Geoff Cox’s Ways of Machine Seeing provides a compelling framework for understanding the role of machine vision in shaping visual (or virtual) environments. Cox examines how machine vision operates through calculations that prioritise operational efficiency and data accuracy while remaining indifferent to symbolic or cultural meaning. This detachment highlights the limitations of technologies that process images such as photogrammetry and VR, which excel in spatial accuracy but often neglect embodied, phenomenological, sensory-rich experiences.
This image was generated using DALL-E by inputting the prompt "make an image of a digital camera shooting a glacier and the camera saying "I don't give a fuck that you are melting" OpenAI (2024) DALL-E: AI image generation tool. Available at: https://openai.com/dall-e (Accessed: 18 November 2024)
Cox also discusses how machine vision generates “operational images” that do not merely represent reality but actively shape it through algorithmic processes, which makes sense through the digital methods of capturing and compressing of visual space. This underscores the idea that digital photographs are not neutral tools but agents that construct specific worldviews based on their technical and algorithmic limitations. When these images are used for photogrammetry then those limitations compound into new forms of virtual representations that contain, or are a result of these non-real features.
Comments