womsactivity1

The Sleep of Reason

Devised by Dean Kenning

Learning Objectives:

To explore and develop understanding of the relationship between individual choices of words to describe an image, and images generated using AI based on these descriptions.

To gain a sense of the human and machine process of interpretation, description and image generation

Nb: This is a good lead-in activity to do before the Diagramming Activity Inside the Black Box of AI

Approximate duration:

30 Minutes

Equipment / Resources needed:

A digital image or another reproduction of Goya's 1798 etching, 'The Sleep of Reason Produces Monsters'. (This image was chosen because of its strange dreamlike qualities and the resonance of 'reason' for computational rationality. However this image could be replaced with another image - the important thing is that everyone uses the same image.)

Interactive Whiteboard, Digital Projector or Overhead Projector (an OHP is nice because it is an old technology and so will add to the strangeness of the image)

All participants needs access to a computer or Ipad, or one between two.

Activity:

1. Project or display a reproduction of Goya's 'The Sleep of Reason Produces Monsters' large enough so that everyone can see.

2. Briefly discuss the image - What do you see? What does it remind you of? When do we think it was made? How do we think it was made? The teacher can add a little more information. 

3. All participants are asked to input a description of the image (prompt) into Craiyon AI image generator, using any words except the title of the artwork. Several attempts can be made altering the prompts with the aim of getting a more accurate resemblance to the original image.

3. Compare and discuss the images generated with the person next to you.

Prompts for discussion could be displayed. Why is each image so different? What commonalities arise? Why do certain things tend to appear, and others not?

4. Feedback to whole class. Ask pairs to share interesting results, both similarities and differences. 

Extension activity

5. Each participant to create another image related to something they are doing at school or in education/AI research (work by an artist, a ‘theme’, etc.).

Compare and discuss the images generated.