In an inkblot test, people see different shapes, derived from their inner psyche and past visions. What would a bot see with the same test?


Test #1 Using Clairify Image recognition

The test begins as follows. This is an image that I uploaded unto Clairify. The image is my own and I haven't posted it anywhere else.  Although this is in fact a feltet bat, and does a poor job looking it, the image recognition software available was able to identify the following:

It is inverted, in the air, at a studio (perhaps maybe because of the white background) and gymnastics. Similar images includes the likes of dragons, bats, and gargoyles. So, spot on. Releif that the bat i had sewn together looks like a bat, but more importantly it recognizes it as it is.

So its not a surprise that when I submit an inkblot that most people would see a bat in, it would give me an inkblot for the similar images. But my experiment wants to work with the predicted tags and mostly, what does the bot see inside the inkblot?