ArtiNova.ai
An AI Image Generation Platform for Teenagers
ArtiNova is a course project. It is a creative platform for teenagers to learn, explore, and practice AI image generation.
About the Project
ArtiNova is a course project that I collaborated on with my classmates at Teachers College, Columbia University. Our inspiration stemmed from our experiences with AI image-generation platforms. We identified two major deficiencies in the existing platforms. First, some expert-level platforms are too complex for users to learn, while many amateur platforms fail to produce high-quality images consistently. Second, a significant number of these platforms may contain harmful information. It led us to consider how teenagers can experience these issues. As they are in their prime of creativity and curiosity, teenagers deserve an appropriate AI image-generation platform.
MidJourney is simple but unpredictable
ComfyUI is complex but powerful
Our design centers on creating an AI image-generation platform specifically for teenagers. This platform is designed to block harmful content while also making it easy for teens to learn about and use AI image generation to create high-quality images. We combine features from both expert and amateur platforms by visualizing the AI image generation process, allowing users to generate quality images in a more accessible and educational manner.
Our Research
The Overview of Our Research
Our research includes two steps: a literature review and a user research. In the literature review, we explored the opportunities and challenges of AI for teenagers, which inspired our idea of designing ArtiNova. In the user research, we followed the standard process: emphasizing the users with a questionnaire and interviews, building user personas, creating user scenarios, etc.
Literature Review
The Growing Generative AI Market Size
According to a report by Fortune Business Insights, the global generative AI market is projected to grow from USD 67.18 billion in 2024 to USD 967.65 billion by 2032. North America dominated the market, with a share of 49.78% in 2023. The market in the U.S. is projected to grow significantly, reaching an estimated value of USD 220.27 billion by 2032.
Educational Potential
Creative Expression
Career Development
Misinformation
VCSAM
Content title
Content
Exploit the Opportunities
- Teach the key process and interaction of AI image generation, helping teenagers understand the underlying logic
- Provide teenagers with an inspiration library, triggering their creativity when learning AI
- Help teenagers understand the basic concept of AI, offering them a chance to further explore their career in AI
Overcome the Challenges
- Introduce teenagers to deepfake knowledge, helping them better understand AIGC and become less susceptible to misinformation
- Provide teenagers with a secure environment without the existence of any harmful information and minimize the risk of VCSAM
User Research
An excerpt from the questionnaire we used to research teenagers' interests, experiences, and expectations of AI image generation:
- How interested are you in learning how AI can assist with creative work?
- How familiar are you with AI-generated creative content, such as images or music?
- If you had the opportunity to use an AI image generator, what kind of content would you want to create?
- What do you most want to achieve with an AI image generation platform?
- If you were to use a platform that teaches AI image generation, which features would you consider most important?
- How much time are you willing to spend learning the basics of AI image generation?
Some Important Data
77% of the subjects show interested in how AI can assist in creative tasks.
An Empathy Map
54% of the subjects prefer generating school-related content using AI.
38% favor artistic creations.
38% want to generate social media content.
54% of the subjects want to use the platform to learn new skills.
46% want to realize their own artistic thoughts.
38% wish to ease the process of artistic creation.
69% of the subjects prefer ready-made templates for ease of use.
61% need detailed tutorials.
46% want customized features for personalized creation.
An excerpt from the interview we used to research teenagers' interests, experiences, and expectations of AI image generation:
- Have you used any AI tools to create content (like AI art)? If yes, what did you think of it?
- What do you enjoy most about these kinds of AI tools you use?
- Have you used any AI tools to generate images? If yes, what motivated you to use it? If no, what prevented you from using it?
- What features would you want in an AI image app (such as ready-made templates for easy image creation, style choices, editing options, inspiration library，adjustable settings to customize images, an option to share creations on social media, tips and tricks for improving image quality)?
- How would you mostly use this app?
- Would you prefer to learn how AI image generation works and understand the fundamental structures behind it, or would you rather just use it?
- What challenges have you experienced with AI image generators?
