Delete By Default

Jon Uriarte is curator of the Digital Programme at The Photographers’ Gallery.
More on Jon Uriarte

This essay steams from the research and thinking for the 2022 edition of Getxophoto Festival themed “To Imagine” and from accompanying Marloes de Valk during her research on the environmental impact of networked images in a collaborative PhD programme by The Photographers’ Gallery and the Centre for the Study of the Networked Image. The text is available to download as a pdf here, and translated into Japanese and Portuguese.

“Pensar es olvidar diferencias, es generalizar, abstraer”
– Funes el memorioso, Jorge Luis Borges
“A culture that is merely preserved is no culture at all”
– Capitalist Realism: Is There No Alternative?, Mark Fisher

Living and resisting through the effects of late of capitalism requires critical imagination. In times of climate emergency, back-to-back economic crashes, wars and hardship, the philosopher Marina Garcés asks how can we separate their real implications from the stories that stem from them. She warns that when facts automatically become unavoidable predictions there’s only room for compliance or denial, for utopias or dystopias. Garcés makes the case for the need of an ecology of imagination. A critical and emancipatory process consisting of three aspects: first to incorporate otherness in order to become strange to oneself, then to stop living against the past and the future to live in time, and finally to learn not to know. The first of the dimensions doesn’t only suggest being hospitable to the other but also to be strange of oneself by opening the possibility of a common world that we can recognise as shared, not of our own. The second dimension implies to learn in a communal time where, instead of chasing past notions of the future, we can imagine in time. The third proposal suggests a threshold where knowledge and ignorance meet without clashing. A dimension in which our saturated attention can learn not to know and therefore make questions that might become “a starting point of a critical and imaginative commitment to emancipation in times of the scheduled Apocalypse.”

A portrait of Russell A. Kirsch’s son Walden, regarded as the first digital image.

Photography contributes to the anxious reality of today in which alternatives seem to be nowhere to find. Even if many of its core elements –its materiality, circulation, value, and its human and non-human uses– have radically changed since its digitalisation and networking, those changes have been mostly driven by the extractive and accumulative goals that have evolved by the hand of the late capitalist system. The origin of the first digital image offers a good example. In 1957 almost all digital input data required manual processing. A human being was needed to visually inspect the characters on punched cards or magnetic tapes and describe them to a machine so it could process them.

“The problem is one of making directly available to a computer pictorial information which would ordinarily be visually processed by human beings before being fed to a data processing system.”

Engineers and computer scientists like Russell A. Kirsch –who also invented the pixel­– and his colleagues explored how machines could see in order to “reduce the amount of human intervention required during the input process.” A photograph of Kirsch’s new-born child was publicly presented as one of the first pictures ever scanned becoming a poetic metaphor of the new-born technology on mass media, and a landmark in the history of photography. However, that particular image was not published within the scientific paper of their research. A decision that hints at the conflicts between the representational pervasiveness and the underlaying complexities of its materiality, pinpointing them on the announcement of the very first digital image ever made. The wide and multifaceted implications that the development of machinic visual literacy would end up having were already shadowed by the charming story of the baby and the humanist drive of scientific curiosity. Kirsch and his team inaugurated the field of image processing choosing photographs as the starting subject to look at without considering the codification that photography itself already adds to the act of seeing. They used photography as the transparent medium that traditional photojournalism claimed it was. However, their intent was not to show the world as it is or to give a voice to the invisible as some of the journalistic patronizing clichés would put it. Image processing was first developed to make a computer visually literate enough to remove humans from performing certain activities in order to streamline and increase its computational capabilities.

