What seems most basic often isn’t so. It can reflect a complex cultural landscape with roots in remote antiquity. This includes perception.
Richard Nisbett at the University of Michigan conducted an eye-tracking study that found that Caucasian Americans usually attend to focal objects more than most East Asians do; they’re more inclined to analyze separate objects and their attributes and place the objects into categories. East Asians usually attend more to contexts and make more judgements based on relationships. Pictures of foregrounded animals and inanimate objects against complex backgrounds (e.g., a tiger standing on a river’s bank, or a fighter jet flying in front of a mountain scape) were presented to subjects. Caucasian Americans looked at the foregrounded objects sooner and longer than Chinese participants did, and Chinese looked more at the backgrounds. Nisbett concluded that the Americans encoded more visual details of the objects than the Chinese, and the latter balanced objects and their backgrounds into more holistic perspectives.
Several of the most commonly used Mandarin words for perceiving also convey a holistic orientation. Kanjian (to see) combines two words. Kan means to look for, while jian means to see or to meet. It thus includes the whole process rather than a focus on distinct objects.
Most AI and virtual reality systems are heavily oriented to vision. In order to avoid biases, they’ll need to accommodate different cultures’ ways of seeing the surroundings. Deep dialogs between AI and cultural studies will be necessary if the leading makers of AI-based systems want them to fully understand humanity. Otherwise, artificial intelligence will be very artificial indeed.