Exploration

  • different style of AI music generated (k-pop style)

    different style of AI music generated (k-pop style)

    by Kael 내로남불(Rules for thee)- Kael Lyrics[Verse 1](Rules for thee, not for me, that’s the way it’s meant to be)Real problems ignored, but you tell me to stay quiet, why?That’s just an excuse, your petty alibiReal problems ignored, but you tell me to shut it up, why? (Pre-Chorus)눈 감고 귀 닫고 “규정상 괜찮다” 말해Sitting in…

  • [LegAL AI] Part 2-1-2: A Quick Guide to Using Open Law Korea

    [LegAL AI] Part 2-1-2: A Quick Guide to Using Open Law Korea

    In this update, Daniel discusses his experience using the Open Law Korea API to retrieve legal data for a project. The API offers access to a variety of legal information in formats like JSON, XML, and HTML. He shares insights on API structure, including how to efficiently fetch law IDs, manage response formats, and understand…

  • Why Feeling ‘Special’ Matters: hash tag “music”

    Why Feeling ‘Special’ Matters: hash tag “music”

    by Kael I remember a short animated movie(I guess it was You are special) I watched when I was kiddo. At that time, it felt like such a powerful and comforting message. However, as I grew older, the concept of “specialness” started to feel vague. “You’re special” — phrases that can be both uplifting and…

  • Creating Video from Images: The Role of Generative AI and GPU

    Creating Video from Images: The Role of Generative AI and GPU

    Kael shares her blogging journey, highlighting her experimental approach compared to her husband’s steadiness. Despite lacking formal training in music, art, and video production, she utilizes various tools to create content. Although she feels her work is amateurish, she uploads it as a record of her learning experience, despite imperfections.

  • [LEGAL AI] Part 2-2: Structuring Legal Data for AI

    [LEGAL AI] Part 2-2: Structuring Legal Data for AI

    The project on South Korean laws focuses on continuously refining datasets. It includes core legal components necessary for AI applications like LLM fine-tuning and retrieval. The dataset, converted from JSON to a structured format and uploaded to Hugging Face, aims to enhance understanding of laws, precedents, and multilingual capabilities.