30,000 GPs, 300 hospitals and every individual by a “Healthspace” web portal. In order to do this you need:

• Multiple (not necessarily trustworthy) actors to input data
• Multiple (not necessarily trustworthy) actors to read data
• An accurate consensus on that data
• A way of checking who had interacted with the data
• A way of restricting what data people can and cannot interact with

The solution: A huge centralised database providing secure and audited access.

This was doomed to failure from the outset: you have to include disparate records, from multiple areas while new data feeds in. The centralised datacentre is begging for an attack from hackers as the personally identifiable information (PII) is incredibly valuable, even more so if you add in that this is medical information, and most systems are woefully ill prepared for an attack even those considered secure (as shown by JP Morgan, Sony, and the Pentagon).

With the advent of the bitcoin protocol, the ability to secure this with public and private key technology became not only possible but, in my opinion, sensible. Furthermore, using an open source solution like Eris was unlikely to cost around £12bn and more importantly, would actually have worked as implementation of the database is no more than installing the software on servers connected over the public internet.

This could be built by small companies from local economies and requires a much lower hardware cost than the previous design.

For more information on the UK Digital Currency Association, please visit their website. Follow them on Twitter @UK_DCA.