Reading Levin’s essay really changed how I think about what it means for a computer to “see.” One of the most striking points was that, unlike humans, a computer has no built-in way to understand what it is looking at. Levin writes that video is “just a big, dumb array of numbers”, which really stood out to me, people automatically recognize faces, objects, and events, but a computer has to be taught step by step what to look for. This helped me appreciate how much of computer vision is really about extracting meaning from raw data rather than simply displaying images.
I also found it interesting that helping the computer “see” is as much about the physical environment as it is about the code. Levin mentions that careful control of lighting and contrast, even using infrared light or retroreflective materials, can make the difference between a system that works and one that fails. That made me realize that when we design with computer vision, we are not just writing software, but we are shaping the entire stage so the camera has the best chance to succeed.
Finally, the discussion about surveillance made me think about the double-edged nature of computer vision in art. Works like Standards and Double Standards or Sorting Daemon use tracking not just for interaction but to make us feel observed, even judged. I then reflected on how computer vision can be playful but also unsettling and how artists can deliberately use that feeling of being watched to make a political or emotional point. It reminded me that computer vision is not neutral, it carries cultural and ethical weight, which is part of what makes it powerful in interactive art.