1 of 39

Creation of Multimedia in the Age of Generative AI

Anastasia Pantazopoulou, Jade Hoyer, Kate Brooks, Wiebke Kuhn

LTC Lunch, March 4, 2025

2 of 39

AI or Not AI?

"It's only AI when you don't know how it works; once you know how it works, it's just software."

3 of 39

Outline

  • What kind of media creation is possible with generative AI?�
  • Faculty stories: Anastasia�
  • Faculty stories: Jade�
  • Table discussion�
  • Copyright implications

4 of 39

What can we do with generative AI right now when it comes to multimedia content?

  • Image
    • Gemini,
    • Adobe Express/Firefly,
    • ChatGPT (DALL-E)
  • Video
    • Midjourney
    • Sora
  • Music and sound
    • Elevenlabs
  • Avatar

5 of 39

CLAS 132:

Fantasy and Science Fiction Then and Now

Anastasia Pantazopoulou

Visiting Assistant Professor

Classics Department

Winter 2024

6 of 39

Course Topic

CLAS 132: Fantasy and Science Fiction Then and Now

Course Topic: Lucian’s True History (2nd c. AD novel) and the tropes of the fantasy/sci-fi genre(s) from antiquity to modernity

Assignment: Comic Book-Group Project (Final Course Assignment)

Project Title: Lucian’s True History. The Comic Book

7 of 39

The Assignment

Assignment Description:

“In small groups, you will be assigned (by lot) one of the scenes/episodes from Lucian’s True History and re-imagine it into a comic-book format. For the visual art of the comic book, you can use Adobe Firefly or do your own drawings. For the panel design, we will use Canva. On the last day of the class, you will present your work, reflecting on your choices and the process.”

Assignment Objectives:

  • foster collaboration and creativity
  • enhance students’ understanding of written and visual storytelling forms and formats
  • encourage students to think of ways that would allow them to effectively share their academic interests and work with audiences outside academia.

8 of 39

Assignment Preparation

Week 3: Workshop led by Dann Hurlbert (on Adobe Firefly)

Week 6: Comic Book Workshop (on Canva and creating a comic book)

Lorem Ipsum

9 of 39

Did it Work?

Lorem Ipsum

10 of 39

What did the students say about their engagement with Adobe Firefly?

  • Enjoyed the overall result

But

  • Time-consuming process
  • Not producing exactly what they had in mind

11 of 39

“But is it Art?”

AN ARTS 278

AI Assignment

Jade Hoyer,

Assistant Professor of Art

Department of Art & Art History

12 of 39

The Assignment

  1. Students read both Walter Benjamin’s “The Work of Art in the Age of Mechanical Reproduction,” and a NY Times op-ed on AI + Art, and then responded to guiding questions.
  2. Students used Adobe Firefly to generate “intaglio-esque” imagery.
  3. Students created an intaglio print responding to the their Firefly exercise, assembling a composition of components from different AI generations.

These steps in tandem offered students a means to evaluate their own artistry in the context of machine collaborations.

13 of 39

LEarning objectives

  1. (Conceptual) to consider the “Printmaking Question” (valuation of art due to reproducibility and machine engagement) in contemporary frameworks, including AI
  2. (Technical) to engage with a new tool, Firefly, as a means for image generation and ideation
  3. (Conceptual + Technical) to synthesize this conceptual framework and skill development towards the creation of unique artwork

14 of 39

AI EXERCISE

Rembrandt, Christ with two thieves, 1653

15 of 39

AI EXERCISE

Text from left: �intaglio many TV sets / intaglio many TV sets with bunny ear antenna in love/ intaglio TV bunnies and small background hills / intaglio just plz give me a cosmic landscape perhaps the apocalypse?

16 of 39

Guided Questions

  1. What’s your take on AI + Art?
  2. How does your conception of art relate to the readings, + the notion of artistic worth with regards to the artist’s hand, makership and agency, media use, technology, and automation?
  3. With this in mind, what is your concept for your next print? �(write a 1-sentence summary imagining a future work)*

17 of 39

Student Approaches + Responses

18 of 39

Prompting text: Kaylah’s physics homework!

19 of 39

It was fascinating for me to use AI to generate images/ create things, and then incorporate elements into my art that I usually would not consider.

Kaylah

20 of 39

Prompting text: “cow-themed alien spaceships”

21 of 39

“While I think there is a possibility for AI art to be used in the future as a tool, in its current state it hurts artists more than it helps us. But I REALLY liked my cows!”

