Return of the poor image

Curator’s choice

In her essay on the “poor image,” Hito Steyerl argues that a class system hierarchy of images exists in visual culture, where high-resolution “rich” images exist in opposition to low-resolution and degraded “poor” images. “The poor image is a copy in motion,” she writes, “...distributed for free, squeezed through slow digital connections, compressed, reproduced, ripped, remixed…” But where does the poor image exist in art NFTs? Is it possible for an art NFT to be classed as such?

An art NFT is not a copy; its originality—its one-off nature— is not distributed for free, and is not crammed through slow connections. It might be remixed (through generate-based networks to create multiples of similar images) or ripped (Ctrl-Shift-V), but an NFT’s inherent nature requires it to exist within blockchain-based tokens that each represent a unique asset. NFTs in general are not poor, lossy or continual copies; their non-fungibility does not allow for exchange. They are an irreversible digital certificate of ownership, and their authenticity and lineage is always transparent.

The poor image “transforms quality into accessibility, exhibition value into cult value, films into clips, contemplation into distraction.” An art NFT does much the same, translating the quality of a physical commodity into accessible (sometimes) cheaper commodities. They turn exhibition value into cult status with PFPs and a sense of higher social status in being seen to own something of speculative high value; they are sometimes GIFs, clips of a larger work or a meme remixed and redubbed; and in some circles, they have become a distraction from an original intent of art NFTs to provide digital artists with a new market to sell work, receive royalties, and disrupt the traditional art market, creating new communities of collectors and giving artists more access to their fans and buyers. (This is, of course, only referring to a specific type of art NFT, and not a universality that all art NFTs share; some are more based on functionality and many art NFTs are not related to art at all.)

Where does the poor image debate leave GAN creations, text-to-image, NLPs? Can we bracket these with the poor image because they are often more low-res, generated, and distributed? Is the clarity of an image becoming less important when it is surpassed by computational algorithmic output? The “poor” in the poor image is no longer necessarily related to its resolution.

If we consider how contemporary culture has adopted the memeified, low-resolution, low culture image as cult, then we can say that the high-resolution, slickly rendered, glossy image is in opposition. Left to those in the more commercial, more streamlined 9-5 of product-to-consumer turnarounds, is the high-res image a sell-out of surface-level engagement with a subject, the basic bitch of non-conceptual output?

Outside of the art world, some millennials are celebrating the return to “poor media” of the 1990s and 2000s—disposable Kodaks, letter writing, digital detoxes, and life-hacking (the newer make-do-and-mend)—feeling nostalgic for a time they never encountered. The irony is that the fastidious proliferation of these trends becomes widespread by the very tool they’re shunning, Web3. The nature of this nostalgic plight being correlated to the impossibility “to grow up in today's economy” feels somewhat coupled to the rise in emerging artists experimenting with NFTs to regain agency in how they create a sustainable living from their work.

This selection of NFTs elevates works that defy more commercial trends in NFT art (of enhanced realism, slick renders, and imitation) to foreground the necessity of including low-res, computationally generative, and non-fixed format in the mix.

1 Steyerl, H. (2009). In Defense of the Poor Image - Journal #10 November 2009 - e-flux. [online] Available at:

2 Thanks to Dane Sutherland for pointing this out <3

3 Hoffower, H. (2022). Gen Z brought the ’90s back because it feels impossible to grow up in today’s economy. [online] Business Insider. Available at:

Lorna Mills

Singspiel Faster in the collection of FOMOBOY
Singspiel Faster in the collection of FOMOBOY by Lorna Mills

Working in internet-based digital art since the 1990s, Lorna Mills has developed a technique of pixel ripping and deconstruction to create her GIF works. Often with cryptic titles and hard-coded low resolution output, the works recontextualize the origin of the content and meaning and are instead imbued with feelings of anxiety, vulgarity, endless motion, and stepping into the realm of niche subculture. The compositions of jiggling eggs, or sleeping puppies and various obscenities, rendered as an intense, never-ending loop in gloriously low resolution, reflect the rapid consumption of online cultures and memeified images, but also the strange system in which they move.

Rafaël Rozendaal

Infinite Entropy
Infinite Entropy by Rafaël Rozendaal

Infinite Entropy explores Rozendaal’s predilection for abstraction and geometry in hypnotically colorful ensembles. Although the work seems fuzzy or zoomed into, it is actually the opposite: “With NFTs, a lot of people ask, ‘what is the perfect home or display?’ But I always wanted to make works where there is no perfect display, the displays evolve and the work evolves with it.” The overall effect of these works contrasts with Rozendaal’s usually harder-edged outcomes, though the process of building it from simple shapes and small steps is the same. The circles look as if they are moving toward the viewer, implying depth where there is none and thus creating a sense of space within an overcrowded screen.

