đź’ˇ Applied AI Writing & Research Task Ideas

A list of potential articles and research notes, categorized by theme, to be developed using Claude Code for acceleration and documentation.

1. The Human-AI Workflow and Trust (Trustworthy & Transparent Systems)

Priority Task ID Topic Title Focus Status
High WFA-001 The Model Context Protocol (MCP) as the “Universal Remote” for AI Analyze the technical and user experience benefits of MCP. How does standardizing context solve problems like data silos and model hallucination in enterprise settings? To Do
High WFA-002 Augmentation vs. Automation: A Dignity-First Approach Explore the ethical and practical differences between using AI to augment a human knowledge worker versus full automation, focusing on the goal of removing tedious work. To Do
Medium WFA-003 The Necessity of “Showing Work”: Transparency in Auditable AI Write about the principle of transparency in AI. Why is it crucial for human users to see the evidence, sources, and steps an AI took to reach a high-stakes conclusion? To Do

2. The Future of Work and AI Agents (Organizational & Societal Impact)

Priority Task ID Topic Title Focus Status
High FWA-001 The Skills Gap in the Age of AI: Prompt Engineering & AI Literacy Analyze the new essential skills for workers, such as prompt engineering and AI literacy. How can companies invest in personalized training to help employees adapt? To Do
Medium FWA-002 AI Agents and Organizational Design: Managing Human-Agent Teams Discuss how organizational structures and leadership roles must evolve to manage teams composed of both human employees and increasingly autonomous AI agents. To Do
Low FWA-003 Beyond Job Replacement: The Productivity Effect and New Roles Argue how the massive productivity gains from AI lead to the creation of new industries and job roles, shifting human focus toward creativity and contextual understanding. To Do

3. Practical AI Implementation (Development & System Hardening)

Priority Task ID Topic Title Focus Status
High PIA-001 From Prototype to Production: Hardening LLM Workflows Discuss the technical challenges of moving an AI proof-of-concept into a reliable, enterprise-grade system, covering error handling, security (Zero Trust), and auditable pipelines. To Do
Medium PIA-002 The Craft of AI-Assisted Development (A Case Study) Write about the experience of building a project (like Humaine Studio itself) using an AI coding assistant. What are the trade-offs and how does the human developer’s role change? To Do
Medium PIA-003 Context Engineering for Reliable Agents Detail how structured inputs and clear context are vital for getting predictable and high-quality results from large language models, especially in complex tasks. To Do

Usage Notes

Workflow

  1. Select a topic from this list based on priority and current interest
  2. Update status to “In Progress” when you start writing
  3. Create research notes in writing-topics/notes/[TASK-ID]-notes.md
  4. Create draft in _drafts/[title].md
  5. Publish by moving to _posts/YYYY-MM-DD-[title].md
  6. Update status to “Published” when live

Research Notes Structure

For each topic, create a notes file: writing-topics/notes/WFA-001-notes.md

Example structure:

# WFA-001: MCP Research Notes

## Key Points
- [Research findings]
- [Quotes to include]
- [Examples to reference]

## Sources
- [Links]
- [Papers]
- [Projects]

## Outline Ideas
- [Potential structure]

Integration with Substack

Topic Selection Guide

When you’re energized and have focus:

When you’re low energy but want to write:

When you need a quick win:

When you’re blocked: