TPan here. I’m busy learning about aliens today, so I won’t be writing my usual piece. However, I do have some great content for you from someone else in the space.
One area of web3 (out of many) that I’m less familiar with has been generative art, and that explains why you don’t see as much content about that topic coming from me. As I’ve spent more time here, I’ve grown to appreciate the value and impact generative art brings to the table.
Today we have Nat Emodi sharing his thoughts on this subject. Nat is the Co-Founder and CEO of Highlight, a platform for creating generative art and other types of NFTs on Ethereum and compatible L2s. He’s also been a long-time Web3 with TPan reader! It’s been a pleasure getting to know him and following his journey building Highlight.
If you’re interested in any of the themes in this article, I’d encourage you to check out Highlight and mint the genesis collection (only ~$3 plus gas). You can also find Nat on Twitter here.
“Sum” by Duane King | 42-character hexadecimal Ethereum wallet addresses rendered as generative art
In the midst of the bear market, generative art projects — art created using autonomous systems such as code or AI — are thriving. Look no further than QQL or Opepen for a glimpse of successful projects. While overall liquidity continues to be a shadow of the bull run, the generative art community is actively growing and engaging, bucking the FUD facing other types of NFT projects.
I’m a founder, builder, art collector, and friend of TPan’s and today will make the case that generative art NFTs created using code or AI systems are 10x better than the average monkey jpeg. While your bags may be down, for a glimpse into the bright future ahead, I believe that the core UX around generative assets points to the future of digital ownership.
What is generative art and why is it thriving?
Generative art is a longstanding tradition in the art world. It refers to art that, in whole or in part, has been created with the use of an “autonomous system.” An autonomous system in this context is generally non-human — commonly code, machines, or devices — and can independently determine features of an artwork that would otherwise require decisions made directly by the artist.
Imagine a computer program, written by an artist. The program can run an infinite number of times and produce a rendered image that is relatively similar but not identical each time. The program and code is autonomous from the artist, and is the medium for creating the art. The artist designs the system, but then steps back and gives over control for creating the art to it.
With blockchains and NFTs, this type of art can now be collected. Instead of an infinite number of outputs, artists are able to create projects with finitude: a set number of pieces, a set timeframe for minting, or other limits that impose scarcity and increase the fun and value of collecting.
With the textbook definition as background, the term “generative” gets applied to NFT projects haphazardly, and not all references mean the same thing. Broadly, “generative” is a label applied to three types of NFTs:
Code-based art: these projects are written in programming languages such as Javascript. While as an art form this has existed since the earliest days of computing, it’s become popularized as collectible NFTs by Art Blocks and other platforms.
AI art: using neural networks and other models, and often with extensive post-production editing, artists generate a series of generative outputs. Brain Drops is a favorite example of a platform that curates this type of artwork.
PFP projects: though not purely generative because of the level of human curation involved, PFP-style projects that layer different characters, apparel and accessories from a set of image files are often referred to as “generative” as they are often produced using a script or other tools.
Since I’m writing today instead of TPan, we’ll leave PFP-style projects like Bored Apes out of this post (love you, Apes.)
As a category, generative art is also starting to ripple into the broader art world. In June, the (in)famous “Goose” output of Dmitri Cherniak’s celebrated Ringers collection sold for $6.2M at Sotheby’s. While the buyer was a well-known digital art collector, the sale in the midst of the seeming doom and gloom of crypto winter caught many people’s attention.
Ringers #879 (The Goose) by Dmitri Cherniak | Sold at auction by Sotheby’s for $6.2M on June 15
Others in the broader web3 space are starting to take notice. Mercedes-Benz recently collaborated with pioneering artist Harm van den Dorpel and Fingerprints DAO on a 3-D generative project, created using a neural network, called Maschine which draws inspiration from themes of movement, velocity, and perception — ideas core to Mercedes product line. The project is inspiring an exploration of digital art displays, based on a car owner’s wallet, in vehicle displays.
