AI & Photography: A Personal Story
Why I Chose to Work with AI
My journey into the world of AI was born out of necessity and curiosity. As an artist, I found myself unable to work directly with patients suffering from Chronic Actinic Dermatitis (CAD). The graphic photographs of bleeding skin and severe eczema were too intense for many audiences, limiting the reach and impact of my work.
AI offered a solution. It allowed me to explore new tools and ideas, transforming the raw and often disturbing visuals into something more palatable yet equally powerful. More than just a tool, AI became a partner in my creative process. This collaborative creation, where I input my creative direction and the AI provides its computational interpretation, often resulted in unexpected and inspiring outcomes. It reminded me of Garry Kasparov's Centaur tasks, where human and machine work together to achieve superior results.
Small Data Pool vs Large Data Pool
My recent experience at a conference on AI & Photography, highlighted just how many people are frightened and confused by AI. Not everyone understands that AI is essentially a statistics machine, extrapolating its output from the data it has been exposed to.
For instance, if I enter a prompt into an AI image generator that includes the name of a photographer working in a niche area, the resulting output could be visually similar to the work of that photographer due to the small data pool.
However, when I was making work involving data-rich and non-niche areas, like the portraits of middle-aged Caucasian males for my CAD project, the results are far less likely to replicate the work of any given photographer due to the wealth of available information — this concept evaded the head of a major photography organisation who accused me of plagiarism while I was presenting. Ho-hum…
The possible financial impact, with photographers loosing out on work being generated by AI, was also a huge cause for concern, with many expecting to be replaced by AI in the near future.
Historical Concerns With Technology
The advent of photography stirred up similar concerns among painters. French painter Paul Delaroche, upon seeing the first daguerreotype, famously declared, “As from today, painting is dead!” This statement reflected a fear among some artists that photography would render traditional painting obsolete. Similarly, the realism of Gustave Courbet’s paintings was criticised for being “photographic,” a term used derogatorily to imply a lack of artistic interpretation.
Copyright and the Tradition of Copying
When it comes to copyright, we need to acknowledge the elephant in the room: artists have always copied. At times, it's even been encouraged and accoladed, for example:
Édouard Manet: His work "Déjeuner sur l'herbe", one of the grandest and most copied works of western art history, is based in large part on an Italian Renaissance print.
Marcel Duchamp: Known for his concept of the "readymade", Duchamp's "Fresh Widow" from 1920 is a perfect example. Sturtevant, another artist, created seven replicas of "Fresh Widow", all of which looked exactly like the 'original' Duchamp, which is itself a readymade¹.
Sturtevant: Sturtevant recreated her colleagues’ paintings and sculptures with the same techniques they used, and not via photographic or digital means. She never asked permission, though Warhol once lent her his flower silkscreen. Her works included those of Warhol, Lichtenstein, Haring and more.
Sherrie Levine: An American photographer, critic, and conceptual artist. Some of her work consists of exact photographic reproductions of the work of other photographers such as Walker Evans, Eliot Porter and Edward Weston.
Richard Prince: An American painter and photographer who rephotographed advertisements and presented them as his own art. His work, while controversial, has been influential in the art world.
These artists show that the act of copying can be a form of commentary or critique, and can lead to new, original works that push the boundaries of what we consider art. It's a reminder that creativity often involves building upon the work of others, transforming existing ideas into something new and unique.
Could creator data pools become a thing, like a Pablo Picasso dataset? Combined with blockchain technology, could the usage could then result in micro (or macro) payments everywhere the image is used or shown?
Environmental Concerns
There are legitimate environmental concerns about the energy consumption of AI. However, according to this research paper, it's worth noting that AI can take less time and energy than a human to produce similar results. [Thanks to Sarah-Jane Field and Chris Le Messurier for this].
Moving Forward
As artists, we must be open about our AI usage and navigate its use responsibly, ensuring that the AI's training data is ethically sourced. We should encourage and embrace discourse around topics such as:
Authenticity and Originality: AI may challenge the notion of what constitutes original artwork, as it can replicate styles and elements from a vast database of existing art. Should rap and hiphop music not be valued as an art form because of it’s reliance on sampling from other musicians. Could AI art be seen in a similar way?
Artistic Skill Development: Is there a concern that reliance on AI could lead to a decline in traditional artistic skills and techniques?
Intellectual Property Rights: AI-generated art raises complex copyright issues, particularly when it comes to determining the ownership and authorship of the created works.
Bias and Representation: AI systems can perpetuate biases present in their training data, leading to a lack of diversity or misrepresentation in generated art.
Let's also explore alternative financial models for artists, such as the blockchain payments mentioned earlier. In the ever-evolving narrative of photography, AI is anticipated to contribute a chapter rich with importance and innovation to the medium.