Along with digital technology, the internet is the other main force that has transformed the way in which we live, work and relate to each other through images. Since its very early ages as a military funded research project to its further development to share scientific experiments from the European Organisation for Nuclear Research (CERN), the internet’s purpose has always been to keep the circulation of information going. The first image uploaded to the internet –that of the music group Les Horribles Cernettes– aligns very well with that aim. Showcasing the technical skills of the photographer to capture a relevant moment in time was clearly not the main purpose of that image. It doesn’t document anything rather than perhaps one of the early examples of digital folklore aesthetics. The relevance of the first image on the internet relays on its potential to communicate, to relate to each other, to provoke an emotional response from a viewer who in 1992 might have felt appealed to attend one of the group’s concerts. And that response relayed on its circulation and visibility, a feature brutally exploited years later by social media companies.

Picture of Les Horribles Cernettes, from left to right: Angela Higney, Michele de Gennaro, Colette Marx-Neilsen and Lynn Veronneau.

The social media era, which appears to start its decline these days, can be considered the second successful attempt to monetize the internet after the .com era at the end of the 1990s. In between, other more short-lived but equally frantic developments have also succeeded in creating mechanisms to accumulate wealth on the web, as for instance the recent crypto frenzy and its efforts to transform images into scarce commodities. The role that networked images play in our lives today can’t be denied given that the vast majority of internet traffic is visual. Social media platforms encourage the use of photographs and videos as tools to present oneself and interact with each other, and less as tools for memory. Apps for camera phones that allow instant photo sharing have greatly benefited from current visual practises in which photography is used more as a social communication tool than for documenting, remembering, or artistic expression. While creating and sharing curated images of yourself in front of others has been arguably the main drive on photography-led social networks, that seems to be changing in the most recent platforms. Criticism towards the companies who greedily exploited those carefully crafted online identities seem to have taken hold, favouring an apparently more casual and temporary use of the image. The imposed moral constrains, the controversial sudden changes of its features to copycat competitors and the backlash against the shameless monetization strategies have made some of the biggest platforms struggle. On the other hand, companies favouring algorithmically accelerated visual media rabbit holes are gaining more traction. The openly published media is less directed to chosen family and friends but to everyone aiming only for the widest reach. Meanwhile, the personal interaction is mostly developed through group or individual private chats in which the image remains also important, particularly within the younger generations. The function of photography as a representational tool seems to be shifting from identity shaper –and previously memory keeper– to communication facilitator. The medium has always included those three roles within itself, is the power balance between them what has been rebalancing over time.

Leaning towards a means of communication has multiplied its use exponentially in a time in which the development and massive adoption of consumer memory devices has made digital storage extremely cheaper. At the same time, the poisonous offers by big tech companies to store your data on their servers –the so-called cloud– for free or for very little money makes digital storage look like a bargain. The easy and immediate possibility to produce, share and store images with a phone –a device that most people carry with them constantly– has produced such a gargantuan scale of images that humans have deflected on automated systems to make sense of them. The out of scale constant growth becomes a serious environmental concern as it requires further infrastructures and energy sources. Recent scientific research shows how redundant images lingering in the cloud produce increasing CO2 emissions. Resolution and live stream bandwidth have been constantly augmenting for years along with storage capacity. The mantra of an ever-increasing image and screen size, processing power and storage space seems to be unlimited for the marketers. However, the impact of those ways of mass-producing technology is starting to be questioned and new ideas and environmentally conscious computing practices are arising beyond the margins of consumer culture. The resources of the planet are reaching a tipping point and, even if reluctantly in some cases, the acknowledgment of that new scenario is forcing a call to action.

At this point, going back to Marina Garcés’ ecology of imagination might be useful as a way to critically think beyond the archetypical problem/solution binary. I’m not aiming for a silver bullet that will save us all from the impact of extractive logics. I would like to consider an alternative visual culture and image materiality that could offer at least the possibility to imagine a different kind of photography; maybe a less environmentally harmful one. In order to do so, it would be important to reflect on the ways in which society has already shifted its popular visual practises in response to corporate exploitation, particularly when thinking on the role of the networked image. Reconsidering photography as a means to primarily communicate instead of to record and preserve, as the traditional cannon still insists on, could open different viewpoints to explore.

