Blockchain stores the data interacting between machines in a block, and chains it to the previous ones after verifying its validity using the consensus system in the machines. Encoding the transactions in the prior blocks makes it technically impossible to counterfeit and alter transactions and the related transaction information stored in the block. A chance to attack the algorithm remains at disturbing the consensus process before a block is chained to the prior block, but it is defendable because infecting 51% of a decentralized system is economically infeasible. This issue briefly addresses the fundamental assumptions of the consensus protocol related to the reliability of blockchain.
The first part of the issue briefly sketches the mechanism of centralization using blockchain algorithm. One of the keys to blockchain is that Trusted Third Parties are replaced with the consensus algorithm to select the reliable node (or an individual machine). The algorithms are classified into either a competitive type (e.g., Proof-of-Work) or a cooperative one (e.g., Stellar Consensus). Although the competitive process stimulates the growth of the blockchain ecosystem, the incentive scheme can centralize the distribution of hash power and wealth, and there exists a trade-off between size and distribution. In other words, it is difficult for decentralization algorithm to avoid the faith of centralization.
The second part describes the architecture of blockchain that requires a hub node to source the data off-chain. The problem arises from the fact that data off-chain could be false, and the blocks with defects are false. However, blockchain algorithm verifies the integrity of data on-chain. Blockchain inevitably requires the data off-chain to extend its business models to various service fields (e.g., currency ratio), distributions (e.g., groceries), and physical products (e.g., automobiles). In other words, blockchain calls for some special nodes that guarantee the reliability of data off-chain, which are another Trusted Third Parties that blockchain have expelled from verification.
In summary, blockchain, a decentralization algorithm, induces centralization of information. This is a structural problem rather than a minor side effect, casting doubt on why we have to use blockchain, if it brings centralization that it has assured to replace. This issue briefly demonstrates that machines distributed in a system will replace the Trusted Third Parties, without a doubt, due to its efficiency. However, it reveals that a decentralization algorithm induces centralization in the power of information generated within as well as out of the system. A new regulatory framework is then called for in governing information imbalance in a decentralized system to secure the trust endorsed by machines.