Fixing Taggr community's broken update issue – announcing imminent NNS proposal

I wouldn’t say that ETH is much further along in the decentralization process, and that isn’t meant to throw shade at them, because we love Ethereum.

One of the greatest misconceptions in crypto, is that having thousands of validator nodes running on Big Tech’s cloud infrastructure represents decentralization. I haven’t see the latest stats, but articles like this provide some color from last year 3 cloud providers accounting for over two-thirds of Ethereum nodes: Data – more than half run on Amazon Web Services.

Currently I’m seeing 560,000 validator nodes being reported by Ethereum Ethereum staking | ethereum.org

But, based on that previous breakdown, we might expect 300,000+ of those nodes to be running on Amazon Web Services, which in reality, is just one giant node machine controlled by a corporation. Sure, the network is replicating computation and data hundreds of thousands of times on that giant node machine controlled by the corporation, which is expensive and makes them lots of money, but it doesn’t increase the Nakamoto Coefficient of the network. Amazon could take control of those validators, or switch them off at any time, just like Hetzner just did to 40% of Solana’s network Is Solana Decentralized? Cloud Provider Hetzner Ban Raises Questions - Decrypt

Such Proof-of-Stake architectures result in a situation where the number of real nodes is vastly lower than the number of validator nodes advertised (and I mean VASTLY, as in a tiny fraction of a percent). Many whales also run huge number of individual validators – so validator count is a very misleading measure of decentralization.

Moxie, the developer of Signal, made a really insightful post about Web3 decentralization on these platforms, see Moxie Marlinspike >> Blog >> My first impressions of web3

Proof-of-Useful-Work, a la the Internet Computer, which builds a network on standardized dedicated (sovereign) hardware run by independent parties, which identifies those node providers, so that “deterministic decentralization” can generate required Nakamoto Coefficients more reliably while minimizing replication and cost has major advantages, and one of those is surely increasing the reliability of decentralization.

Likewise, in the end, using an opaque system for updating blockchain networks through network forks, which in are ultimately coordinated by project insiders, also has issues regarding decentralization (aside from the technical risk, inefficiency, politics and so on).

That’s why the Internet Computer network has a permissionless DAO at its core, which automates updating and configuring the network in a transparent way, as well as reducing technical risk, and efficiency – allowing for adaptive operation and rapid evolution.

We still live in a world where crypto promotes thousands of anonymous Proof-of-Stake validators, which are mostly running on Big Tech’s centralized cloud services, as a mark of decentralization, and commercial incentives have persuaded many this is the case, but it’s really the opposite. That’s also true of updating networks via forks.

That the Internet Computer runs on a network of sovereign hardware (much like Bitcoin whose blocks are produced by dedicated ASICs), and is updated and configured by the Network Nervous System DAO, is something this community should be immensely proud of.

For those that want to see a true World Computer, which allows people to build on the decentralized internet, rather than Big Tech’s centralized infrastructure, and allows Web3 builders to create end-to-end decentralized services that run fully on-chain under the control of DAOs, this is the real deal.

The Internet Computer pushes the boundaries of what a compute platform can do, enabling people to build using tamperproof, unstoppable and autonomous smart contracts that can process tokens, which here can do things like web serving and http outcalls, but it also pushes the boundaries of the decentralization of blockchain networks.

12 Likes