This repository contains materials for the Prompt Engineering class. It focuses on practical skills for generating text, images, video, and sound using AI tools — with a strong emphasis on creative experimentation and critical reflection.
- Overview
- Competencies
- Session Summaries
- Tools & Resources
- Ethics & Reflection
Prompt Engineering (PROMPT) equips students with hands-on prompting techniques for different modalities: text, image, music, and video generation. Emphasis is placed on both creative craft and responsible use.
Each week includes mini-projects that simulate real-world creative workflows, supported by ethical discussion and critical reflection.
Professional
- Design effective prompts for diverse AI tools and contexts.
- Adapt prompts to different models and use cases.
- Understand the limits and appropriate use of generative systems.
- Recognize privacy, security, and IP risks when using AI.
Methodological
- Test, compare, and refine prompts systematically.
- Analyze structure and wording to improve outputs.
Personal
- Engage critically and creatively with AI tools.
- Communicate transparently about AI-assisted work.
- Reflect on ethical and societal implications.
Focus: Understanding visual generation models and mastering descriptive, stylistic prompting through the Artistic Telephone game.
- Understand how text-to-image/video models interpret language.
- Construct scene-based prompts that define subject, style, and mood.
- Observe interpretation drift and prompt clarity in creative chains.
- Model demo & style comparison (Replicate / SDXL / Pika Labs)
- Crafting effective art-style prompts
- Writing full visual scene descriptions
- Artistic Telephone collaborative exercise
Focus: Exploring AI music and sound generation — prompting for genre, tempo, and mood across different models through the PromptVision contest.
- Understand how AI systems synthesize and remix audio.
- Write structured prompts for mood, genre, and tempo.
- Evaluate and compare audio outputs for coherence and style.
- Collaborate in creative “AI Eurovision”-style performance.
Focus: Translating aesthetic “vibes” into generative logic using p5.js and interpreting those human-coded rhythms through AI video generation inspired by Rafael Lozano-Hemmer’s Internet.
- Express mood, rhythm, and motion through parameter-based “vibe coding.”
- Understand how procedural systems embody aesthetic language.
- Generate visual outputs and transform them into AI video interpretations.
- Compare human vs. AI perception of visual rhythm and motion.
- VibeCoding exercise in p5.js (no prior coding required)
- Group reinterpretations of Internet
- Frame capture and AI video generation on Replicate
- Reflection on motion, authorship, and interpretation drift
- Image Generation: Stable Diffusion XL, Kandinsky 3, PixArt, DALL·E 3, Midjourney
- Video Generation: Runway ML, Pika Labs, Kaiber, Sora (conceptual), Replicate video models
- Audio Generation: MusicLM, AudioLDM, Mubert, Suno.ai
- Creative Coding: p5.js (browser editor)
- Experimentation Platform: Replicate
- Workflow UIs: Automatic1111, ComfyUI
Topics covered across sessions:
- Copyright and dataset provenance
- Bias, representation, and inclusivity in generative media
- Disclosure and transparency in AI-assisted work
- Synthetic reality, interpretation, and creative responsibility