Jan Robert Leegte

JPEG #0 by Jan Robert Leegte

“JPEG is a collection of generative images in the JPEG file format. The work tries to break with the idea of the fixed file format, therefore creating the JPEG entirely from code using no image material whatsoever. The individual JPEGs are generated responsively at the moment of viewing based on the token hash as a random seed. JPEG pays tribute to interface culture, and as such when the work is resized it is again rendered and compressed. When viewed smaller, details will disappear, yet the composition stays recognisable. Delving into the depths and peculiarities of the digital, tethered to art history, JPEG found its way to the surface. Generated color fields are utilized to trigger the compressing algorithm, resulting in a world of raw visual entities. Where abstract expressionists cut out the depiction in painting, aiming to let the artist convey pure emotion, JPEG cuts out the artist's emotion and lets it emerge from within the algorithm. Carefully, you could categorize the work as compressionism.”

Enjoying the selection?

Subscribe to get new Curator’s Choices right to your inbox.

Simon Denny, Guile Twardowski, and Cosmographia (1999-2001) reimagined by Cosmographia, with Simon Denny and Guile Twardowski (1999-2001) reimagined by Cosmographia, with Simon Denny and Guile Twardowski by Simon Denny, Guile Twardowski, and Cosmographia

“Dotcom Séance is an expansive digital installation exploring technology’s sustained influence on global cultural and political development. For this project, Denny has selected 21 dotcom-era companies declared ‘dead’ in the wake of the dotcom crash in the early 2000s, with the purpose of ‘resurrecting’ their existence as it stands in a contemporary setting. Each company was given a number of new logos produced by text-to-image Artificial Intelligence, created in collaboration with Cosmographia, thus demonstrating the relationship between Web1 and Web3 through a visual manifestation of these companies’ ‘ghosts’ and respective histories. These new logos entitle their holders to register an ENS subdomain and serve as ‘employees’ of the reanimated company.”

Casey Kauffmann

#cybaer by Casey Kauffmann

@uncannysfvalley advocates for a metric of valuation which aligns with Hito Steryel’s conception of an image's quality being measured by its popularity, subverting classicist notions of high resolution and elitist tech-fetishization as an indication of value. “The project features digital collage works and GIFs created using only my iPhone. The pieces I post to this account are an ever-accumulating collection of material from all corners of the internet, sourced from Tumblr, Instagram, and Google. The collages serve as an intervention into the online representation of women. Using depictions of women often authored by men, I supplant their original meaning, recontextualizing them within a framework of my own imagination. I employ sparkly, pink, feminine-coded aesthetics, which commonly inhabit a sexist association with cultural constructs of superficiality.”

Peter Burr


“Every one of the structures in BOOM TOWN represents a unique, living artwork comprised of 1,000 images that will change over time, evolving as a kind of durational stop motion animation over an undetermined period pegged to Ethereum’s block time—that is, the length of time it takes to add another block to the blockchain. Collectors receive the full set of images as part of their acquisition package and can browse or even create prints of the work’s past or future state, but they cannot make the work advance any faster or slower than block time.” In the process of creating the works, Burr highlights unique “fingerprints” of different software, working across multiple platforms to weave together a new aesthetic that reveals the compression and dithering of this approach.

Harm van den Dorpel

Temporality is – 2/100
Temporality is – 2/100 by Harm van den Dorpel

“Temporality is an artwork in the form of screensaver that uses Gaussian smoothing, a function used in noise reduction for digital images, and a frozen loading spinner, to adorn live uploaded images scraped from photo sharing sites. The images are shown for a random number of seconds before disappearing. The blurring effect and static circle emulate a frozen moment in time, eliciting a sense of disruption in the viewer waiting for the theatricality that has become standard in UX culture. The work renders the mundane as a digital artifact without opulence but with an almost unnerving sense of pause in time that requires a still appreciation of a moment where nothing happens.” (@terencesharpe)

Libby Heaney

Venuses (quantum bodies watched by Open Pose algorithm)

Libby Heaney’s Venuses (quantum bodies watched by Open Pose algorithm) are some of the first artworks generated using quantum algorithms. Heaney uses quantum data as a material basis for pluralizing and entangling the human body in a series of venus poses, some taken from Western art history and others re-performed by the artist. The presence of both types of images highlights how our bodies are shaped by culture as we re-perform the poses we have been culturally exposed to. The initial stills emphasize biases in the way bodies are seen in AI and art history. The subsequent frames in the animation were generated by passing the initial frame through quantum computing systems, which, through entangled pixels, fragments and inverts the image. As the animation progresses, we see the body from alternative, multiple perspectives—boundary-less and form-less, disrupting the biased conceptions of the female nude.

Curate your own selection

Sign in to create Curator’s Choices.
The Curated platform supports openness in the Web3 space. Most of the works featured in this Curator’s Choice section are fetched from external Web3 marketplaces and fall under their respective regulations, and remain the intellectual copyright of the artists. The editorials are non-commercial and we do not take any fees or commissions.