- Would you like to share your work in a community with other teens?
SAYS
- “I use AI to generate images that can be used in my schoolwork, such as PowerPoint presentations.”
- “It is difficult to use AI to generate storytelling pictures.”
- “Some AI image generation platforms, like Civitai, have more powerful features, but learning how to use them takes time.”
- “If I can generate some very interesting images, I also want to share it on social platforms, such as RedNote.”
THINKS
- Has a need to use AI image generators in schoolwork.
- Wants an AI image generator that has a better ability to generate storytelling images.
- It is a little bit difficult to use some more expert AI image generators.
- Wants to share interesting AI works on social media.
- Wants to get AI image generation prompts more conveniently instead of asking ChatGPT.
DOES
- Uses AI to generate images for schoolwork, such as in PowerPoint presentations.
- Uses ChatGPT for most of the time and is not very familiar with other AI platforms.
FEELS
- Hesitate to try other AI image generators other than ChatGPT.
- Surprised by some high-quality images generated by expert AI image generators.
- Feels it inconvenient to use GPT to generate prompts and then use them in other AI image generators
- Hindered by the price need to pay to generate more images in expert AI image generators
Personas
Emily 17yrs
High school student
Goals
Quickly generate fun and creative content in different artistic styles.
Share creations with peers or on social media to gain recognition and inspiration.
Goals
Using AI-generated images in PowerPoints to complete her homework.
Frustrations
Sometimes, it is difficult to find appropriate images for PowerPoints.
It is inconvenient to use ChatGPT to generate prompts before using image-generation AI.
Emily is a student at an international high school. While she is not very familiar with image-generation AI, she occasionally uses some AI tools to create images for her homework, such as in PowerPoint presentations. She believes that AI-generated images are often unnecessary for her, but she is willing to learn more about image-generation AI if it can produce high-quality images that she can use for her assignments and share on social media platforms like RedNote. The three most important features of an image-generation AI, according to her, are the Inspiration Library, Editing Options, and Style Choices.
Frustrations
Generated images sometimes lack precision and don’t meet expectations.
Needs clear guidance.
Alex likes technology and creativity. He loves experimenting with innovative tools and learning how they work. In his free time, he enjoys using AI tools to generate imaginative and entertaining content. He often inputs humorous or random prompts to generate creative images.Sharing creations with friends and posting them on social media is one of his favorite activities. Alex is also a big fan of manga and science fiction. Alex often uses AI tools to bring the scenes he imagines while reading science fiction novels to life. The process excites him because it feels like stepping into his favorite stories and making them real. He loves to change the details, like the colors of the sky or the design of the buildings, to match the images in his mind.
A Scenario
Alex 16yrs
High school student
Our Design
An Overview of Our Design
Based on our research, we defined the main components of our platform, which include four parts: a “Create” section, an “Inspiration Library”, a tutorial, and a community.
Inspiration Library
The "Inspiration Library" is a resource for users seeking inspiration for their creative projects. It features a diverse collection of inspirational images that can be filtered by themes such as fantasy, sci-fi, and surrealism. Additionally, the images are categorized by their functions, which include design ideas, creative challenges, trending topics, etc.
The Demo of the MVP Version of ArtiNova
Create
The “Create” section embodies the major function of our platform: image generation. Users can easily generate images by dragging base models and prompts from the sidebar to the main screen. It not only makes the process more convenient with pre-written prompts but also helps teenagers gain knowledge of the basic logic of AI image generation.
Tutorial
After signing up, users are recommended to finish a stater tutorial. The tutorial is a step-by-step guide that uses a gamified learning style to motivate teenagers. It helps users gain a basic understanding of this platform's workflow.
My Role
My major role in this project is leading the team in conducting research, including a literature review and user research.
I organized regular meetings and collaborated with my team members to compose the literature review. I also coordinated the team's efforts to issue a questionnaire and conduct interviews.
Additionally, I collaborated with my team members during the design process to create the prototype of our platform. At the conclusion of this project, I wrote a design rationale to summarize our work and justify our design choices.