The origin of the “Feeling Cute Might Delete Later” meme.

When thinking on how to apply Garcés' principles in order to critically reimagine  photography, the first step –how to incorporate otherness to become strange to oneself– is maybe the most important. Assuming that the current default role of digital photography –when created by humans–is to communicate information, ideas and messages instead of documenting reality or keeping memory, we could liberate the medium from its duty to endure. It could start to leave behind its main modernist function as a questionable tool to categorise and describe the world while embracing a communal function, along with other means of communication. Inspired by speech, photographs could aim to be created and shared but not saved forever. A technical development could delete them by default allowing a new way of resistance against the pervasive storing of digital data.

As professor Viktor Mayer-Schönberger puts it on his book Delete: The Virtue of Forgetting in the Digital Age:

“Comprehensive digital remembering denies us humans the chance to evolve, develop, and learn, leaving us helplessly oscillating between two equally troubling options: a permanent past and an ignorant present“

On his book Mayer-Schönberger examines the impact that the systematic storage and monetization of data has had in privacy violations, reputational damage, and outdated or false information. He argues that the digital era has shifted the default from forgetting to remembering a vast amount of data, which ultimately undermines human capacity for abstract thinking, knowledge exchange and development. In response, he offers a technical solution: an expiration time that all digital files would require to be inserted when saved, forcing the operator to reflect and decide on the file’s purpose and its preferred lifespan. Once the prescribed time has passed, the file would be automatically deleted saving storage space, liberating memory and learning “not to know”, as per Garcés’ suggestion.

Mayer-Schönberger’s book was published in 2009, the same year in which Facebook introduced new privacy settings making all status updates public unless specified otherwise. At that time private data was shared online with little concern by the growing tech companies while civil liberty campaigners and other organisations alerted about its predictable problems –it would take five more years for the EU parliament to approve the right to be forgotten. Even if they were right to be concerned about the impact of private data management –just Facebook has been implicated in electoral influence campaigns in countries like Argentina, Kenya, Malaysia, the United Kingdom, the United States and South Africa– today we know that not all data remains stored or accessible. Storage and accessibility are mostly controlled by corporations who decide about deleting, updating or cutting access depending of its profitability without acknowledging data’s potential cultural value. In response, some public institutions and independent organisations such as the Internet Archive make an invaluable work retrieving and conserving discarded fragments of the internet that they believe are worth keeping. However, even considering all the data that is carelessly wiped out, the quantity that is constantly generated is still so immense that its easily outgrown. Given the relentless increase of data volume, some preservation organisations have changed their initial strategy from archiving all to archiving a curated selection. A very important decision that still doesn’t have an impact on the production of data itself.

Since his book was published, several developments have made Mayer-Schönberger’s proposal to consider an expiration date for digital files more appealing. The book didn’t anticipate the environmental impact that the digital age could have and how deleting could help reduce it. It included analysis on the role of images within the larger framework of forgetting but it didn’t discuss the more abstract conflicts that his proposal would pose to the relationship that photography has with memory. The publication preceded Snapchat –the app that made fleeting photographs popular – by two years. On the first post on the company’s blog, one of the co-founders expressed that “Snapchat isn’t about capturing the traditional Kodak moment. It’s about communicating with the full range of human emotion — not just what appears to be pretty or perfect.” The idea of vanishing photographs and short videos has been adopted by other companies in the following years, although most of them don’t delete them from their servers not even after a one-month period, as Snapchat claims to do. The data is too precious as raw resource to keep the attention economy going and the machine vision algorithms training. Integrating the time expiration function within the metadata of the file itself would help to avoid companies to quietly store the images when claiming the opposite. Setting digital photographs to be deleted by default could offer a mechanism to resist surveillance, to reinforce privacy and to evade monetization from everyday social exchange. It could help limit the nowadays rampant data scrapping for the development of cognitive computing, the so-called artificial intelligence –a research field in the process of becoming a mass industry. The gigantic collections of user-generated images are the essential material to create tools that not only recognise and categorize photographs but also synthetize new images out of them. A technology only capable of imagining past futures yet incapable of imagining “in time”, as Marina Garcés suggests.

