Assignment 01: Problem Solving & Situated AI Study

Overview

Choose a focused creative, cultural, research, or design problem and investigate it with generative AI systems. Rather than treating artificial intelligence as a generic one-click solution, develop a structured body of experiments that probes a specific model, interface, or prompt framework as an artistic material with discrete affordances, biases, and boundaries.

Objectives

• Problem Framing: Articulate a clear creative or conceptual problem, defining target audience and artistic stakes.
• Empirical Experimentation: Conduct and document 6–12 iterative experiments testing varying prompt architectures, temperature/sampling parameters, and model versions.
• Critical Failure Analysis: Explicitly document at least 3 rejected or failed iterations, explaining what caused the system to fail and what the failure reveals about the model’s latent representation.
• Human Authorial Steering: Distinguish between machine default outputs and intentional human steering, addressing themes from required readings (Ted Chiang, Ken Liu).

Deliverables

1. Problem Statement (300–500 words): Written statement outlining the conceptual goal, aesthetic ambition, and chosen model ecosystem.
2. Experimentation & Prompt Log: Comprehensive log including prompt revisions, parameter settings, input assets, and generated outputs.
3. Artifact Portfolio: Curated visual, textual, sonic, or code outputs from the study.
4. Critique Presentation (3–5 minutes): In-class presentation evaluating what the system enabled, what it obscured, and where human decision-making operated.

References & Wiki Links

• Ted Chiang: Why A.I. Isn’t Going to Make Art
• Arvind Narayanan & Sayash Kapoor: AI as Normal Technology
• Jagged Frontier of AI
• System Literacy