Research map / Large language models / AI and society
Designing Characters with AI: An Art & AI Learning Activity
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Presents a four-to-eight-step character design learning activity, delivered via a Google Colab code notebook and activity sheet, that integrates OpenAI models into an art-making workflow, and evaluates it with 39 middle and high school students.
Problem
Traditional art lessons rarely integrate AI in ways that mirror real iterative art-making workflows, leaving K-12 students with limited AI literacy for creative practice.
Approach
Adapts a professional character design process (briefing, referencing, sketching, production, animation, personality design) so that AI is used at several stages (reference generation, visual feedback, personality chatbots) within a guided, low-code Colab notebook, letting students choose where AI fits in their workflow.
Key results
- Pre/post surveys showed significant increases in AI skills and attitudes (3.17 to 3.62, p<0.001) and AI confidence (3.19 to 3.59, p=0.0015).
- Post-activity trust in AI feedback rose from 3.0 to 2.17 reversed-item scale and agreement with AI feedback rose from 3.17 to 3.92 (p=0.02).
- All but one of 39 participants completed the activity; 95% engaged with the optional personality and chat activities.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use the AI-assisted character design activity | Middle and high school students · Character design activity survey | Trust in AI feedback: 3 ± 0.953 -> 2.167 ± 0.577 | Helped |
| Use the AI-assisted character design activity | Middle and high school students · Character design activity survey | Prompt-writing knowledge: 2.417 ± 0.669 -> 2.25 ± 0.622 | No clear effect |
| Use the AI-assisted character design activity | Middle and high school students · Character design activity survey | Agreement with AI feedback: 3.167 ± 0.835 -> 3.917 ± 0.515 | Helped |
| Use the AI-assisted character design activity | Middle and high school students · Character design activity survey | Understanding of AI responses: 2.75 ± 0.866 -> 4 ± 0.603 | Helped |
| Use the AI-assisted character design activity | Middle and high school students · Character design activity survey | Confusion about how AI works: 2.667 ± 0.778 -> 2 ± 0.853 | Helped |
| Use the AI-assisted character design activity | Middle and high school students · Character design activity survey | Perceived AI bias: 3.417 ± 1.379 -> 3.25 ± 0.965 | No clear effect |
| Use the AI-assisted character design activity | Middle and high school students · Character design activity survey | Moral acceptability of AI art creation: 2.917 ± 0.669 -> 3 ± 0.426 | No clear effect |
| Use the AI-assisted character design activity | Middle and high school students · Character design activity survey | Confidence using AI tools in code: 3 ± 1.128 -> 3.167 ± 0.718 | No clear effect |
| Use the AI-assisted character design activity | Middle and high school students · Character design activity survey | Confidence using code notebooks: 3.167 ± 1.193 -> 3.417 ± 0.669 | No clear effect |
| Use the AI-assisted character design activity | Middle and high school students · Character design activity survey | Confidence drawing characters: 3.667 ± 0.888 -> 3.5 ± 1 | No clear effect |
| Use the AI-assisted character design activity | Middle and high school students · Character design activity survey | Confidence using reference images: 3.583 ± 0.996 -> 4 ± 0.739 | No clear effect |
| Use the AI-assisted character design activity | Middle and high school students · Character design activity survey | Confidence explaining AI to friends: 2.583 ± 1.165 -> 3.417 ± 1.084 | Helped |
Surprising findings
- Students' creative workflows clustered into personas such as 'Dependent' AI inspiration-seekers (62% in session 1, 92% in session 2) who replicated AI reference images.
- Many students dismissed AI feedback (54% of session 1), and some found AI feedback generic or unhelpful for physical art.
Limitations
- Language barrier: sample code and materials largely English, excluding one non-English-speaking participant.
- Small sample (39 students across two workshops) with no control group.
- Middle school students found the activity harder and may need more scaffolding in Python and Colab.
Experimental setup
Metrics: pre/post Likert survey shifts qualitative workflow coding activity completion rate
Builds on: Dreaming with AI DAILy curriculum PopBots
Closest papers by meaning
- Co-designing AI Education Curriculum with Cross-Disciplinary High School Teachers (AAAI 2024 · similarity 0.85)
- Advancing Research on Equitable AI Education Through a Focus on Implementation: Insights from a Middle School Computer Vision Module Beta-Test (AAAI 2025 · similarity 0.85)
- “From Unseen Needs to Classroom Solutions”: Exploring AI Literacy Challenges & Opportunities with Project-Based Learning Toolkit in K-12 Education (AAAI 2025 · similarity 0.84)
- Constructing Dreams Using Generative AI (AAAI 2024 · similarity 0.84)
- Mimicking the Maestro: Exploring the Efficacy of a Virtual AI Teacher in Fine Motor Skill Acquisition (AAAI 2024 · similarity 0.83)
- ArtWhisperer: A Dataset for Characterizing Human-AI Interactions in Artistic Creations (ICML 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.