The Countdown to Blast Off is on! 
Join the 400M $SUPRA token airdrop
Robot Image

Generative Art NFT: An Introduction & Examples

July 21, 2022 - 9 min read

Generative art can most easily be understood as having been produced in collaboration between artists and computer software which produces digital art based on sets of given parameters.

generative art NFTs


The rise of blockchains and digital ownership, NFTs in particular, has made it possible for a generation of artists to better safeguard their work and monetize it. This has drawn artists new and old to Web3 as they can display and sell non-fungible versions of their art via smart contracts, where collectors and speculators can buy and sell these digital pieces on secondary markets. 

Since the rise in popularity of NFTs, the term generative art is increasingly being applied to prolific NFT collections like Bored Apes, whose features are algorithmically generated based on provided rarity scores for their characteristics. 

Generative artists even have their own dedicated NFT platform, Art Blocks, where they upload algorithms that buyers can use to mint their own NFT iterations on Ethereum’s blockchain. Art Blocks has generated hundreds of millions in sales for artists, which is extraordinary for digital creatives who have until recently struggled with how to prevent their art from simply being copied.

As they are recorded on blockchains with characteristics of non-fungibility, generative art enjoys a new layer of complexity where users, creators, and code all engage in a certain kind of creative dialogue. By minting a generative art NFT, the buyer tells the code to generate a unique image which becomes an NFT.

Generative Art Genesis: Harold Cohen’s AARON

Modern applications of generative art can be traced back to the 1960s when artist Harold Cohen began exploring the concept, inspired in part by a graduate student where he worked at the University of California, San Diego. Programmers at the time used a system of instructions (FORTRAN) based on punched cards fed into ‘batch process machines,’ after which the computers produced newly punched cards or printouts

Cohen applied his artistic passion to this technology and began programming computers to draw pictures for him. By 1971, he had developed a functioning system called AARON which he presented in a special exhibit at the Fall Joint Computer Conference in Las Vegas.

The exhibition, despite its presentation at a tech conference, marked a new beginning for Cohen as an artist-programmer-researcher. This new beginning is characterized by his total switch-over from painting as a profession to programming-for-painting. It was also the origination of what would eventually become generative art NFTs; history was about to be made.

generative art harold cohen AARON
A comparison of Cohen’s AARON drawings from 1976 (left) to the color-enabled version unveiled in 1995. Image credit: Collection of the Computer History Museum, Catalog # 102627449 and #102741168

In a 1988 paper, Cohen wrote that AARON is best conceived of as a tool for researching and expansion rather than an all-encapsulating finished product for use by others. Nevertheless, after 15 years of development, AARON was able to autonomously generate sorts of freehand drawings of humans in garden-like environs.

Generative Art, Profile Pic NFTs, and On-Demand Custom Designs

Generative art NFT collections are often generated by stacking PNG images on top of one another. First, character traits are chosen, with variations of each trait being the same dimensions and relative position within the final graphic. For instance, the desired finished collection might consist of 10,000 Hippie Dog NFTs, and one of the traits happens to be sandals.

generative art nfts
Profile picture NFTs are often generated by combining a set number of layers into a variety of combinations, with some of the combinations being rarer than others.

Designers can decide how many possible variations in sandal colors could be generated, (for instance, with red sandals being the rarest variation, and blue being the most common color). The same can be done for other traits like hair, eye color, sunglasses, t-shirt, and so on. Each layer is given a transparent background so that they stack neatly and coherently within the image.

Many of the most popular NFT projects were not one-off designs made by artists and turned into tokenized versions of themselves. Rather, they were also generated using an algorithmic engine in order to quickly produce a neat collection of 10,000 NFTs for release. For instance, while CryptoPunks all have unique attributes, their creation was not affected by audiences. That is, the creative process did not necessarily take place on-chain.

cryptopunks profile pic nfts
Though CryptoPunks used a generative art script to assemble the collection of NFTs, the art itself was created off-chain, and was only thereafter referenced via smart contracts.

On the other hand, there have been several instances which have taken this concept and applied it with varying levels of success. We’ll discuss two successful examples of generative art platforms as they’ve been applied to NFTs to include buyers as part of the creative process via minting, Autoglyphs and Art Blocks.


Canadian software developers John Watkinson and Matt Hall founded the firm Larva Labs, which has been responsible for NFT drops like CryptoPunks, Autoglyphs, and more recently Meebits.
Autoglyphs’ was an experiment in generative art, having been the first on-chain generative art deployed on Ethereum. Users had the opportunity to mint a total of 512 glyphs before the generator became inactive and no more new glyphs could be minted. Now, the only way to acquire one of these rare generative art NFTs is to purchase it on secondary markets from a willing seller.

Autoglyph generative art NFTs
Autoglyphs were perhaps the first completely self-contained mechanism for not only the creation, but also the ownership of digital artwork.

