Outcomes of the workshop produced as part of the Peckham Digital festival on 16 October 2025.
This workshop explored notions of reality and non-reality in analogue and digital photography, while thinking critically about the potential value of AI when developed for specific contexts and tasks, rather than as broad, all-encompassing tools (like ChatGPT or general text-to-image generators).
Modern-day smartphone photography is a useful area to explore this: even a “simple” still image is made with the help of multiple other unseen images, manipulations and AI processes, in order to resemble analogue photographs.
These processes help:
As a result, digital images are increasingly further from reality, constructed through a series of automated decisions that shape what we see.
Photography (digital or analogue) has always contained contested elements of truth and fiction. But digital images today are mediated at every step, from capture to final render, with degrees of complexity that are difficult to demystify.
Byproducts inherent to these processes produce interesting effects in themselves, which are becoming more present in visual languages used in sci-fi and speculative storytelling. 3D point clouds, which often depict verification tools, fragmented memories, surveillance footage, or data access, have been used in recent films and video games like The Creator, Alien: Romulus, and Cyberpunk 2077.
In this workshop, we invited participants to experiment with these ideas using a custom-built digital tool which incorporates MiDaS, a local/offline AI tool that creates 3D depth maps from still images, which can be used to simulate depth of field and point cloud effects. Through this process, we reflected on what it means to "capture" reality, and what creative possibilities open up when we move beyond it.