Promotional banner by Snapchat announcing it’s delete by default functionality

The requirement to set an expiration time when saving an image would not only help consider the impact of its materiality, it would also give certain agency to the person in charge of saving it. Mayer-Schönberger suggests other functionalities that would make this idea more flexible. The expiration date could be updated if the relevance of the image changes over time and alarms could be set to give the opportunity to reconsider its imminent deletion. Given some of the fundamental features of digital media and the unprovable adoption of this feature by all file formats –especially those created and controlled by private interests– it’s safe to assume that the role of visual culture as a tool for memory or representation would still be available, maybe just not predominant.

Setting digital photographs to be deleted by default would not only require a very complex technical endeavour, it would most importantly ask to forget and rethink some of the traditional photographic cultures. It would challenge what it means to preserve a medium, proposing that it is not the perpetuation of certain normative historical practices but the acknowledgment and critical investigation of its changing nature. It would discard endurance as its main trait rebalancing it towards circulation, effect and interaction. It would force to reconsider the market value of digital visual art in relationship with its lifetime, challenging some of the recent technical attempts to perpetuate its existence and further commercialize it. It would ask to consider photography as a live performing medium, leaving behind -at last! - the associations with death that have constantly accompanied the medium over time. A rather outdated analogy in a time in which photography is anything but immobile or an emanation of the referent –pointing Barthes–. Furthermore, photography is far from being dead itself, against what is being claimed every time that a new technical development is announced in the field. It carries many deceased practices and notions within itself and it can often be more difficult to define –as it exists blended with other technologies– but fuelled by the digital liberalization of its means of production ­–I deliberately wouldn’t call it democratization– it has been more vivid on its last twenty years than in most of its previous history.

Photography is a living artefact in continuous transformation, hardly controllable as it’s reproducing an amalgam of itself and its conditions of production along with their surroundings. It might remind us more of other kinds of entities capable of going viral and replicating themselves only within a network or an organism. We could consider it similar to the virus, one of the smallest and most numerous types of biological entities. An agent that challenges the notion of what living means, even if it is capable of performing two of the functions attributed to living beings: relating to each other and reproducing. A referent popularly linked only to diseases while its role as fundamental element for the evolution of life remains unknown for many. A very contradictory and slippery thing of its time that is hard to even agree on a term for it: photography, expanded photography, post-photography, image, networked image, computational image... Incorporating delete by default would force the medium to review its status as knowledge keeper across different fields and mediators, from the everyday practitioner to the cultural organisations and the companies that make profit out of it. Photography would need to sustain its own ecology of imagination to explore other possibilities that would question the basis of its own raison d'être.


  • Borges, Jorge Luis (1944). Funes el memorioso [Funes the memorious], Ficciones [Fictions]. In Fervor de Buenos Aires, Obras completas [Complete works] (p. 485). Emecé. 
  • Dewdney, Andrew (2021) Forget Photography, Goldsmiths Press.
  • Fisher, Mark (2009) Capitalist Realism: Is There No Alternative?, Zero Books.
  • Jurgenson, Nathan, (2019). The social photo: on photography and social media, Verso Books
  • Mayer-Schönberger, Viktor (2009) Delete: The Virtue of Forgetting in the Digital Age, Princeton University Press.


  • Barthes, Roland (1982) Camera Lucida: Reflections On Photography, Hill and Wang.
Jon Uriarte is curator of the Digital Programme at The Photographers’ Gallery.
More on Jon Uriarte

Suggested Citation:

Uriarte, J. (2023) 'Delete By Default', The Photographers’ Gallery: Unthinking Photography. Available at:
< Prev Next >