Foundational Learning First

Meghdad Asadilari, School of Film and Animation
Meghdad Asadilari, Assistant Professor, Film and Animation

I developed a foundation-first approach to teaching animation, computer programming, and compositing that balances strict creative ethical boundaries with practical, automated workflows. In my classroom, students are required to master manual labor and foundational principles, such as hand-rotoscoping moving subjects and writing basic code-based animations, before they are permitted to use AI tools. By teaching students to manually struggle through these foundational steps, they build the conceptual knowledge necessary to effectively direct, troubleshoot, and validate AI outputs. Furthermore, I demonstrate live, Socratic fact-checking of AI in front of the students to teach information literacy and expose AI hallucinations, and I encourage transparent, ethical dialogue when faced with concerns about AI and creative ownership. Ultimately, this approach empowers students to control the technology rather than be controlled by it, preserving original artistic touch and technical rigor.

To integrate generative tools while maintaining rigorous skill development, you can implement this four-step workflow:

Mandate manual processes (such as frame-by-frame rotoscoping or writing codes using topics covered in the lecture) before allowing automated alternatives. Experiencing this struggle engraves core mechanical principles into the brain, providing the exact foundation needed to diagnose, evaluate, and correct flawed AI outputs later on.

Ensure students master foundational theory (like variables, logical operations, and loops) before they use AI for scripting. Without understanding the underlying syntax, they will lack the vocabulary to write effective prompts, troubleshoot errors, or validate the code AI generates.

Query AI platforms live in class and execute their recommendations step-by-step. Let students observe hallucinations and errors in real-time, training them to treat AI as a statistical language model requiring strict fact-checking rather than a flawless oracle.

Address student resistance and skepticism by engaging in transparent, ethical dialogues. Rather than mandating AI use or banning it outright, open conversations about copyright and artistic ownership establish mutual trust, allowing you to guide students toward controlled, responsible use of AI as an analytical tool rather than an uncritical shortcut.

  Click here for another example of how RIT faculty have adapted their teaching methods in response to AI.