Research & Application
This section contains all of the research that aided the development of this project.
Literature
These sources used in the literature of my project (such as the report) have been very useful in developing my understanding of the topic as these sources date from 1992 - 2022. So alongside providing different insights on the topic, when comparing these sources with different publication dates it is clear to see the progression of editing technology with each different source.
Mitchell (1992) examines the impacts of digital photography on visual representations and how these can be easily edited. He then compares that to analogue photography, where the images taken are a real representation of reality that cannot be easily altered. Mitchell’s work supports my exploration of people’s realities potentially being altered by images being published in the media.
The Reconfigured Eye
The Photographic Image In Digital Culture
Artificial Intelligence, Deepfakes, and Disinformation
After Photography
Subjective and Objective Evaluation of Deepfake Videos
This book by Lister (2013) evaluates several essays that explore the significant changes to photography going into the digital age and how photography has transformed from a mechanical process to a digital one. Part of this book explores the emergence of smartphones and social media and how this has made it even easier for people to edit images and easily share them online. This book helps to build on the question of the extent to which people's realities could potentially be distorted on social media.
Helmus (2022) examined the threat of AI editing and deepfakes in relation to their role in spreading disinformation. The report looks at specific examples where these editing techniques can be used, such as spreading political disinformation, and how this can easily distort the realities of the general public. This report helped to build my knowledge, specifically on deepfake technology in relation to my report and how more should be done to address the spread of misinformation (including deepfakes) online.
This paper by Korshunov and Marcel (2021) explores how widely accessible deepfake technology is and how practically anyone can create a deepfake video. They believe that deepfakes will erode the truth of video evidence because it is still unclear whether the average person will be able to detect a deepfake video. This paper is useful to my work, as they carry out participant observations to see how hard it is for humans to tell if a video is a deepfake or not. This is a question that my research question explores.
Ritchin (2009) explores how digital technology has changed how photography is created, distributed, and interpreted in society. He highlights how this technology has created many positives but also negatives. One negative explored is how the authenticity of an image can be lost when an image is digitally altered and how hard these altered images can be to spot. This book helped develop my research, as part of my project added to Ritchin’s research and explored how the context an altered image is placed in can influence how easy an edited image is to spot.
Related Practice
These sources used in related practice in my report have been instrumental in providing me with a level of understanding of the topic through different forms of media. This research helped to provide insights into the topic that I did not know about before, and also helped me with the brainstorming of ideas for my project outputs, including the exhibition.
The Social Dilemma Film
This Person Does Not Exist Website
Truth Is Dead Exhibition
The Social Dilemma (2020) is a Netflix film directed by Jeff Orlowski that explores the dangerous societal impacts that social media causes and how the technology industry is behind these dangerous societal impacts. The film uses different methodologies, such as real interviews, to research the topic, something that I did in my project. Part of the film explores misinformation such as deepfakes and how the algorithms can promote this content to users, potentially distorting their realities. Some of the research undertaken in this film directly impacted the style of research I carried out.
ThisPersonDoesNotExist (2025) is a website created by Philip Wang that aims to highlight the capabilities of AI technology in the realistic image creation of human faces. The website creates faces of people that do not actually exist, and every time the page refreshes, a new face is created. The website promotes the positive advancements in AI technology but also the negative ethical implications of AI-generated images and how they can be distorting people's realities online. This is useful to my work, as the website was created to highlight this topic to the public, something that my project aims to do. The website was also useful, as this is one of the tools I will be using for my final pieces of work.
Truth Is Dead (2021) is a visual art exhibition by Alison Jackson that seeks to explore the difference in contemporary celebrity culture between ‘true’ and ‘false’, ‘real’ and ‘imaginary’, ‘original’ and ‘copy’. This is a premise Jackson believes to be blurred and confused by people and that sometimes fake can be mistaken for real as manipulation becomes a lot easier to do. This exhibition was detrimental to my work, as this is what my project aims to explore visually, similar to Jackson. This exhibition also gave me some inspiration for my visual photography.
Social Media
As my project looks at the use of edited images and AI on social media, it was important for me to do some form of research for the project on social media. For this research I browsed social media forums such as Reddit to read and even participate in some discussions about editing images on social media and the use of AI, particularly deepfakes.
Primary Research
During this project I undertook a number of interviews to get a first-hand understanding of people’s opinions on the topic and to also show them images that were real and fake to see if people could tell the difference. Originally this was going to be one of the main aspects of the project, but after discussions it made more sense to focus my project on one main output: photography. So these interviews were just undertaken to aid my research.
The findings from the interviews found that many people had not actually thought about the fact that a lot of content online is edited, potentially even created with AI. Many participants just consumed this content without thinking of the potential impacts and how it may impact their ideas of reality. Most participants from the interviews all agreed that editing of images is a completely fine thing to do. However, the amount of editing people thought to be acceptable varied from person to person. In general participants had varied views on AI images, with people stating that it is a useful tool but only in some scenarios and contexts. This was explored more in the observations where people were able to guess which image from a series of images was AI only 60% of the time. Through the observations, it was clear that people looked for potential deformities in images that can sometimes happen with AI portraits (such as AI finding hands hard to generate) to guess which image was AI. Many participants also stated that they did not know AI-generated images of portraits were so well created and realistic and were shocked at how quickly AI has progressed. Overall, the findings suggested that people's realities are being distorted online and that most people do not even realise it.
Exhibition Research
As part of this project I researched many exhibitions online and in person, some relating to my topic and some not, so I can get a better understanding of what to include in my exhibition. Below are some images from the exhibitions that I found very useful in relation to my project.