Autoglyph #376 was the most recent NFT in the collection to be sold, going for 248.16 ETH according to DappRadar. Funny enough, these NFTs were initially able to be minted by anybody after paying the gas fee to do so, 0.2 ETH. Proceeds from the sale of Autoglyphs was contributed to an organization aimed at renewable energy activism. Pairing the sale of NFT collections with tangible organizations or groups which might positively influence their communities is a solid strategy for driving up prices and garnering even more attention in the process.

Art Blocks

Art Blocks is unique as an NFT platform as it doesn’t just showcase and sell generative art, but also hosts the generative code, allowing buyers to become participants in the process. The final output is custom NFT artwork, complete with a number of randomized variables derived from ‘seed code’ making it completely unique. The seed determines the artistic output of the generative code as variables like color, geometry, and other rarity features are derived from its contents. 

Essentially, when collectors mint NFTs using Art Blocks, they create Ethereum tokens which are recorded to the blockchain’s ledger and held at the collector’s wallet address. As with any NFT, the generated token is unique and non-fungible. So minting generative art NFTs with Art Blocks is similar to commissioning custom, on-demand artwork using the power of computation, and the added security of Ethereum’s blockchain to prove the artwork’s originality and authenticity.

Art Blocks has several tiers of NFTs, some of which are rather exclusive, as might be expected in the creative realm. Art Blocks’ Curated Tier is the platform’s top tier for generative art NFTs. Artists selected to participate at the Curated Tier are given the privilege of creating NFT drops which are considered officially part of the Art Blocks collection.

Projects found in the Curated Tier are enviable amongst generative NFT artists for their sophistication as well as the reach of their popularity. While many Curated Tier NFT projects were fairly inexpensive to mint at their release, they’ve since appreciated in value rather significantly. The term ‘starving artists’ may have to eventually be retired to the dustbin of history. 

Chromie Squiggle was the first Curated Tier NFT drop on Art Blocks, made by the anonymous founder, SnowFro. Chromie Squiggles used minted token seeds to determine characteristics like the gradient angle, colors, and number of segments.

chromis squiggle NFT
Currently, the cheapest Chromie Squiggle on OpenSea still sells for nearly 13 ETH at the time of publication.

Any artists already accepted into the Curated Tier may also create generative art NFT drops in the Playground Tier. These projects aren’t considered part of the official Art Blocks collection, but are nevertheless a sought-after corner of their website. Indeed, the Playground Tier provides a space for artists to expand on experimental NFT ideas and launch selected drops for collection. Artists dropping NFT collections at the Playground Tier must have already previously dropped a Curated Tier generative art NFT project.

Finally, Art Blocks has their Factory Tier, offering the lowest and cheapest barriers for entry for newer artists. If artists don’t want to wait for acceptance into the Curated Tier, they may launch a generative art NFT collection in the Factory Tier. Factory Tier NFT projects may be less refined than what’s found in the more formal Curated Tier, but it seems they are leaning into this concept rather than it happening by chance. Whereas Curated and Playground Tier projects hold the line in terms of making generative art into its own emergent NFT discipline, Factory Tier NFT drops exude playfulness, spontaneity, and artistic quirk.

Generative Art NFTs: Taking Pleasure in the Process

Blockchain and NFTs more specifically have opened up the door for artistic expression to be recorded and monetized, and so it’s easy to understand generative art’s rise in popularity as an accompanying creative avenue. NFT drops and secondary marketplaces have quickly become popular venues for collecting and appreciating generative art in its many forms. Though not covered in detail, Tezos-based fxhash is another example of generative art’s ongoing rise in popularity, and the movement of artists online where they find increasingly collaborative and financially rewarding communities.

More than anything, it’s clear that there is to be a new genre of digital art which includes a novel relationship between artists and collectors, one in which participation and patronage is greatly enhanced. Still in its infancy, generative art NFTs have yet to hit the mainstream and thus have yet to reveal their true potential. Only time and our imaginations will limit their manifestations and real-world implementations, and we can’t wait to see what the world is dreaming up.


  1. Boden, M. A., & Edmonds, E. A. (2009). What is generative art? Digital Creativity, 20, 21-46.
  2. CHM Editorial. (2017, 20 Oct.). Into the archives: A look back with “robotic artist” Harold Cohen. Computer History Museum. 
  3. Cohen, H. (1988). How to draw three people in a botanical garden. AAAI.

Related Articles

Learn More



Get news, insights, and more

Sign up for the Supra newsletter for company news, industry insights, and more. You’ll also be the first to know when we come out of stealth mode.


The Tech Behind Supra

Learn how Supra’s breakthroughs in Containerization and Parallel Execution are creating the perfect Super dApp storm.

Week 6

May 29, 2024 | 9 PM HKT

Webinar Speakers
PrivacyTerms of UseWebsite Data Usage & CookiesBug DisclosureBiometric Information Privacy Policy

©2024 Supra | Entropy Foundation (Switzerland: CHE.383.364.961). All Rights Reserved