Prototyping with v0
Prototype Description
A web app / resource for ʻōlelo noʻeau that allows visitors to interact with the text in a dynamic way.
Unique Problem, Curiosity, or Goal:
These text are a unique resource into lāhui consciousness of the past. These sayings help to define an ʻōiwi mindset, and give insights to different wahi. ʻŌlelo noʻeau stand out as digestible packages of ʻike that serve as a window into deeper frameworks at play.
Found in PDFs across the internet, there's only one searchable, online resource (now archived) created by a researcher at U.H. with little affiliation with the Hawaiian Studies department. One of my favorite thought exercises is imagining a timeline where Ke Aupuni Hawaiʻi hadn't the inconvenience of annexation—how would our ʻike be circulated and shared? What IP would we have been created, and what IP governance would we have stepped into? How would our meta tagging systems be organized?
Your Solution, Exploration, Thesis:
Would be neat to have a deeply searchable database for ʻōlelo noʻeau that centers around kaona, maps, and even oral histories. Something that reflects the translation happening between ʻāina > kanaka > poetry. Being provide unique context aligns with Pukui's legacy for the text. Being able to talk with a chat bot would be useful for folks who want to cite source.
Demo / Prototype
What was the process?
V0 takes in a natural language prompt, and allows you to refine it while building in real time. Free plan allows you to generate features, publish chat for team, save up to 200 projects, and deploy.
From this prototype idea, I think the biggest investment would be in getting the data and tagging it with kaona. Fun features could include giving ʻōlelo noʻeau based on IP address, and locking localized ones down. Ideally you can toggle between Hawaiian / English, too.
Findings and Reflections
What did you learn?
Part of this process took me down an OCR / ICR / LVLM rabbit hole, and got me thinking about all the audio content in ʻŌlelo, too. In college, the librarian once told me, if they aggressively cataloged audio content, it would take 12 years and many open ended full-time positions. I also imagine, there is protocol around doing this kind of thing, and wonder how to build that into a process so it presents as a framework.
How do you want to iterate?
This would be a neat project to publish on a password protected site for students, educators, and practitioners. Would love to get it up and running.
