{
  "$type": "site.standard.document",
  "description": "I use several flavours of AI as my pocket crit. Today, Gemini is helping me synthesize a few important points. I'm microblogging them from my Obsidian vault so it's easier for colleagues to join the discussion.",
  "path": "/2026/06/02/leveraging-my-digital-critic.html",
  "publishedAt": "2026-06-02T09:38:56Z",
  "site": "at://did:plc:ki5px4zg2sa2zhmojphgtjys/site.standard.publication/3mn7r5odk2w2g",
  "textContent": "I use several flavours of AI as my pocket crit. Today, Gemini is helping me synthesize a few important points. I’m microblogging them from my Obsidian vault so it’s easier for colleagues to join the discussion. 1.The Cognitive Load of Visuals: Intro & Theory. Tie Kahneman's System 2 thinking to how humans process dense data vs. clean visualisations. Explain how AI mitigates human cognitive offloading risks. 2.The AI Video & Audio Landscape: Tools & Platforms. Introduce the practical tech stack for multimedia production (including mobile journalism workflows). Contrast surface-level content generation with deep, sustainable knowledge gathering. 3.Prompting for High-Precision Visuals: Framework Application. Show a side-by-side comparison of a lazy prompt vs. a highly structured GCAO / RACE prompt used to clean and graph complex R&D data sets. 4.Breakout Challenge: Automated Pipeline: Hands-on Exercise. We need to give clear, numbered instructions for participants to upload a technical document, ask an AI to extract key patterns, and map it into a functional workflow diagram. 5.The Sovereignty Warning: Data Security & PKM. I don’t think people realise how maintaining an air-gapped second brain offers a clear advantage when trying to keep data private. I believe in the absolute necessity of backing up prompts (they’re in my Swipes folder) and workflows locally (using markdown/Obsidian) to prevent total data dependency loss.",
  "title": "Leveraging my digital critic",
  "updatedAt": "2026-06-02T09:39:28Z"
}