BEGIN ARTICLE PREVIEW:
Blockchains solve a hard problem: how to ensure consensus across a distributed, decentralised network, where messages arrive out of order if at all, where individual nodes may fail, and where a certain proportion may be actively malicious.
The original blockchain, bitcoin, was designed to support a novel digital currency, and the issue its consensus algorithm solved was preventing double-spend. It also successfully introduced game theory for security: adversaries would have to spend more money on an attack than they could expect to gain financially. All this and the original protocol was just a few hundred lines of code.
But this achievement came at a high cost in terms of energy use and performance.
With bitcoin, a new leader is required to verify each block of transactions, that leader being the first device to complete a computationally heavy challenge (Proof of Work, PoW). As a result, the blockchain’s throughput is painfully slow at around seven transactions per second (Visa claims it can do 56,000) and the whole process is massively wasteful of energy. These drawbacks have been surmounted, to some degree, in newer blockchain designs using overlay networks, sharding and different types of “proofs of” and by non-blockchain directed acyclic graphs (DAGs), …
END ARTICLE PREVIEW