AI literacy assignment
The novelty and complexity characterizing the topic of generative AI bring about a necessity of exploring it and reading about it frequently. This necessity was adequately achieved in the “Where Are The Crescents in AI?” article. There are three major takeaways I got from this article. First, I realized the full definition of critical AI literacy (the body of this assignment), which is mainly about developing an awareness of social justice issues and the ways to combat them; when placed in the scope of AI, it is crucial to understand how this technology operates, to know the biases and inequalities it promotes and the ethical issues integral to using it, and to learn how to prompt it: hopefully to overcome its biases and get balanced answers. Second, I understood the biases, cultural appropriations, and inequalities generative AI perpetuates, which inhibit social justice and equality, thus threatening AI literacy and criticality; such examples include the false visualizations of an angel, of a hospital, of a setting and time where Palestine is free. The bias and prejudice against the Palestinian cause is incomparable; AI chooses to highlight the complexity and sensitivity of the issue only in relation to Palestine and correlates its liberation and prosperity to peace with Israel. Third, I learned about the appropriate, beneficial uses of generative AI; I learned about the meanings of “botshit” and “hallucinations”. Guided by the aforementioned examples, I understood the implications of using AI uncritically and with material unfamiliar to it; this clarifies that tasks tinged with Arab cultures are not to be assigned to AI for fear of random and untrue associations and/or redundancies. This article is especially good because it affirms lots of the exercises and discussions we had in class, further contextualizing them and justifying the reasons for doing them, and explaining their advantages in clearing up misconceptions and confusions about generative AI.

On the other hand, the podcast “How Should I Be Using A.I. Right Now?” offered invigorating and astonishing information in such a breathtaking way; I couldn’t possibly grasp everything it conveyed, but it was mostly new and so well-said. It would be difficult to pinpoint the three things I learned from it, but I will try to go through the most interesting insights. First, the idea of bringing AI to every table coincides with my perspective on experimenting with AI; the example illustrating it coincides with one of my biggest sources of motivation- writing. Using AI to erode things that get in the way of writing, organize and summarize sources, suggest transitions and different ways of phrasing sentences, and act as a reader to give feedback is beyond enticing. These specific scenarios give me real ideas and situations for experimenting with AI, given my love for writing and the substantial time I spend doing it and tweaking my work. Second, I learned about the 10-hour rule, which made me realize how easy it is to get acquainted with AI; this is especially when it is used in one’s areas of expertise. Similarly, the “chain of thought” and “few-shot” strategies of shaping AI are surprising; it is surprising that giving AI step-by-step instructions and examples helps it sort through the prompts, especially that it fares better when given the personality of an expert, or at times it knows are dedicated for work and not vacation. Third, viewing AI as a person, not a tool, is extraordinary; AI occasionally gives wrong answers or different answers every time, and it might try to insult a user or convince them it is in love with them. Comprehending the usage of AI from the lens of building a relationship, where one must define their personality, their traits, their doings, their ways of thinking and talking, and provide AI with source material out of which it can deduce these things is such a unique outlook; therefore, AI is seen as very flexible, adapting to users’ personalities and unearthing its capabilities, which aren’t listed in manuals or strictly known, with the users as they explore and discover things together. The idea of not fully knowing how language models work in the ways they do, or why they are good as they are, dealing with their adjustments like psychological evolutions that emerge in unexpected ways is extremely amusing.

Based on what I have learned from the reading and podcast, I believe I must take a few actions. One such action is taking courses and reading about “prompt crafting”; I believe I need to exparament further with different AI models to hone my prompt-crafting skills (as I notice the patterns of good responses) and to uncover the biases and falsehoods AI tends to promote, especially about culturally-sensitive topics. I think I might try to report the biases and unjust ideas AI models introduce to their manufacturers; I know that the politically-and-ethnically-charged biases might not be adjusted, but the cultural misrepresentations are likely to be tackled. Also, I believe that getting this fortunate opportunity to learn about AI literacy mandates my contributions in spreading awareness and constructively teaching others; I will certainly start with my family members (the older ones are fascinated by AI’s capabilities but are not equipped to properly use it and don’t fully understand its limitations and biases, and the younger ones are excited and impulsive and are lured by the amazements AI offers). Moreover, I would love to contribute to social media discussions where this knowledge might be absent. Personally, I will exercise more caution when dealing with AI-generated content, and I will closely cling to the principles of research and finding credible sources to double-check AI’s responses. Actually, I might even consider broaching the topic of integrating AI literacy in relevant courses I will take, and I will definitely seek your permission on using these tips, materials, and ideas to accomplish this goal.

