Are Blockchain Art and Finance Really Decentralized?

Inspired by a recent post by Moxie Marlinspike on the subject of Web3, I wish to elaborate on the promise of decentralization of blockchain applications. I will focus on two major use cases of blockchain technology, namely art and finance, which beg the questions: What does decentralization mean? Are blockchain art and finance truly decentralized? And if not, what are the forces that drive us towards centralization? Is centralization inevitable?

What is decentralization?

According to the Oxford English Dictionary, to decentralise is “(1) to undo the centralisation of and (2) to distribute administrative powers, etc. which have been concentrated in a single centre.” The blockchain enacts exactly this transfer of control and decision-making from a centralized entity to a distributed network.

However decentralization — as a way to protect our privacy — is also a landmark for the cypherpunk movement. As much is clear from the following excerpt of The Cypherpunk Manifesto, written by Eric Hughes in 1993:

Privacy is necessary for an open society in the electronic age. [...] Privacy is the power to selectively reveal oneself to the world. [...] We the Cypherpunks are dedicated to building anonymous systems. We are defending our privacy with cryptography, with anonymous mail forwarding systems, with digital signatures, and with electronic money. [...] Cypherpunks write code. We know that someone has to write software to defend privacy, and since we can’t get privacy unless we all do, we’re going to write it. [...] We know that software can’t be destroyed and that a widely dispersed system can’t be shut down. 

In Information Science, a centralized organization is typically represented by the tree data structure. Trees are hierarchical structures typically associated with the following principles:

1. Centralism. The tree structure has a unique root node from which all other nodes descend. This expresses a concentration of power and authority in a central actor of the system.‍

2. Finalism. It describes the unidirectional, linear courses of trees: Any point in a tree is reachable from the root by a unique, linear, top-down path of intermediate nodes. Trees therefore embody an organization devoid of multilinearity or feedback loops.
   

On the other hand, a decentralized organization is generally associated with the network data structure. Networks are rhizomatic structures defined by the following principles:‍

1. Decentralization. Networks are webs without a spider: There exists no central authority or root that regulates their growth, but they evolve in a self-organized way.‍

2. Multilinearity. Networks allow feedback loops. This means that actors of the system can be reached through multiple paths.‍
   

Rhizomatic (network) structures offer a new model for knowledge and society that acknowledge decentralization, autonomy, flexibility, creativity, diversity, collaboration, altruism and, ultimately, democracy.¹

Networks match and sustain the proliferation of information typical of the postmodern condition — the co-existence of a multiplicity of heterogeneous discourses — instead of a simple, central discourse that unifies all forms of knowledge:‍

Those who have a nostalgia for a unifying metanarrative — a dream central to the history of Western metaphysics — experience the postmodern condition as fragmented, full of anarchy and therefore ultimately meaningless. It leaves them with a feeling of vertigo. On the other hand, those who embrace postmodernism find it challenging, exciting and full of uncharted spaces. It fills them with a sense of adventure.²

An important example of the modern perspectival shift from tree (centralization) to network (decentralization) is knowledge. Denis Diderot and Jean le Rond d’Alembert, in their Encyclopédie published between 1751 and 1772, depicted the genealogical structure of knowledge as a tree with three prominent branches: Memory and history; reason and philosophy; and imagination and poetry, wherbey only a small circle of scholars could participate as contributors. 

By contrast, the contemporary encyclopedia is incarnated by the knitted network of Wikipedia in which everyone can potentially contribute an article. As another example of this shift, corporations are evolving more fluid and dynamic models of decentralized autonomous organization (DAOs) out of the typical monolithic hierarchies that have characterized business in the past.

My intention here is to investigate if and how the decentralization promised by the cypherpunk movement is present in blockchain art and finance — arguably the two major applications of digital ledger technology.

‍

Decentralization in blockchain finance

The financial crisis of 2008 led to the collapse of a number of major banks, such as Lehman Brothers, and forced many states to bail out others out with public money. This crisis triggered an economic recession that has not yet been fully absorbed. On a psychological level, 2008 undermined the trust on which the relationship between citizens and their banks, and citizens and their states were based.

On Halloween of that fateful year, something else happened. A mysterious figure hiding behind the pseudonym Satoshi Nakamoto circulated an article, titled “Bitcoin: A Peer-to-Peer Electronic Cash System,” preceded by a statement that left no doubt as to their motives:

I’ve been working on a new electronic cash system that’s fully peer-to-peer, with no trusted third party.

