The Aesthetic Measure of an NFT

With these words, Sol LeWitt established at the end of the 1960s that it was not the final execution of a work of art that was decisive, but the underlying idea. When someone purchased a work by LeWitt, they received a document with instructions on how to install it on the wall. The question of who ultimately executed the work of art was considered an irrelevance. The value lay in the concept and not in the materialized wall installation.

His contemporary, Lawrence Weiner, subsequently formulated this creative principle in more detail in his well-known “Declaration of Intent” of 1970:

1. The artist may construct the piece

2. The piece may be fabricated

3. The piece need not be built

Each being equal and consistent with the intent of the artist, the decision as to condition rests with the receiver upon the occasion of receivership.²

In retrospect, the instructions set out by these artists were nothing more than an analog prototype of today’s digitally automated smart contracts that deploy blockchain technology.

The conceptual art of LeWitt and Weiner emerged at a time when the first mainframe computers were being developed and pioneers of computer art such as Frieder Nake, A. Michael Noll, and Herbert W. Franke were drawing the first plotter drawings on Zuse Graphomats. In these instances, an algorithm was used to specify the parameters and the machine completed the work of art. Mathematical serendipity played a major role. In his theory of computer art, Max Bense — who coined the term “Generative Aesthetics” — did not speak of works of art in the conventional sense, but rather of aesthetic states or structures that were generated by a machine according to a program.

With an aesthetics based on mathematics, the quality of an aesthetic state demanded a rational formula. George D. Birkhoff saw the “aesthetic measure” (M) of a work of art in the ratio of complexity (C) and order (O). His formula was: M = O/C. Yet the question always remained as to whether the machine was really capable of replacing the innate creativity of humans. 

In 1965, the American artist A. Michael Noll attempted a Turing simulation based on Piet Mondrian’s Composition in Line (1916/17), which consisted of vertical and horizontal lines of different lengths. According to Turing, whose Automatic Computing Engine (ACE) was one of the first designs of computer, a machine has an artificial intelligence insofar as a human cannot distinguish whether an aesthetic product derives from the machine or another human. 

Noll put his audience to the test by juxtaposing a copy of a real Mondrian with an artificial Mondrian created by a computer according to the same parameters, asking the public to judge which of the two was made by a human and which by a machine. Reviewers were also asked which work they preferred. In the event, only 28% of respondents recognized the real Mondrian with well over half preferring the computer-generated version. 

While early computer art of the 1960s developed some interesting experiments and novel forms of algorithmic aesthetics, it was never accepted into the postmodern canon of art and remains largely absent from most museum collections today. With the exception of a few galleries and collectors, there wasn’t a significant market for digital art until 2017 when the ERC-721 allowed digital artworks to be tokenized as NFTs (Non-Fungible Tokens) on the Ethereum blockchain. While a digital work of art is still reproducible, possession of the token in a digital wallet can be used to determine the rightful owner of a work.

In 2019, Noll’s computer compositions were reincarnated in the form of Autoglyphs — the first NFTs whose digital asset is stored on the Ethereum blockchain. This series of 512 generative glyphs were themselves the product of an experiment by John Watkinson and Matt Hall, better known as Larva Labs. The duo’s breakthrough came in the summer of 2017 with the introduction of CryptoPunks. The product of a random number generator, at the time these 10,000 different characters could be accessed free of charge, reminding us of the NFT’s early gift economy. Their pixelated heads in three-quarter profile have various distinct identities and attributes: From zombies to aliens with beanies, hoodies, whistles, etc. 

CryptoPunks are now traded in the millions of dollars on the secondary market, with their price usually determined by a simple mathematical principle. Their different value categories depend on the rarity of the attributes. Among the most expensive CryptoPunks are the aliens, with only nine in total. Then come the apes with 24, the zombies with 88, followed by certain rare combinations or especially popular Punks — characterized by beanies or hoodies — with which the tech community in particular strongly identifies. It is common knowledge that every new employee of Google receives a beanie hat with a propeller that looks exactly like the CryptoPunk beanie.

The name of the art project also harks back to the cyberpunk and cypherpunk movement of the 1980s and ’90s, which anticipated the libertarian precepts underlying Bitcoin. In under four years, more than 20,000 CryptoPunks have sold on the secondary market, with a total turnover of over $2 billion. Reliable rankings and price estimates are now available on various platforms based on the liquidity of the market, existing sales data, and the attributes specified in the artworks’ metadata. 

With the Punks’ growing popularity, their collectors have started to adopt them as avatars across social media channels. On Twitter, a comparison with the owner’s digital wallet verifies the owner’s claim to the image. That is, unless they’re a right-click saver. Profile Picture Projects (PFP) have consequently become the digital status symbols of our time. 

Perhaps inevitably, the CryptoPunks have spawned countless spin-off projects, with Bored Apes currently the most sought-after collectibles. Their hype has been driven by star members of the Bored Ape Yacht Club (BAYC) such as Steve Aoki, Justin Bieber, and Eminem. One can argue about the artistic merits of such work, but when it comes to aesthetic measure, everyone is in agreement. What counts is rarity. In this respect, NFTs are comparable to the Bitcoin principle of artificial limitation, by which NFTs are now considered a reliable store of value by investors around the world.

Looking back at the Mondrian experiment of the 1960s, an interesting parallel has emerged in Mario Klingemann’s project, Botto, launched in October 2021. The German artist has created the first autonomous AI artist based on a DAO (Decentralized Autonomous Organization), whose community’s consensus determines the artwork auctioned weekly on the SuperRare marketplace. The proceeds from these sales flow back into the DAO to sustain its creative ecosystem. Members must, however, own BOTTO coins in order to vote. Botto’s ultimate stylistic and aesthetic orientation therefore corresponds to the tastes of its DAO members, exercised according to the credit of contributors. In this context, aesthetic measure corresponds directly to a new form of democracy. The work of art has become a reflection of the prevailing tastes of its public.