toplogo
Resources
Sign In

Leveraging Generative AI to Construct Multimodal Voice in Student Writing


Core Concepts
Students can leverage generative AI tools to purposefully construct a coherent and unique voice in multimodal writing by strategically remixing textual and visual resources.
Abstract
This study explores how students use generative AI tools, such as Midjourney, to construct their authorial voice in multimodal photo essay assignments. The analysis reveals two key remixing practices that students employed: Layering: Students carefully considered the relationship between text and AI-generated images, using visuals to complement, reinforce, and evoke emotions in their storytelling. They approached image generation with clear intentions, prompting the AI system to create images that align with the overall tone and rhetorical goals of their work. Blending: Students drew on their knowledge of other genres, such as profiles and biographies, to structure their photo essays. They also blended academic and public-facing registers in their writing to establish an intimate yet scholarly voice. The study also highlights the affordances and challenges students experienced when using generative AI tools. Key affordances include the high quality of AI-generated images, the pleasant composing experience, the satisfactory outcomes, and the facilitation of voice expression. Challenges include difficulties in visualizing abstract ideas, articulating prompts effectively, and dealing with the unpredictable and biased nature of AI outputs. The findings suggest that incorporating generative AI tools in multimodal writing assignments can provide valuable opportunities for students to develop prompt literacy, multimodal awareness, and critical perspectives on the ethical implications of AI technology.
Stats
"It was really hard to figure out what components should be in the photo. Like how do I represent English? How do I represent an abstract idea into those photos?... I can't visualize the education act itself. I also can't visualize English itself without finding a symbol of English" (Ana, interview). "Mine would be generally like the perspective, the colorization, like how Sue would look in the photo, because I was providing a pretty general description about her" (Nathen, interview). "For this first image, I kind of wanted the people around the table looking a little bit more upset or confused, but I could only get them kind of also looking excited and rowdy. I mean, they do show some concern. But I think that AI kind of yielded the full image that I was expecting" (Linda, interview). "Because I said Thai students, I guess that's where the background came from. Perhaps AI assumes that Thai students wouldn't be as rich to have a fancy movie theater, like a fancy projector. So that's what came out. It has prejudice, I guess" (Sandy, interview).
Quotes
"The first picture here [See Figure 10] I wanted to put the flag of South Africa because that's where the lead quest starts from. And there are different great people in history who have done great deeds. Unfortunately, I don't know what every single person did in that picture...I keep looking at all the people in there and I think I searched for a few of them, but I wish that AI generated a picture that could tell me who those people are actually in the collage in the beginning of the picture, there's so many people there." (Harry, interview) "It would give me images that I felt like just didn't relate to what I was trying. They didn't convey the image that I wanted to convey." (Emma, interview)

Key Insights Distilled From

by Xiao Tan,Wei... at arxiv.org 03-29-2024

https://arxiv.org/pdf/2403.19095.pdf
Purposeful remixing with generative AI

Deeper Inquiries

How can writing instructors effectively scaffold students' prompt literacy and multimodal awareness to help them leverage generative AI tools in constructing voice?

To scaffold students' prompt literacy and multimodal awareness for leveraging generative AI tools in constructing voice, writing instructors can implement several strategies. Firstly, instructors can provide explicit instruction on how to formulate effective prompts for AI tools. This includes teaching students how to articulate their ideas clearly and concisely, ensuring that the prompts align with their intended rhetorical purposes. Additionally, instructors can guide students on how to adjust and refine their prompts based on the feedback received from the AI-generated outputs. Furthermore, instructors can incorporate activities that enhance students' multimodal awareness. This can involve analyzing and deconstructing multimodal texts to understand how different modes interact to convey meaning. By engaging in activities that explore the relationships between text and images, students can develop a deeper understanding of how to effectively combine various modes to construct a coherent voice in their writing. Instructors can also encourage students to reflect on the ethical implications of using generative AI tools in multimodal composition. By discussing issues such as bias, stereotypes, and the limitations of AI technology, students can develop a critical awareness of the ethical considerations involved in leveraging AI tools for creative expression. This reflective practice can help students make informed decisions about how to ethically and responsibly use AI in their writing. Overall, by providing targeted instruction on prompt literacy, facilitating activities that enhance multimodal awareness, and fostering discussions on ethical considerations, writing instructors can effectively scaffold students' skills to leverage generative AI tools in constructing voice in multimodal compositions.

What are the potential ethical concerns associated with using generative AI in multimodal composition, and how can we address them in the classroom?

The use of generative AI in multimodal composition raises several ethical concerns that writing instructors should address in the classroom. One major concern is the perpetuation of biases and stereotypes in AI-generated content. AI systems can inadvertently reinforce existing biases present in the training data, leading to the reproduction of harmful stereotypes in multimodal compositions. To address this concern, instructors can educate students about the potential biases in AI technology and encourage them to critically evaluate the outputs generated by AI tools for any signs of bias or stereotypes. Another ethical concern is the issue of authorship and originality in AI-assisted compositions. Students may face dilemmas regarding the ownership of AI-generated content and the extent to which they can claim authorship over materials created with the assistance of AI tools. Instructors can guide students on how to navigate these ethical considerations by emphasizing the importance of transparency, attribution, and integrity in their use of AI technology. Encouraging students to clearly acknowledge the role of AI in their compositions and to uphold academic honesty can help mitigate concerns related to authorship and originality. Additionally, privacy and data security are significant ethical considerations when using generative AI tools in the classroom. Instructors should ensure that students understand the implications of sharing personal data or sensitive information with AI systems and provide guidance on how to protect their privacy while engaging with AI technology. By promoting awareness of data privacy issues and implementing measures to safeguard student information, instructors can address concerns related to privacy and data security in AI-assisted multimodal composition activities. By proactively addressing these ethical concerns and fostering a culture of ethical awareness and responsibility in the classroom, writing instructors can help students navigate the complexities of using generative AI in multimodal composition while upholding ethical standards and values.

In what other genres or contexts could generative AI tools be meaningfully integrated to enhance students' creative expression and critical engagement with technology?

Generative AI tools can be meaningfully integrated into various genres and contexts to enhance students' creative expression and critical engagement with technology. One potential application is in the field of creative writing, where AI tools can assist students in generating plot ideas, character development, and dialogue for fiction writing. By leveraging AI for creative writing tasks, students can explore new narrative possibilities, experiment with different writing styles, and enhance their storytelling skills. In the realm of visual arts and design, generative AI tools can be used to create digital artwork, graphic designs, and multimedia presentations. Students can collaborate with AI systems to generate visual elements, experiment with different visual styles, and explore innovative approaches to visual storytelling. By incorporating AI tools into visual arts education, students can expand their creative repertoire, develop technical skills in digital design, and engage with cutting-edge technologies in the creative process. Furthermore, in the field of game design and development, generative AI tools can empower students to create interactive narratives, dynamic game environments, and intelligent game characters. By integrating AI into game design projects, students can explore the intersection of storytelling, technology, and gameplay, fostering critical thinking, problem-solving skills, and collaborative teamwork in the context of game development. Overall, generative AI tools have the potential to enhance students' creative expression and critical engagement across a wide range of genres and contexts, including creative writing, visual arts, design, game development, and beyond. By incorporating AI technology into various educational settings, instructors can provide students with opportunities to explore new modes of creative expression, experiment with innovative technologies, and develop essential skills for the digital age.
0