Advanced Exploration of Digital Self-Representation in New Media(Nilufar Kargasova)-PART 1

 Student-Nilfuar Kargasova

Professor-Jessica Laccetti

Course- New Media Narratives: Writing & Publishing in a Developing Field


In today’s world, a few years ago, a boom in AI happened. It is a fascinating tool that helps generate information and analyze different sources, bringing them together for users' needs. On the one hand, AI is an amazing tool that makes people's lives much easier. For example, it can assist in creating dissertations, writing essays, and generating CVs based on user commands. Generally speaking, it is like a human with an IQ of 1500, who stores all the information in their mind around the globe and different timelines. However, many people find AI a concerning tool that can be dangerous at some point. For instance, in the research "AI, the New Frontier – Opportunities and Challenges," Darrel Guntur refers to biases and inaccuracies of AI. There was a case where a pregnant woman was arrested and accused of a crime (Darrel G, 2023). However, it turned out that AI didn't accurately recognize her face. Therefore, AI was accused of inaccurate prediction, and, more importantly, the woman was in jail because of it. Moreover, AI being biased can create one-sided generated information that can harm a person or an organization. Besides, the commands that you put in an AI generator are going to stay there forever. Basically, you will provide your personal info to the database of AI.

I have used different types of AI to create my digital self-representation. I used free websites such as https://vanceai.com/ and https://www.bing.com/images/create?toWww=1&redig=EB41AFC562524791A00489E01A245147. Both of the websites do not require any money for creating an image. I chose these platforms to learn about the free AI tools since not everyone can afford to pay for the apps. In the first one - vanceai.com, I put a regular photo (Figure 1) of mine and got an avatar in anime style (Figure 2). It looked quite realistic; however, the skin was smoothed, blemishes were added, and certain highlights appeared on the face, and the skin tone was even and smoothed. Other than that, the transition turned out quite realistic. So, this digitalized selfie can be used in different contexts, such as an avatar for gaming or it can be used as a profile picture on, or it can be a form of self-representation online on the YouTube platform. I would personally use the generated picture for gaming if I decided to play any anime game. After creating the animation comes the question - why is it free? In reality, most "free" websites track your behavior to use it for advertising segments (Acar G, 2014). Many people are unaware of the meanings of first-party cookies and third-party cookies. People usually just accept all the cookies which can lead to some unpleasant consequences. There are so-called two types of cookies - first party and third party (Chen, Q., 2021). First-party cookies are the ones that remember your data for the sake of your own comfort. For example, there are online shopping stores where you can accept first-party cookies, and it remembers the list of items that you marked as your favorites. That way, it is more convenient for you to search and pay for the items. However, third-party websites can track your behavior online and provide your data for advertising sectors based on your choices and preferences. Moreover, recent studies show that first-party cookies are using JavaScript code to track the behavior and store the data of the users (Chen, Q., 2021). This specific website is actually using third-party websites, even though I can regulate the settings; however, I cannot be sure that they have not used the data, specifically the access to images on my PC.

Figure 1                                                                                                 Figure 2 A person taking a selfieA person taking a selfie

Description automatically generated



Another website that I chose is a text generator. This website converts texts into images. I think in this case, the way we give a command is the most important thing. So, I had several trials to make sure I include different aspects of my life, not only referring to my physical appearance but also including social and political aspects. So, first, I gave it a command such as “a girl (24 years) from Tajikistan (Figure 3)." It showed me a girl with a more Slavic appearance, blue eyes, big lips, and very symmetrical features; however, actual Tajik women do not look like this. They have more Iranian features. So, that way, the AI generator can give you a false representation of the race of the particular nation. Next, I gave it a more specific command, such as “a Pamiri girl (24 years old) with dark curly hair, dark-skinned, who has the features of a Persian girl and who has been living in Russia for more than 10 years as a migrant. A girl who has got a high school diploma in Russia and a college degree in Kyrgyzstan. Also, that girl went through cultural shock while moving from Tajikistan to Russia. She experienced the transition from a collective society to individualistic” (Image 4). Giving a concrete command allowed me to create a more accurate avatar of myself. Image 5 includes some of my hobbies, and Image 6 shows the political and social reality of the place that I have experienced and lived in.

It is interesting to have a look at the background in each image. Figure 3 shows the background which looks like the capital of the city with tulips. It is an interesting observation because the creator of the AI might have just taken the image without asking. This can lead to a privacy issue. However, as the command changes, so does the background. Referring to migration to Russia created a more familiar background consisting of spruce, which is a very common tree in Russia (figure 4). Besides me, referring to education created a girl with a very clear look of a conscious and thoughtful person (figure 4). But if we look at Image five, the background is more concentrated on my creative hobbies, which also changes the character. The avatar’s face does not help all the sharp edges. She is soft and clever. If we look at the last figure 6, I have mentioned all the struggles, both social and political, which resulted in a woman with very sharp features who looks quite serious and sad and whose background looks like the walls of a prison. After analysis, I wonder where the background was taken from. This is a little concerning since the creator of the AI might have just taken or will do it in the future without people knowing about it (Yowono B, 2023). It might lead to serious problems such as a privacy issue and data thifting. Besides, the way AI portrays a certain race might create a huge concern. The lack of data about certain nationalities can lead to misinformation from the side of AI to the global audience.

In conclusion, the foray into AI for digital self-representation unveils its transformative potential and inherent challenges. The essay illuminates the intricate dynamics of AI, from its applications to the ethical considerations surrounding user data and racial representations. As we navigate this digital frontier, a nuanced understanding of AI's complexities becomes paramount, ensuring responsible engagement and fostering a dialogue on its evolving impact on identity and representation in the digital realm.

Figure 3



Figure4


Figure 5

A screenshot of a computer

Description automatically generatedFigure 6

References


Acar, G., Eubank, C., Englehardt, S., Juarez, M., Narayanan, A., & Diaz, C. (2014, November). The web never forgets: Persistent tracking mechanisms in the wild. In Proceedings of the 2014 ACM SIGSAC conference on computer and communications security (pp. 674-689).


Chen, Q., Ilia, P., Polychronakis, M., & Kapravelos, A. (2021, April). Cookie swap party: Abusing first-party cookies for web tracking. In Proceedings of the Web Conference 2021 (pp. 2117-2129).


Gunter, D. (2023). AI, the new frontier – opportunities and challenges: Research information yearbook 2023/2024


Yuwono, B. (2023, November). THE TRANSFORMATION OF DIGITAL IMAGING IN PHOTOSHOP WITH THE PRESENCE OF ARTIFICIAL INTELLIGENCE GENERATORS. In SETI International Conference (Vol. 1, No. 1, pp. 33-36)

Comments