Sofia

22 of 39

TEACHING TAKEAWAYS

23 of 39

LEarning objectives

  1. (Conceptual) to consider the “Printmaking Question” (valuation of art due to reproducibility and machine engagement) in contemporary context(s), including AI
  2. (Technical) to engage with a new tool, Firefly, as a means for image generation and ideation
  3. (Conceptual + Technical) to synthesize this conceptual framework and skill development towards the creation of unique artwork

24 of 39

What happened?

“Glows”

  • Students enjoyed imagery generation
  • After set-up, Firefly was user-friendly
  • The assignment offered a relevant, experiential way to contemplate artistry + the machine
  • Students readily met first conceptual and technical SLOs

“Growth”

  • Synthesizing AI imagery to create new work proved to be a challenge!
  • At the point in the term, students were still developing their creative voices; adding AI complicated this trajectory
  • Many used Firefly as a generator rather than a leaping off point, and felt constrained or overwhelmed as a result

25 of 39

The Assignment

  1. Students read both Walter Benjamin’s “The Work of Art in the Age of Mechanical Reproduction,” and a NY Times op-ed on AI + Art, and then responded to guiding questions.
  2. Students used Adobe Firefly to generate “intaglio-esque” imagery.

  1. Students created an intaglio print responding to the their Firefly exercise, assembling a composition of components from different AI generations.

These steps in tandem offered students a means to evaluate their own artistry in the context of machine collaborations.

I need a step �(or a stop) here!

26 of 39

For next time!

  1. Evaluate the timing of assignment: provide more scaffolding for Benjamin and an intermediate step between AI composite imagery and an intaglio print; move assignment later in term, perhaps run in advanced class
  2. Evaluate the scope of assignment: consider using this assignment as a visual, creative exercise, such as creating a drawing, rather than a large project
  3. Keep the fun parts: students enjoyed engaging with Firefly and framing it as a silly and irreverent activity; the exercise functioned well as a means of contextualizing print theory

A caveat: I would not employ this exercise for printmaking media that engage extensively with digital output (silkscreen, risograph, photolithography)

27 of 39

image reference: Jackson Pollock, Untitled

28 of 39

“I fed my friends’ band’s lyrics into it:

‘if it weren’t so big the world would hold me / sometimes when I’m walking in a straight line.’

Then I kept making iterations, using each new version to depart more from where I had started.”

Maya

29 of 39

30 of 39

31 of 39

32 of 39

Questions for Anastasia and Jade

33 of 39

Table discussion

What kind of multimedia content might work for your class?

What would you need to make this happen?

34 of 39

Generative AI and Copyrightability

Considering copyrightability with students who are using generative AI in the classroom provides a framework for them to assess their role as a creator and the role of generative AI as a tool.

Document discussed:

U.S. Copyright Office. Copyright and Artificial Intelligence. Part 2: Copyrightability. By Shira Perlmutter. January 2025. https://www.copyright.gov/ai/.

35 of 39

An idea cannot have copyright, but a human’s expression of that idea can.

Three Kinds of Engagement with Generative AI Systems

  • Prompts that instruct an AI system to generate an output
  • Expressive inputs that can be perceived in AI-generated outputs
  • Modifications or arrangements of AI-generated outputs

36 of 39

Prompts that instruct an AI system to generate an output

  • (Generally) prompts alone will not count as human authorship.
  • Given the current state of AI technology, prompts “do not appear to adequately determine the expressive elements produced, or control how the system translates them into an output” and may “reflect a user’s mental conception or idea, but they do not control the way that idea is expressed.”

A Recent Entrance to Paradise, Wikimedia Commons

Thaler v. Perlmutter (2023)

37 of 39

Expressive inputs that can be perceived in AI-generated outputs

If a system allows for the input to be maintained in the output (example of inputting an original sketch and asking a system to modify it), those elements would be copyrightable.

Rose Enigma, Kris Kashtanova, website

Letter of application for copyright

38 of 39

Modifications or arrangements of AI-generated outputs

  • When selection coordination and arrangement of AI-generated material shows significant human authorship the “resulting work as a whole constitutes an original work of authorship.”
  • If there are enough (generally 3+) edits, enhancements,

or modification to the original output copyright would

extend to the modified material, but not the underlying

AI-generated work.

Zarya of the Dawn, Kris Kashtanova, website

39 of 39

Successful copyright claims must demonstrate control of output

Factors That Favor Copyrightability

Factors that Disfavor Copyrightability

  • Range of expressive output is constrained by the user
  • AI-generated content has been demonstrably altered by a human author
  • Input and output are different media (requires the model to interpret)
  • Human authorship is limited to prompts alone