Three blockchain benefits, above and beyond traditional databases

If you’re thinking about business opportunities for blockchain technology, it’s useful to avoid getting over-excited about an idea that a traditional centralized database could do just as well. How do you parse which is which?

In building BigchainDB, we started with a traditional centralized database that had scale and built-in querying, and added blockchain characteristics. This forced us to think deeply about the characteristics, as benefits. We asked the following.

What specific benefits do blockchains bring compared to traditional databases? We found three.

1. Decentralized / shared control,
2. Immutability / audit trail, and
3. Assets / exchanges.

You can turn the crank on any one of these benefits to find potential application in a given domain. Let’s explore further, with examples.

Benefit 1: Decentralized / shared control

This benefit comes in three forms.

First, blockchainsallow enemies to work together for a common benefit. It makes blockchains a political tool. Examples include: R3 for banks, or Open Music Initiative for the music labels.

More generally, blockchains can reduce friction in forming consortiums, because the entities involved do not have to give control of the infrastructure to some single entity. So for some contexts, a blockchain is a “consortium database.”

Second, when the network is decentralized and sufficiently open and for anyone to use, then it becomes a new public utility, like electricity, the internet, or the Web. Examples include Bitcoin as an e-money or e-gold utility, Ethereum as a decentralized processing utility, IPFS as a decentralized file system utility, or IPDB as a decentralized database utility.

Third, interoperability protocols can further reduce friction, especially for public networks. If networks interoperate, then which network you’re using matters less. You aren’t necessarily stuck with your initial choice of network. The internet is a network of networks; do you care which sub-network you’re in? Two key protocols are:

• Interledger. It connects two networks, even if those networks speak a different language. You can think of it like a well-defined protocol for an exchange. Or, like a router for the internet, but instead of handling just data, it also handles assets. How: escrow in the forward direction from source to target, then release funds backwards. That’s basically it. But it enables easy connection between Bitcoin, Ethereum, SWIFT, BigchainDB, and more.
• IPLD (Interplanetary Linked Data). This allows data blobs to flow through the walls of one network to another. It makes the networks permeable. How: an opinionated way to hash the structured data format JSON. That’s basically it. But it means you can initially store your JSON data on, say, IPFS, but know that it can also flow to IPDB or whatever network supports IPLD.

Benefit 2: Immutability / audit trail

When you write data to a blockchain, it’s like etching the data into stone. This could can be used for education credentials (like Recruit), land registries (like BenBen), and more.

If you have a series of transactions over time, you gain an immutable audit trail, which is useful for financial audits, art provenance (like ascribe), food history (like Provenance), and more.

Benefit 3: Assets / exchanges

Once you have a data store that no single entity owns or controls (i.e. decentralized), and no one can change what’s already written (i.e. immutable), then it unlocks the possibility for assets themselves to live on the data store. Put another way: it gives benefits similar to double-entry bookkeeping (less chance of errors, etc.) but much easier to apply.

Once you have native assets, you’ve lowered the friction inimplementing exchanges. Blockchain application in traditional exchanges include stocks (NASDAQ / Chain), currency (like Bitstamp), or energy (like RWE / EEX / GridSingularity). Multi-sided platforms are also exchanges; examples include social media (like Synereo), ride sharing (like Arcade City), and online retail marketplace (like OpenBazaar).

Appendix

This section is for the nerd purists. One can make an argument that some databases traditionally did support each of {decentralization, immutability, and assets}. And they’d be right! But only now are these ideas gaining momentum. I’ll elaborate.

• Decentralization for permissioned blockchains — that’s Byzantine Fault Tolerant databases, as Leslie Lamport theorized in the early 80s, and with the first practical algorithm by Castro & Liskov in ’99. For the public blockchains, credit is certainly due to Satoshi for at least two things: a novel approach to Sybil attacks (“attack of the clones”); and a new approach to the CAP thoerem by using economic constraints to maintain both availability and (pseudo) consistency in the face of a network partition.
• Immutability work goes back decades too. Interestingly, even RethinkDB was going for immutability in its early days. Readers of the BigchainDB paper will know that I prefer the label “tamper resistance,” as “immutability” is too absolutist.
• Finally, the concept of assets is closely related to transactional databases, which goes back decades too. And banks and financial institutions do store assets on their databases and already have double-entry bookkeeping. It’s just not as clean.

So, is any of this really that new? Probably not! But it certainly was under-appreciated. Bitcoin and the movement it sparked has raised awareness to some, and helped everyone to better appreciate the importance of these characteristics. It certainly inspired us: we built BigchainDB to bring these revived ideas into modern distributed databases.

Trent McConaghy is founder & CTO of BigchainDB, the scalable blockchain database. In the past, he designed AI algorithms to help drive Moore’s Law and explore man-machine creativity.