This master thesis aims to map the impact of Artificial Intelligence systems, dealing with images, on the world of photography. Is an information visualization project that consists of a digital atlas and a printed volume.
Extensive research work was divided into two parts: on one hand, theoretical research on the fields of artificial intelligence and photography, on the other hand, data collection and analysis on AI systems that process images.
A set of collection methods enabled me to reach a list of ai software: those that edit images, generate images, or assist in shooting were considered.
For each software, a set of data were collected including name, available features, main means of use, claims, website etc.
A qualitative and open research approach was chosen in order to obtain an overview of the subject to bring out preponderant themes. Focus was placed on the claims used by AI services and the functions performed by the different features. Feature study mapping and qualitative repeated reading of the claims was produced to bring out the main narratives.
Therefore, three research protocols were developed. Each protocol constitutes a section of the website.
A printed volume and a website were produced to communicate the theoretical research and make the data visualizations explorable.
WEBSITE SECTION 1 (FIRST RESEARCH PROTOCOL)
The first protocol outlines the functions performed by AI services. The features have been grouped on the basis of the function performed into clusters and subclusters. An initial mapping was able to provide an overview of the system.
From the map, an explorable version on website was designed, which allows the user to access different levels of depth via zoom-in. Here: macro categories of functions performed by AI features.
WEBSITE SECTION 2 (SECOND RESEARCH PROTOCOL)
The second protocol deals with the narratives that emerged from the claims of artificial intelligence services. These themes were then cross-referenced with the functions performed by the services themselves.
WEBSITE SECTION 3 (THIRD RESEARCH PROTOCOL)
A high percentage of claims contain the word "you/yourself", placing the focus on the human subject rather than the AI. The third protocol treats all filtered expressions present in claims that contain the words "you/yourself", "your/yours".
This project was challenging because of its complexity but at the same time very interesting because of the impact of the issue on today's world. The theoretical research work siding data visualization allowed me to make a broad analysis of technical, social and cultural implications.
Individual project.
