When Images No Longer Capture Reality
Since the advent of photography and film in the 19th century, humans—who primarily rely on vision—have become accustomed to seeing images as projections of a specific place and moment in reality. As carriers of reality, images capture fragments of time and space, allowing us to perceive the reality they reflect and shape our understanding of the world.
The internet broke the spatial and temporal constraints of images, enabling them to spread more efficiently online. With the rise of social media and multimedia platforms, people have increasingly relied on images to convey and receive information, even replacing pure text as the primary medium of communication. Over time, images have become the dominant means through which we perceive reality.
Because of this, photo editing tools emerged, allowing content creators to modify images to guide audiences toward a desired interpretation of reality. Companies are willing to invest heavily in endorsers and professional teams to craft advertisements that shape a “perceived reality” of their products, enhancing audience appeal. Likewise, news media can manipulate public perception by selectively editing footage, influencing the direction of public opinion.
Although these images may be retouched or cropped, their original materials are still rooted in the physical world, maintaining some connection to reality—albeit perhaps at a deeper level of Plato’s cave. However, the emergence of generative AI has changed everything. Now, with just a few prompts, we can create images that have never existed in the real world, completely overturning humanity’s long-held understanding and reliance on images.
Generative images introduce a rift between perception and reality. They no longer serve as representations of the real world but instead become mere simulations of concepts, styles, or emotions. In other words, images are no longer tethered to the physical world; they have become purely symbolic carriers of information, entirely detached from reality. As a result, their nature increasingly resembles storytelling and artistic creation, serving as a medium for expressing ideas rather than mirroring the real world.
When images are no longer evidence of reality—when they no longer reflect the truth—what happens if those who were supposed to “convey reality” start replacing traditional photography with generative imagery? If audiences discover that an advertisement was AI-generated rather than filmed in real life, can they still establish a meaningful perception of the product? Likely not, because once people recognize that an image is merely an illusion, they struggle to trust or connect with the information it conveys.
Even if the creators of generative images intend to convey conceptual information, their detachment from reality significantly weakens their persuasiveness. This is not due to the "unreality" of generative images themselves, but rather because of the way humans process information. We are accustomed to perceiving the three-dimensional reality around us—or, when we know an image is an artistic creation, we focus on its intended message. However, if fictional images are forcibly labeled as reality, they risk falling into an awkward middle ground: neither gaining the audience’s trust nor securing their rightful place within the framework of artistic creation.