teaching

Links to Syllabi:


Teaching Statement

From Pinhole to Pixel: Cultural Contexts and Technological Evolution

In my Computational Photography course, students begin by piercing blackout drapes with pins and projecting images of the world on their hands. From there they build dark boxes and try to capture those fleet images with iron and silver. This is how they begin to experience fundamental operations of light and image capture. The journey from pinhole to pixel lets students trace with their own hands how each evolution in technology has given us new ways of seeing and understanding the world around us.

Over more than five years of teaching at Scripps College, I have developed an approach to teaching technology through direct, hands-on experience. In my classroom, students begin by encountering fundamental principles through physical engagement. For example, when teaching digital imaging, students create portraits using software that exactly reproduces the specifications of the first digital camera built by Steve Sasson in 1975: 100x100 pixels with 4 bits per pixel, allowing just 16 shades of grey. By working within these stark constraints students directly experience how digital images are constructed from discrete units of information. The exercise bridges art and computer science: students must make creative decisions about portraiture while understanding how binary numbers represent grey values, how memory constraints shape image quality, and why early digital cameras made specific technical trade-offs.

This foundational understanding prepares students for more advanced exploration in the second semester of Computational Photography. Here, they trace the evolution of imaging technology through hands-on recreation of pivotal innovations. Students manipulate image matrices with OpenCV, reverse-engineer Instagram filters, and reconstruct color photographs from Prokudin-Gorsky's revolutionary three-color separation plates from the early 1900s. Each project builds upon their earlier understanding of digital fundamentals while revealing how technological advances have transformed not just our technical capabilities, but our ways of seeing and representing the world. Through this progression—from basic binary representations to sophisticated computational techniques—students develop both technical mastery and critical awareness of how imaging technologies shape our perception and creative expression.

This exploration of how technology shapes perception continues in my CalArts course Fuzzbox Physics, where students discover the fundamentals of sound processing through direct experimentation. The class begins outdoors, where students map natural acoustic phenomena—measuring how sound waves bounce off buildings, experimenting with echo timing, and documenting how different materials absorb and reflect frequencies. These acoustic principles then inform their work with electronic sound manipulation: students physically modify speakers to understand frequency response, build circuits to understand signal amplification, and ultimately create their own distortion pedals. By the time they analyze classic circuits like the Fuzzface's germanium transistor clipping or the Octavia's frequency-doubling transformer, students understand both the mathematical principles of waveform manipulation and the musical implications of different circuit designs. This hands-on progression helps students bridge the gap between physical acoustics, electronic theory, and artistic application.

In Tangible Media, students integrate their understanding of visual and sonic systems to explore communication and interaction design. The course culminates in a 'message relay' project where students create devices that must transmit signals to their neighbors within precise timing constraints. Each student builds a unique communication mechanism—some using light sensors, others sound waves or mechanical triggers—forming a chain of diverse information transfer methods around the classroom. When working correctly, the message cascades through each project in sequence, ultimately triggering a celebratory burst of confetti. This shared moment transforms individual technical achievements into collective success, while demonstrating how different forms of mediation can create unified experiences. Students learn not just about signal processing and timing systems, but about designing robust interfaces between different communication methods and the importance of graceful error handling in interactive systems.

In Machine Learning for Artists, students explore contemporary technological mediation through direct engagement with AI systems. Starting with foundational techniques, students recreate seminal works like Mario Klingemann's neural style transfers and Anna Ridler's tulip-generating GANs, learning not just to use these tools but to understand their underlying mechanisms. The course progresses from basic image classification—where students discover how neural networks learn to distinguish features and patterns—to more complex generative projects. For instance, when training their own StyleGAN models, students must grapple with dataset curation, understanding how their choices in training data shape the system's output and biases. In their final real-time installations, students might use pose detection to drive generative art, or create interactive systems that reveal machine learning's interpretive processes. Through these hands-on projects, students experience how ML systems mediate between human and machine perception, discovering both the creative possibilities and critical implications of AI art.

A proposed course, Data Structures in Place: From Linked Lists to Songlines, explicitly examines how different cultural frameworks encode and organize knowledge. Building on my previous courses' exploration of technological mediation—whether through light, sound, or AI—this course reveals how knowledge systems themselves shape our understanding of place and memory. For instance, when teaching linked list data structures, we compare Western computer science's node-pointer model with how Cherokee oral traditions encode interconnected information through place-based narratives. Students discover how a linked list's linear traversal parallels yet differs from how Indigenous knowledge systems navigate interconnected stories and locations. By examining how different cultures solve similar information organization challenges, students gain deeper insight into both computational and cultural approaches to knowledge representation. The course demonstrates how Indigenous ways of knowing can illuminate fundamental computer science concepts while revealing the cultural assumptions embedded in conventional data structures..

My assessment approach reinforces these pedagogical principles through a structured rubric that emphasizes process, discovery, and critical reflection. Adapted from Golan Levin at Carnegie Mellon, the rubric evaluates students' curiosity in asking probing questions, tenacity in solving complex problems, precision in execution, and inventiveness in exploring unexpected solutions. Rather than focusing solely on technical success, students create detailed project portfolios that demonstrate their engagement with both intended and unexpected outcomes. For example, when a Computational Photography student's recreation of Sasson's first digital camera reveals the artistic possibilities of extreme pixel constraints, or when a Fuzzbox Physics circuit produces unplanned harmonics, these 'departures' become opportunities to understand system behavior at a deeper level. Through comprehensive documentation requirements, students not only record their technical process but analyze how their discoveries connect to broader principles of technological mediation. This emphasis on documentation and reflection ensures students engage deeply with both the technical and cultural implications of their work.

Throughout this curriculum—from early digital cameras to Indigenous data structures—students develop a technological literacy that transcends mere technical competency. Through structured peer critique and hands-on exploration, they learn to articulate how their technical and artistic choices engage with larger questions about representation and cultural understanding. A student training an AI model must consider not just classification accuracy but dataset bias; a student building a sound installation must understand both circuit design and acoustic psychology. Most importantly, they discover that technological systems are not neutral tools but culturally embedded frameworks for understanding and representing the world. By experiencing how different cultures and time periods have approached similar challenges of encoding and transmitting knowledge, students develop critical awareness of how all systems—whether computational, cultural, or historical—shape our perception and expression.