AI fabrication raises existential questions about photography's reality

“What the Photograph reproduces to infinity has occurred only once: The Photograph mechanically repeats what could never be repeated existentially…”

French semiotician Roland Barthes’ reflection in his book Camera Lucida encapsulates the essence of photography: It captures a fleeting moment forever fixed in its unique temporal and spatial context, inherently incapable of being replicated. In other words, a photograph is a fixed representation of a reality that can never be experienced in exactly the same way again.

But hold onto your lenses! This very notion of photographic authenticity is getting a serious shake-up with the rise of generative AI, which is flipping our understanding of memory on its head. With its power to fabricate hyper-realistic images, this technology can conjure events that never happened, challenging our very understanding of what is real. Take, for example, Google’s latest Pixel 9 phones. These devices come with an “Add Me” feature that merges two photos into a single, seamless image, blending reality with the magic of AI and augmented reality. Moreover, the “Reimagine” function allows users to alter the very reality of a photograph.  Similarly, almost all modern smartphones, including the recently launched iPhone 16 series, offer capabilities to remove unwanted elements from images and reframe them. In this new world, traditional photography skills (Photoshop prowess) are rapidly becoming obsolete — and the very nature of memories and perceptions is undergoing a radical transformation. Does the authenticity of these AI-generated creations hold significance? The answer leans heavily toward “yes” when these virtual fabrications are accepted and utilised as real, potentially triggering tangible consequences in the real world.

Within the labyrinth of human consciousness, something intriguing happens: False memories are created; they are occasionally stirred by old photographs of cherished events, recollections of which may be marred by inaccuracies. While such false memories are typically personal and relatively benign, AI-induced hallucinations represent a far more troubling development, affecting our grasp of reality itself. Generative AI’s foray into memory manipulation begins with its ability to create highly realistic simulations with convincing narratives — simply, the nuisance and ethical minefield called “deepfakes”. 

At its core, this process leverages algorithms trained on vast datasets to produce content that mimics real-life experiences. The implications of such technology are far-reaching, impacting everything from virtual reality  to psychological research. Recent studies show that spending time in AI-generated virtual environments can lead people to develop memories of events they’ve never actually experienced. Further complicating matters, a 17-year-old study at University of California, Irvine, reveals that digitally altered photos can distort our memories and attitudes toward historical events. Participants who saw doctored images of the Tiananmen Square protest and an anti-war demonstration recalled these events as being more dramatic and violent than they truly were. The implications of AI-induced false memories stretch well beyond academic circles and into the real world. AI-driven technologies are transforming content creation by analysing consumer behaviour and crafting advertisements that resonate on a deeply personal level. This technology subtly shapes consumers’ perceptions and memories of products or brands, leading them to recall past interactions or preferences they may not have genuinely experienced.

Moreover, we find ourselves at the intersection of art and artificiality. The intricate craft of photography, once defined by mastery over lighting and colour, now faces formidable challenges in an era dominated by AI. In a culture enamoured with idealised images, the raw and genuine essence of lived experiences risks being overshadowed by artificial perfection.

In a world where reality is increasingly malleable, are we at risk of becoming isolated within our own personalised versions of reality, losing touch with shared experiences?

On one hand, AI-driven memory alteration holds therapeutic potential, such as helping individuals reframe traumatic experiences. But on the other hand, it poses risks of misuse, such as undermining the integrity of legal systems with AI-generated evidence that could lead to wrongful convictions.

American photographer Alfred Stieglitz once observed: “In photography there is a reality so subtle that it becomes more real than reality.” In an age where AI can fabricate and manipulate memories, the photograph, a synonym of immutable reality, finds itself grappling with an existential question: What remains of the singularity of a moment captured in time when the very essence of memory is in flux?