At last, an innovation has emerged that fundamentally redefines the very concept of photography. Here, we introduce the groundbreaking camera 'Paragraphica,' developed by Dutch designer Bjørn Karmann.
A Camera That Captures Data, Not Light
Paragraphica is a camera without a lens to collect light. Instead, it gathers contextual data such as geolocation, weather, and time, and uses this information as input for the AI model Stable Diffusion to generate images. In other words, rather than directly capturing the physical landscape, it creates an 'image' of the place through the interpretation of data and artificial intelligence.
©︎ Bjørn Karmann
On the rear monitor of Paragraphica, text is displayed; when the shutter is pressed, the AI generates an image based on this prompt.
A Paradigm Shift in the Concept of Photography
The revolutionary aspect of 'Paragraphica' lies in transforming the camera from an output device for information into an input device. Traditional cameras were tools to capture and record reality as it is, but 'Paragraphica' collects data about a location and, through AI interpretation and reconstruction, generates new images from nothing—creating something from zero to one. This process fundamentally questions the purpose of photography and suggests that the evolution of camera mechanisms is shifting from documentation to creation.
©︎ Bjørn Karmann
(Left) is a photograph taken with a conventional camera, (center) is the prompt at the time of shooting, and (right) is the 'photograph' generated by Paragraphica.
The system is connected to the internet and collects real-time data such as geolocation, time, and weather at the shooting location. The gathered information is converted by the camera into a prompt, and based on this prompt, generative AI creates the image.
A Sensibility That Imposes Physical Constraints at the Moment of Capture
A key point reflecting the designer's sensibility is that image generation by AI is shaped by the imposition of 'physicality.'
While general AI image generation tools can create arbitrary images from vast datasets, the images generated by 'Paragraphica' are based on data from a specific location captured by the camera. In other words, Paragraphica generates images that reflect the characteristics and atmosphere of a place with higher fidelity, grounded in precise information. This subtle attention to detail becomes the key to offering a photographic experience that fuses the uniqueness of the location with the photographer's personal encounter.
©︎ Bjørn Karmann
Screenshot of visual design and dynamic data in Noodl. To generate prompts and images based on geolocation, the camera establishes communication with multiple APIs.
AI-Generated Imagery and Human Perception
How does AI 'see'—or more precisely, 'generate'—the world? Considering the differences between AI-generated photographs and human perception, the very notion of photography may be fundamentally transformed. 'Paragraphica' emulates non-human perception, and the resulting images differ from the visual representations of reality traditionally offered by photography. At the same time, if we begin to perceive these emulated images as authentic, it will become increasingly difficult to claim that photography is a mere mirror of reality. If experiencing the world through the lens of AI influences our own cognition, then photography may open up new possibilities as a tool for interpretation and expression, rather than simply a medium for documentation.
As illustrated in the following example, there is already considerable debate regarding the regulatory frameworks for photographs (or photograph-like images) involving AI.
Will photographs taken with 'Paragraphica' redefine the concept of photography, or will they fade into ambiguity, neither photographs nor generated images? The future evolution of this technology warrants close attention.
