2026-02-13 Fri
- AI/Public/Sentiment
- AI has become the 2nd most divisive Rorschach test in public opinion, just behind the political elephant in the room.
- Stanford puts out an annual study of public sentiment about AI (AI Index 2025: State of AI in 10 Charts | Stanford HAI) and the US ranks higher in nervousness and lower in excitement about AI than other countries. Other surveys show similar results (Google Ipsos poll says Americans wary of AI, Views of AI Around the World | Pew Research Center).
- Whenever I go to talks that prominent individuals give, it seems like the vast majority of the Q&A is basically "holding people of power to account for what they're doing to us."
- AI has become the trendiest scapegoat in the room, and tech absolutely, 100% needs to address this head on. I recently heard this take:
2020: blue and tech against red
2024: red and tech against blue
2028: blue and red against tech
- AI/Coding
- Claude/Code
- MCP
- Discovered via
claude mcp --helpithatclaude mcp serveserves an MCP server for claude itself. Woah. Inception Recursion. It's not the first time I've heard of this, I think. There was a talk at OReilly/Event/25/09/AI Codecon that involved someone using MCP to control AI coding agents to use different models for comparing how they coded. Even so, I didn't know that it's now supported by Claude Code natively. - Updated
- Researched auth reuse pattern for optional Atlassian/MCP sessions:
- Shared auth is feasible via mcp-remote token cache (
~/.mcp-auth) when URL/resource/headers are identical across sessions
- Shared auth is feasible via mcp-remote token cache (
- Reviewed related existing KB notes:
- Filed
- Discovered via
- MCP
- Claude/Code
- CLI/Craftsmanship