In my country, privacy is a legal concept and easily trumped by commercial interests. One example is the use of AI to collect and personalize data in physical stores.
This involves using AI to process high-resolution security camera footage installed in stores and link it to customer and purchase information. Thanks to strong lobbying by major domestic companies, this has recently become legally possible with just a notice posted in the store, and there is no way to opt out.
The following information is stored, processed, and inferred: face, height, gait, bone structure, gender, fashion, gaze, products viewed, preferences and thoughts inferred from products viewed, purchased and fashion, as well as number of visits to the store.
For those of us who value privacy, a surefire strategy if you can’t opt out is to simply avoid the store. However, these stores are huge supermarkets, and this is nearly impossible.
So, when I looked into countermeasures for tracking and data collection by AI cameras, I often saw discussion in privacy forums about how wearing masks, sunglasses, eyeglasses, and loose, solid-colored clothing can fool AI. However, at the same time, there were just as many comments emphasizing the limitations of these techniques, which may fool classical computer vision but not modern AI.
I have noticed that AI like ChatGPT has become much more advanced in recent years. How intelligent should we assume the latest commercially available AI cameras to be? Furthermore, how effective are current countermeasures (masks, sunglasses or eyeglasses, and creative clothing) when it comes to the data that AI is trying to collect and infer, or are they essentially meaningless in practice?