On 3 January, 2009, Nakamoto’s idea was realized and the first block of the Bitcoin chain was laid. Indeed, they included in the first block an indelible message, quoting an article of the same day in The Times that a second public bailout was imminent.

From that point onward, blockchain finance was a reality. However, despite the original motivations of fully peer-to-peer operations with no trusted third party, a large share of blockchain finance today is in fact Centralized Finance (CeFi), and not Decentralized Finance (DeFi). A CeFi platform uses blockchain technology but it is owned by a private company. Users have to trust a third party to monitor transactions and ensure the security of assets. Such services also require one to submit one’s personal information for verification in order to operate. This is known as the Know Your Customer (KYC) process. 

The three pillars of blockchain finance are stablecoins, exchanges, and lending or borrowing services. The most popular stablecoins, like USDT and USDC, are centralized and are collateralized by the US dollar, which, like all fiat currencies, is centralized. USDT is owned by Tether, while USDC is owned by Centre which was in turn founded by Circle and Coinbase. 

In contrast, DAI is the first decentralized stablecoin, maintained and regulated by MakerDAO, a decentralized autonomous organization comprising the owners of its governance token, MKR, who may vote on changes to certain parameters in its smart contracts in order to ensure the stability of DAI (which is pegged to the dollar). DAI is collateralized by Ether and other cryptocurrencies (including some centralized stablecoins), and not by fiat money. 

Terra USD (UST) is an algorithmic stablecoin that is fully decentralized. Algorithm-based stablecoins have no associated collateral, but are governed by an open-source smart contract that helps to maintain their stability. It is possible to change the rules of the contract only by leveraging social consensus or through voting with governance tokens.

The most important cryptocurrency exchanges by trade volume, such as Binance, Coinbase, FTX and Kraken, are centralized under private business identities as centralized exchanges (CEX). This effectively means that, when a customer moves funds on these exchanges, the funds are deposited in the exchange’s wallet, and the customer doesn’t own them anymore. The customer is forced to trust the third party, as one would when depositing money with a traditional bank. All transactions cost a fee that goes directly to the exchange. The same is true of centralized lending and borrowing services, like BlockFi, Celsius, Ledn, and Nexo. In these instances, one receives a reward for making a deposit at a percentage rate or one borrows funds against deposit of collateral in cryptocurrency.

However, on a decentralized exchange (DEX) like Uniswap, customers rely on smart contracts in order to operate: To swap coins or deposit liquidity. Smart contracts are public, audited by third parties, and deposited on the blockchain. Hence, they always do what they are supposed to do (including executing bugs), and cannot be altered unless a new contract is deployed according to the will of a community of token holders. 

Market takers (those who swap coins) pay a fee that remunerates the market makers (the providers of liquidity). Notably, when a user swaps coins, the exchanged funds are deposited directly into their wallet. Similarly, when they interact with decentralized lending and borrowing services like Compound and Aave, all transactions are mediated by a public smart contract. When using these decentralized applications, users aren’t forced to trust a private company, but rather a public piece of code (the smart contract) and the developers behind it.

It is therefore wrong to say that decentralization is trustless. In fact, the target of one’s trust has changed, from humans to code written by humans.

‍

Decentralization in blockchain art

In one of the first articles about crypto art, Jason Bailey identified some common factors of this artistic trend, including democratization, whereby everyone is encouraged to participate regardless of their education, skills, age, class, gender, race, or creed; and decentralization, whereby tools and guidelines are designed to reduce the power of gatekeepers and middle men and increase the autonomy of artists. As an early crypto artist, I can attest that these were the original premises. But has crypto art kept its promises?

Despite crypto art’s original commitment to decentralization and equal distribution of resources, the crypto art market is now highly concentrated around a small number of artists and collectors. On SuperRare, I computed that 80% of the sale volume (between April 2018 and January 2022) was made by 14% of the top artists (Gini index equal to 0.83) and collected by 8% of the top collectors (Gini index of 0.89). By comparison, in the US, the Gini index of household incomes was 0.48 from 2014 to 2018. On other marketplaces, such as Foundation, the situation is similar or even worse. In other words, if crypto art were a country and sales reflected a population’s income, then wealth would be as concentrated as it is in the wider economy.³  

‍

The crypto art market, in statistical terms, follows a long-tail distribution, in which a vital few dominate the scene at the expense of the trivial many. Sales follow a cumulative advantage (or preferential attachment) process: “For to everyone who has will more be given, and he will have abundance. But from him who has not, even what he has will be taken away” (Matthew 25: 14-30). As a consequence, the democratic principle of crypto art does not hold up in practice.

