Achieving Decentralized Scale: ETH vs BTC vs ALGO

We will in the following paragraphs use Bitcoin, Algorand and Ethereum as examples to argue that If blockchain technology is to fulfill its purpose as a trusted shared public ledger and its promises as “General Purpose Technology” it has to have  the right “incentives mechanisms” (Vitalik Buterin) in place to leverage its access to human ingenuity in order to achieve scale while fighting off many possible emerging points of centralization arising from potential economies of scales.

Blockchains can nowadays have many forms from private and permissioned to public and permissionless or anywhere in between.

While a public centralized blockchain, assuming a completely trusted and trustworthy central authority that does not seek to profit from its position or the “cost of trust”, could in theory serve as a shared public ledger and reduce the costs of verifying the attributes of a transaction or information leakage (Catalini & Gans) it would most likely fail in having the networks effect of a truly decentralized one.

In fact, its is very likely that even if a centrally designed public blockchain was to achieve scale most of the technology´s intended purpose and benefits to the public would fade over time or disappear instantly.

One of the main benefits of blockchain is its ability to create assets that cannot be owned (Catalini & Gans). Centralizing ownership of the ledger and its network effects would surely result in the central authority charging the public for the cost of trust and exert market power which could come in the form of fees, censorship, or other inefficiencies. Those inefficiencies would in turn reduce participation undermining the technology´s very purpose. Centralization also creates “single point of attacks” undermining security and therefore trust, once again undermining the core principles of the technology.

The debate about centralization comes as many blockchain including Bitcoin have failed to scale without centralization. Scaling a blockchain by extending the parameters makes it harder to participate (Vitalik Buterin) thereby increasing centralization and reducing safety.

We agree with Professor Silvio Micali that a lot of the scaling and centralization problems of Bitcoin are inherent to the incentive scheme embedded in its protocol. In Bitcoin, miners are incentivized and rewarded to secure the network by transaction fees and newly minted coin for their POW work. The number of resources needed to mine under Bitcoin POW has resulted in barely profitable and unfamous centralized mining operations while the requirement of running a full node is affecting participation in the network. While miners are remunerated, “Full Nodes” which have to run a full copy of the blockchain are not and neither are “Lightnodes” even though they must keep a full copy of the UTXO set (affect memory vs disk space), the network´s most expensive resource. (Andreas Antonopoulos).

In fact, one could argue that nodes are encouraged to hoard the coins as their only remuneration will come from the appreciation of bitcoin.

According to Wikipedia, 0.5% of Bitcoin wallets owned 87% of all Bitcoin in March 2018. This concentration of mining and ownership could be the inadvertent result of a combination of Bitcoin deflationary fixed monetary supply and incentive mechanism, but the fact is that it has resulted in very little innovations apart from Segwit and Lightning Networks which in the latter case only offer an off chain centralized solution scalability problem.

While Algorand´s “Pure POS” consensus “algorithm” uses randomness and cryptography as a source of trust to create a scalable blockchain solution that not only allows high transaction throughput but also a cheap participation (Silvio Micali), it remains to be seen whether “Pure POS” gather enough active participants to secure its network without suffering from the “tragedy of commons” or “nothing at stake” as no incentives are in place for good behaviors and no punishment for bad ones. Being rewarded for merely holding and using ALGOs could well be enough of an incentive.

To solve its scalability issue Ethereum is switching from POW to “Bonded POS” consensus mechanism while putting in place new technologies such as “Sharding” which split the database and “ZK Rollups” which allows for a large part of the computation and data to be done and kept off chain. Zero Knowledge Cryptography (also used in Zkledger) has tremendous scalability potential as it cryptographically proves the validity of a computation or message without the need to reveal the information. When combined, those measures should; by pushing 90% of the data and 99% of the computation off chain, enable 100,000s of transactions per second (Vitalik Buterin).

By actively fighting and identifying emerging sources of centralization like “MEV” (Miners/Stakers Front-Running etc..) so they only exist for a limited time Ethereum is showing that the with the right incentives in place human ingenuity can overcome Blockchain’s current challenges and fulfill its initial purpose: Individualism is funded and promoted by the public as a way to develop innovative and transformative applications of general purpose that remain in public hands.