When seeing an image, each element can be interpreted as a shape or a symbol.

When we focus on the relationship between elements, we observe that there is an occlusion relationship, where some elements appear in front of others. This reflects the distance between different elements and us.

When we consider our relationship with each element, we employ perspective to establish this connection. This perspective places the viewer in a specific position within the image.

Both the physical relationships between elements and the viewer’s position within the image, as established by perspective, attempt to replicate physical space, treating the image as a representation of the physical world. However, in our current era where AI generates images and videos and we spend increasing amounts of time with images rather than physical spaces, this approach is no longer practical.

It needs to be reaffirmed that images are read as images.

Images are symbols(image can be symbol or shape), and these symbols do not have a specific sequence or distance from one another.

For instance, Chinese pictographic text can be read out of order. If some characters in a 50-character Chinese sentence are rearranged, most people won’t notice because we read in chunks, not one word at a time. This principle also applies to symbols in images, rendering a stable spatial order unnecessary.

Therefore, we can extract symbols separately and then adjust the relationships and perspectives within the image.

Then, it is the game of shape.