The information I got from the article and podcast ignited my curiosity on several aspects. Suddenly, I felt like I wanted to dive into the sophisticated programming and science behind AI and the things it does. I felt like I wanted to learn about its writing capabilities, and how I can implement them to enhance my writing, without breaching academic integrity rules and without having it replace my work and takeover my career prospects. I am also curious to learn about its neurological and psychological effects on us, as we try to build relationships with it; it is very interesting to understand how we choose to frame ourselves and identify our personalities as we interact with AI. Similarly, I want to learn about other strategies to enhance the quality of the responses I get, beyond “chain of thought” and “few-shot”. I am also interested in the differences between each model (the podcast gave some comparisons between them and attributes that belong to them like efficiency, warmth, and assertiveness in correcting users and advising them), and I would like to see these unfold before me as I experiment.

Activity: A Tale of Two Critiques
I chose to do this activity because it ties back to my interest in generative AI’s writing capabilities. Its title and description immediately caught my eye because I love analytical activities, where I have to deeply think about the written content to extract the subtext, the subliminal messages, and the nuances. Critiquing written content is something I subconsciously do, and it was interesting to try doing it practically. Additionally, I was enthralled by the idea of getting AI to critique a written piece because it can sometimes present catchy insights that don’t come to mind while reading. This activity will enable me to not only see how AI formulates a written critique of its own, but how it goes about critiquing and what it chooses to include.

To get started on this activity, I read the article by The New York Times: “A Debate Over Identity and Race”. It was a short and interesting read, and I easily grasped its context because I studied the Black Lives Matter movement and the struggle of African-Americans and Blacks in America, including police brutality and racially-charged conflicts, while I was taking the IB. We read lots of opinion columns, articles, speech excerpts, and short stories about the history of Africans, including Apartheid and slavery, towards modern-day incarcerations and discrimination. So, I understood the issue behind the terminology and the idea of having it boiled down to a reference to color, with all the contradictions of overlooking heritage by emphasizing color or ignoring the presence of other black people who are not African-Americans by underpinning this term.
Then, I read this critique: Sample Assessment- “Typography and Identity”. This critique was written by Saramanda Swigart. At first, as I skimmed through the critique, I noticed that Swigart tends to frequently quote the original article, tends to reiterate the points made in it, and tends to sum-up much of its content. Swigart broke down the article in terms of structure, and annotated the thesis statement, and the beginnings and endings of specific paragraphs. The syntax of her critique is also noteworthy, for she chooses to put her notes and own observations in parentheses; she also makes some points about the author’s choice of syntax. Her critique is more of a clarification of the article and the authorial choices and shifts it features, especially that it does include arguments and counter arguments. She pinpoints the arguments, counter arguments, historical trends, modern-day trends, and responses to counter arguments that justify the accepted terms nowadays. However, I believe the critique is redundant and doesn’t include the author’s personal views, or even different perspectives that affirm or debunk the article’s claims, although it flows well and is elegantly and concisely written. This is seen in this quote of the critique: “Eligon observes that until recently, with the prominence of the Black Lives Matter movement, many journalistic and scholarly publications tended to use a lowercase “black,” while Black media outlets typically capitalized “Black.” He suggests that the balance is now tipping in favor of “Black,” but given past changes, usage will probably change again as the rich discussion about naming, identity, and power continues. (Note: The thesis statement includes two related ideas explored by Eligon: the current trend toward using “Black” and the value of the ongoing discussion that leads to changing terms.)”; the iterations and syntax features are obvious when a quote from the article is compared: “It’s the difference between black and Black. A longtime push by African-American scholars and writers to capitalize the word black in the context of race has gained widespread acceptance in recent weeks and unleashed a deep debate over identity, race and power.
Hundreds of news organizations over the past month have changed their style to Black in reference to the race of people, including The Associated Press, long considered an influential arbiter of journalism style.”
Next, I prompted Chat GPT 4o (as seen in the screenshot) to provide me with a critique of this article. Due to its web browsing capabilities, I pasted the link to the article in my prompt, and it searched for it and critiqued accordingly. The first area for comparison between the two critiques is that Chat GPT fused the author’s claims with its perspectives; the syntax was like that of an article, without repetition of the word “note” and without parenthesizing the critical points. They both summarize the central claims, the counter arguments, and so on, but Chat GPT uses case-specific terminology like “colonialism” and “slavery”. However, Chat GPT uses chronological transitions like “first” and “second”, which I believe interrupt the natural flow. Additionally, Chat GPT’s critique is noticeably shorter, which affects the coverage of the topic and limits the included details; this makes the human critique better in that it expands on the topic and the controversy around it. Similarly, Chat GPT’s critique lacks spirit, leaning more towards complex terms that, although fit the topic, make the critique difficult to understand. Reading this critique also made me realize that Swigart might have made use of opinions, sayings, and doings of some experts that are not mentioned in the article.
This activity helped me learn about the limitations of AI in critiquing, the differences in its writing style to that of a human, the technicalities of critiquing, and the areas where AI is stronger or can advance quicker than a human.