Maschine #118 by Harm van den Dorpel in collaboration with Fingerprints DAO & Mercedes-Benz NXT | Live rendered with interactive animation here
Meanwhile, a proliferation of tools and experiments have emerged on Ethereum and L2s to fuel the passion of artists and collectors alike. These include Art Blocks, Alba, and my own platform, Highlight, all on Ethereum, as well as fxHash (Tezos), Prohibition (Arbitrum), and EditArt (also Tezos).
Why is generative art 10x better than other NFTs?
Generative art created through code and AI systems are emerging as the most interesting and exciting part of the 2021 bull run. While VCs and threadbois like TPan wax poetic about the meta defining a new era of digital collecting, any 30,000 foot theory should be rooted in the important ground truth required for any game-changing technology shift: a magical consumer UX. More specifically, how do the moments spent minting a given NFT feel? For generative art, the experience is electric and converts casual speculators into generative art believers, and in some cases (my own included), addicts.
The key components of this UX are important to unpack because they provide a mental model not just to understand which NFT projects are most interesting, but also the broader multi-year change in consumer experiences around generative and AI systems that’s beginning to unfold.
If you peel back the layers of the generative UX, you’ll find a few compelling dynamics:
Dynamically revealed. No one – not even the artist – knows what will appear until the collector mints. Things get revealed, which unleashes a massive amount of kinetic energy as everyone discusses and compares what happened, who got what, surprises, etc., after the mint.
Fully (or largely) on-chain. Code (eg: javascript libraries such as p5.js or SVG) in the smart contract renders in a browser. The NFT doesn’t require a static file in an off-chain file store. The art is as permanent as the blockchain it’s based on, which makes it more valuable in the eyes of many collectors.
Co-created. With code-based art, a seed — a piece of data – is required to generate a random number that’s needed to run the algorithm. Typically, this means the collector’s transaction hash, generated at time of mint, is a direct, deterministic input that creates the artwork. Without the collector pressing the mint button at that exact moment in time, the NFT output wouldn’t exist. Unlike any other art form, the minter summons the work, forever tying themselves to the piece.
An intrinsic store of value. Art is one of the oldest stores of value. Onchain generative art is a paradigm shift in the traditional art world because it’s a new art form that can’t be made without the blockchain. MoMA, Christie’s, Sotheby’s, and dozens of other major art institutions are now collecting generative art on blockchains. Meanwhile, as we’re seeing with PFP projects, when prices drop, critics dismiss them as having no intrinsic value unless there’s real-world utility associated with owning them.
Made by exceptionally talented artists. The world’s top generative artists are exceptionally talented creatively and produce work that takes months of painstaking development and deep understanding of programming, rendering, hardware dynamics, and more. Imagine a fine oil painting and the years of skill and technique that are required to create something beautiful. Generative artists are similarly talented in terms of mastery of their medium and the effort required to create something exceptional.
Transcendence. One of the most interesting things about code-based and AI generative art is the concept of “transcendent grails,” related to the deeper concept from the sciences of emergence. This term refers to surprising, human-readable outputs that appear unexpectedly and beyond the control of the artist. Some famous examples of transcendent grails include the (in)famous “Goose” output from Dmitri Cherniak’s celebrated Ringers collection, shown above; the “Starry Night” output of Matt Deslauriers’ incredible Meridian collection; and The Tulip from Tyler Hobbs’ iconic Fidenza collection. With generative art, artists typically run through thousands or even tens of thousands of outputs before a collection is finalized. The probability of an emergent work — a “transcendent” grail — is low. And while some see these as a Rorschach test, the mimetic appeal of these pieces helps the collections they are from gain greater fame and appreciation.
Other non-generative facets of the NFT space have up-leveled their categories in various ways: music, photography, collectibles, and more now have on-chain provenance, interoperability, improved monetization, tradability, etc. These are significant, laudable improvements. But the core experience around creating and collecting generative art is particularly special, and can serve as a roadmap for pushing the broader concept of digital ownership forward.
Nat lays out the mechanics of generative art in a simple and understandable way. If you’ve ever heard the saying ‘code is art’, things start to make more sense in this context.
If you enjoyed this piece from Nat, check out Highlight and mint the genesis collection.
See you next week!
