Introduction: The Controversy Surrounding AI-Generated Photography
The topic of whether AI-generated photography should be recognized as authentic photography has sparked heated discussions among scholars and practitioners in recent years^[1]. There are contrasting viewpoints on this matter. On one side, proponents assert that AI-generated images differ fundamentally from traditional photographs, thus disqualifying them from being considered genuine photography. Conversely, others contend that AI-generated images embody a novel form of photography that expands upon and complements traditional techniques^[2]. In this discourse, we will delve into these divergent perspectives and examine their implications for the essence of photography.
The Essence of Photography: Capturing Light and Time
To initiate our exploration, it is valuable to ponder the essence of photography itself. According to André Bazin, photography possesses a distinctive ontological bond with the world as it records the light reflected by objects^[3]. This process of capturing and preserving light onto a surface generates an image that serves as “an emanation of the referent,” thereby forging a special connection to reality^[4]. Similarly, Roland Barthes argues that photography seizes a singular moment in time, endowing it with the capability to document the world in a distinctive manner^[5].
The Emergence of AI-Generated Photography
The advent of AI-generated photography has raised doubts about the intrinsic relationship between photography and reality. Unlike traditional photographs that directly capture reality, AI-generated images often involve composites or manipulations of existing visuals. From this standpoint, one could argue that AI-generated images do not possess the same connection to the world and, therefore, cannot be deemed genuine photographs^[6].
Will AI Art Supplant Photography?
The question of whether AI-generated art will supersede photography is a multifaceted and contentious one. Some believe that AI-generated art has the potential to revolutionize the art world, while others view it as a passing trend or a threat to the authenticity of conventional art forms.
One argument favoring the replacement of photography by AI-generated art posits that the latter has the capacity to produce novel and distinctive images that elude replication by human photographers. AI algorithms can be trained on vast datasets of images, leveraging this information to create innovative compositions and styles, resulting in an abundance of fresh visual aesthetics.
Conversely, critics argue that AI-generated art lacks the inherent meaning and intentional essence that characterizes traditional art forms. Photography, for instance, is often regarded as a means of capturing the world as it exists, whereas AI-generated art is seen as a means of simulating the world based on existing data. Some contend that AI-generated art is inherently derivative and devoid of the creative spark found in human art.
Another concern centers around questions of authorship and ownership. Who possesses the rights to an AI-generated artwork, and who should be credited as its creator? These questions remain subjects of ongoing debate, with future resolutions yet to be determined.
Are AI-Generated Images Eligible for Copyright?
Determining whether AI-generated images are subject to copyright is a complex matter that lacks a definitive answer. In general, the copyright for an artwork or image rests with its creator. However, in the case of AI-generated images, identifying the creator can be challenging.
One possibility is to consider the person or organization responsible for developing the AI algorithm used to generate the image as the creator and copyright holder. This parallels the way software companies hold copyright for the code they create. Nevertheless, this perspective has been contested by legal scholars who argue that since the AI algorithm cannot exhibit original creative expression, it cannot be regarded as the author of the work.
Another possibility is to attribute the creator status to the person who trained the AI algorithm or provided the training data. This is akin to how a photographer is recognized as the creator of a photograph, even if they use a camera created by someone else.
Ultimately, determining copyright ownership for an AI-generated image will likely depend on the specific circumstances surrounding its creation. As AI technology advances and grows more sophisticated, it is expected that more legal challenges and debates regarding the ownership and copyright of AI-generated works will emerge.
May AI-Photography is a treat to photographers and photojournalists?
The emergence of AI-generated photography has sparked concerns regarding its potential impact on the field of photography and the work of photographers and photojournalists. While AI-generated photography can produce remarkably realistic and precise images, there are still notable distinctions between these creations and those captured by human photographers.
Photography encompasses the ability to seize moments and emotions in a distinctive and genuine manner. Photographers undergo training to anticipate and capture the precise instant that conveys a specific message or story. They make artistic choices involving framing, lighting, and composition. These elements are pivotal in creating images that resonate with viewers and convey a particular perspective or message.
Conversely, AI-generated photography relies on algorithms and data, limiting its capacity to capture emotions or convey nuanced messages due to its programmed nature. While AI-generated images can exhibit high realism, they lack the emotional depth and artistic interpretation inherently present in human photography.
However, it is important to acknowledge that AI-generated photography may impact certain specialized areas of photography, such as product or architectural photography, where the primary objective is to accurately and meticulously capture specific objects or spaces. Additionally, AI-generated images may find utility in fields like marketing or advertising, where the emphasis lies on creating polished and visually captivating visuals.
