(2-month internship)
Project Overview
This two-month internship supports the development of a local interactive artwork, where users respond to philosophical questions about AI. Their answers generate both AI-created imagery and commentary, integrated into a dynamic, surreal digital collage inspired by The Garden of Earthly Delights by Hieronymus Bosch.
The goal is to complete a working local pipeline that updates a composite image in real time using web technologies, Python scripting, and external AI APIs.
Your Tasks
1. Frontend – HTML Interface
You will improve an existing local web page that:
- Introduces the artwork and asks ethical questions (text provided).
- Collects answers and sends them to a Python backend.
- Displays both the original and updated image, plus AI-generated commentary.
What needs to be done:
- Connect the site to the chosen AI API (at no cost incurred by the student).
- Optimise usability and performance.
- (Bonus) Optimise usability and performance for a mobile version of the website
2. Backend – Python Server & API Integration (Most Important Task)
A lightweight Flask server is already in place. You will enhance it to:
- Construct prompts from user input and send them to an AI service (e.g. OpenAI).
- Receive and store:
- a Bosch-style generated image,
- a short ethical/philosophical comment.
- Update the correct section of a composite image using Pillow.
Enhancement goals:
- Benchmark analysis to find the best-suited AI API to generate real images and text dynamically.
- Craft and use the API to generate visuals and commentary as instructed
- Ensure seamless display of updated visuals and commentary in the website.
- Add clear, readable code comments in English.
- (If needed) Debug and maintain responsiveness.
3. (Optional 3rd Task in case of Thesis Work) Backend – Integration with Pre-Existing SoMax Architecture for Music Generation
We are working to expand this project to Conservatory students in Electronic Music, who will create an interactive music experience that responds to the generated images. In the case of thesis work, the work of the student could be greatly enriched by the single and straightfoward task of supporting the integration of the music experience. To do so, it is just required that, once the AI image is generated, it is also automatically sent to the software used by the Electronics Music students to translate the image into sound.
What needs to be done:
- Ensure that the generated image is also uploaded to the software used by the Electronic Music students.
- (If needed) Debug and maintain responsiveness.
*** Required Skills
- Basic HTML
- Python: file handling, Flask (REST APIs), image processing with Pillow.
- (Bonus) Basic UI/UX knowledge for smoother user interaction and aesthetics.
- (Bonus) Passion for art and immersive art exhibitions
*** Expected Outcomes
- A fully working local interface connected to an AI API.
- A reliable script for real-time image composition and commentary display.
- A local system functioning offline (except for API calls).
- Logging of user responses in .json or .csv.
*** Final Deliverables
- Project Folder with:
- HTML/CSS/JS frontend
- Python backend & image processing scripts
- Optional documentation
- Technical Report (potentially constituting the basis for a thesis project) covering:
- System architecture and logic
- Image update script design
- Design choices and test results
Timeline
The enhancements described above should be performed within a 2-months timeframe, from early April to early June, due to the immersive art experience being scheduled to take place in late June/early July.
Highly motivated students with different time constraints are invited to reach out to the post author (Chiara Natali) to work out possible different internship configurations (e.g., only taking care of Backend – Python Server & API Integration). However, such limited-form internships cannot be a basis for thesis work on this topic.