How We’re Bootstrapping Marconi
In the previous post we explained how 51% attacks pose a major problem for new proof-of-work networks. Here we’ll explain our plan to address this problem.
An attacker requires two essential properties in order to pull off a 51% attack. The first is the ability to select which transactions make it into each block. The second is anonymity. The former provides an attacker with the ability to generate nefarious blockchain forks. The latter prevents an attacker from being identified and prosecuted.
Normally any miner possesses both of these properties. Interestingly though, there are common situations accepted by the blockchain world where these properties no longer hold. One case is mining pools, where miners contribute their hash power but no longer have control over transactions in each block. Another case is Know-Your-Customer (KYC) processes, where a user gives up anonymity in order to gain the ability to interact with a financial enterprise or to invest in specific products. KYC has almost become standard with all blockchain investments. Even Tezos, which did its investment last year, ended up retroactively requiring KYC for token investors when it launched its network.
For Marconi, to protect the network while it grows, we’re going to allow each miner to choose either one of these properties, but not both. If they value anonymity, they can point their hash power at a mining pool, and just like with Bitcoin and other proof of work chains they’ll be unable to choose which transactions make it into a block. If on the other hand they value the ability to choose transactions in a block, or if they want to run their own mining pool, they’ll have to complete a KYC process to help ensure they remain a good actor.
In addition, to help jump-start the mining community, the Marconi Foundation will be setting up several mining pools which do not charge any fees. Any miner can point their hash power at these pools and enjoy their fee-less rewards.
Overall this setup will, in the early days, make Marconi similar to delegated proof-of-stake chains (e.g. EOS, BitShares), in which users vote to determine who is allowed to create new blocks on the chain. With delegated proof of stake, voting is based on how many coins a user has, whereas with Marconi, voting is based on the user’s hash power. And in the case of Marconi, users have a strictly larger set of options, since they may decide to fully participate in the block creation process by going through KYC, rather than redirecting their hash power as a vote to signal which mining pool they think should be creating new blocks. The KYC case is pure proof of work, while the redirection case can conveniently be thought of as delegated proof of work since miners are delegating the power to create blocks to whichever mining pool they deem worthy.
Finally, once the network reaches a robust size, these safeguards will become unnecessary and will therefore no longer be in effect. We think this solution will work well for the various stakeholders in the ecosystem, and we look forward to announcing the specifics as we get closer to our launch.