Another point of concentration concerns marketplaces. In principle, an artist can write and deploy on the blockchain their own smart contract and use it in a pure peer-to-peer fashion to interact with collectors and sell their art. In practice, few artists have these technical skills and prefer to use smart contracts developed by marketplaces like SuperRare, Art Blocks and OpenSea. In all cases, only the NFT is deposited on the blockchain, while the artwork’s metadata and media containing the actual digital art are typically stored on either a centralized server such as Amazon Web Services (AWS), as in the case of Art Blocks, or on the InterPlanetary File System (IPFS). The transaction minting the NFT on the blockchain records (but doesn’t verify) the link between the token and the metadata and artwork files.

It is true that IPFS is a decentralized, peer-to-peer network. Yet there remains no incentive for nodes on IPFS to share and download files from other nodes, with the consequence that artwork files are typically stored only on the node of the artist (if they can run a server with an IPFS node) or more likely on that of the marketplace. It is possible to mint an NFT with a custom smart contract, to pin the artwork on IPFS with a free service like Pinata, and then immediately remove the file from the Pinata node. In all likelihood, no other IPFS node has grabbed the file in the meantime, and the minted NFT is probably linking to nothing.

That said, a Web3 marketplace is fundamentally different from a centralized Web2 application. Neither the NFTs nor artwork files are stored on a centralized server. The only centralized element is the front end of the marketplace website. The case of Hic et Nunc exemplifies this. When its creator, Rafael Lima, decided to shut down what was a highly successful NFT marketplace — not to mention a progressive ecosystem based on the Tezos blockchain — many users felt scammed and some even claimed it was a rug pull. Yet only the front end of the website disappeared, whilst all the smart contracts and NFTs were still immortalized on the blockchain and visible via other NFT aggregators, like Objkt.com. 

‍

The forces of centralization

Moxie Marlinspike, the founder and CEO of messaging app Signal, recently addressed the effective decentralization of Web3. In his view, the Web3 architecture follows the client-server model but virtually all clients who wish to access it do so by trusting the outputs of servers run by a small number of companies like Etherscan, Infura, OpenSea. However, this inevitably leads to some form of centralization because:

1. People, even the “nerds” described by Marlinspike, don’t want to run their own servers and never will.

2. The burden of building software is high. At this point, software projects require an enormous amount of human effort.

3. A protocol like the blockchain moves much more slowly than a platform like OpenSea. If something is truly decentralized, it becomes very difficult to alter, and often remains stalled in time.
   

If, as Marlinspike indicates, enough money has been made at this point that there are enough faucets to keep the NFT system going, it is worth reflecting on how to avoid Web3 becoming Web2 but with even less privacy and efficiency:

Once a distributed ecosystem centralizes around a platform for convenience, it becomes the worst of both worlds: centralized control, but still distributed enough to become mired in time.

After Marlinspike posted his critique, Ethereum founder Vitalik Buterin replied with a partial rebuttal on Reddit: “Moxie’s critiques in the second half of the post strike me as having a correct criticism of the current state of the ecosystem…[sic] But they are missing where the blockchain ecosystem is going,” adding that the reason Web3 has yet to live up to its own ideal is due “to limited technical resources and funding.”

Centralization is delegation, decentralization is responsibility

Owning cryptocurrency is emblematic. Buying one Bitcoin affords two options to store it. Either one can make a deposit on a centralized exchange like Binance, or save it in a physical wallet like Ledger. The first involves delegating to Binance all issues related to custody and security of the cryptocurrency. For example, if a user forgets their password to the Binance app, they will receive a new one as soon as Binance certifies their identity. However, as the recent confiscation of crypto wallets tied to the Canadian “freedom convoy” proves, delegating responsibility is not the same as ownership.

The second approach involves purchasing a Ledger wallet, learning how it works, and carefully securing the pass phrase associated with the wallet. The user remains  the only owner of their Bitcoin but with the added responsibility to protect it: If someone steals my passphrase, or I simply forget it, I might find one morning that my Bitcoin is not mine anymore.

As Marlinspike hinted in his post, not all individuals have the will, skill, and time to adapt to a complex interdisciplinary field like blockchain. A dose of delegation and centralization is therefore inevitable in even the most decentralized system. This is not a bad thing, so long as we understand what is going on behind the scenes.