Overall, while AI-generated photography may exert some influence on the photography industry, it is improbable that it will completely replace the work of human photographers and photojournalists. These professionals bring a unique perspective and creative vision to their craft, an essence that sets their work apart.
AI Tools for Generating Photography
There are a variety of AI tools and techniques that can be used to generate photography, including:
- Generative Adversarial Networks (GANs): GANs are a type of deep learning model that can generate new images by learning from a dataset of existing images. GANs work by pitting two neural networks against each other – a generator network that creates new images, and a discriminator network that tries to distinguish between real and fake images. The generator network gets better at creating realistic images over time as it learns from the feedback of the discriminator network^[7].
- Neural Style Transfer: Neural style transfer is a technique that uses deep learning to apply the visual style of one image to another image. For example, a photograph could be “stylized” to look like a painting or a sketch^[8].
- Image Manipulation: AI tools can also be used for image manipulation, such as removing or adding objects to an existing image, or changing the lighting or color of a photograph. These techniques can be used to “enhance” or “correct” images, or to create entirely new compositions^[9].
The Changing Nature of Photography in the Digital Age
The debate surrounding the authenticity of AI-generated images as “real” photographs may be overshadowing the more significant cultural and social implications tied to these images. Susan Sontag points out that photography has traditionally served as a means to capture and safeguard cultural memories, but with the widespread use of digital images and the emergence of AI-generated visuals, our connection to the past and present could undergo profound transformations^[10]. Looking through this lens, the philosophical nature of photography might take a backseat to the evolving social and cultural roles that images play in the digital era.
Conclusion: The Complexities of Defining “Real” Photography
To wrap up, the matter of whether AI-generated images can be classified as authentic photography is intricate, sparking crucial theoretical and practical inquiries regarding the essence of the medium and its ties to technology, culture, and history. While diverse viewpoints exist on this matter, one thing is evident: AI-generated images are progressively gaining ground in today’s visual culture.
- [1] See, for example: Lev Manovich, “AI Aesthetics,” artnet News, February 28, 2018, http://manovich.net/index.php/projects/ai-aesthetics; Joerg Colberg, “AI and Photography: A Historical Survey,” Conscientious Photography Magazine, May 17, 2019.
- [2] See, for example: Trevor Paglen, “Invisible Images (Your Pictures Are Looking at You),” The New Inquiry, April 10, 2016.; Jason Bailey, “The Art of Artificial Intelligence: Here’s How Creatives Are Using AI Today,” Artnet News, March 12.
- [3] André Bazin, “The Ontology of the Photographic Image,” in What Is Cinema?, vol. 1, ed. Hugh Gray (Berkeley: University of California Press, 1967), 9-16.
- [4] Ibid., 12.
- [5] Roland Barthes, Camera Lucida: Reflections on Photography, trans. Richard Howard (New York: Hill and Wang, 1981), 5-6.
- [6] See, for example: Colberg, “AI and Photography.”
- [7] Ian Goodfellow, “Generative Adversarial Networks,” in Proceedings of the International Conference on Neural Information Processing Systems (NIPS), 2014, 2672-2680.
- [8] Leon A. Gatys, Alexander S. Ecker, and Matthias Bethge, “A Neural Algorithm of Artistic Style,” arXiv preprint arXiv:1508.06576, 2015.
- [9] See, for example: Adobe, “Photoshop Features: Image Manipulation,” accessed April 23, 2023, https://www.zilliondesigns.com/blog/editing-tools-photoshop-image-manipulation/.
- [10] Susan Sontag, On Photography (New York: Picador, 1977), 3-14.
Bibliography
Bazin, André. “The Ontology of the Photographic Image.” In What Is Cinema?, vol. 1, edited by Hugh Gray, 9-16. Berkeley: University of California Press, 1967.
Bailey, Jason. “The Art of Artificial Intelligence: Here’s How Creatives Are Using AI Today.” Artnet News, March 12, 2019.
Barthes, Roland. Camera Lucida: Reflections on Photography. Translated by Richard Howard. New York: Hill and Wang, 1981.
Colberg, Joerg. “AI and Photography: A Historical Survey.” Conscientious Photography Magazine, May 17, 2019.
Goodfellow, Ian. “Generative Adversarial Networks.” In Proceedings of the International Conference on Neural Information Processing Systems (NIPS), 2014, 2672-2680.
Gatys, Leon A., Alexander S. Ecker, and Matthias Bethge. “A Neural Algorithm of Artistic Style.” arXiv preprint arXiv:1508.06576, 2015.
Manovich, Lev. “AI Aesthetics.” artnet News, February 28, 2018.
Paglen, Trevor. “Invisible Images (Your Pictures Are Looking at You).” The New Inquiry, April 10, 2016.
Sontag, Susan. On Photography. New York: Picador, 1977.