Topics

Incident and accountability infrastructure for AI-agent-caused production damage — vendor postmortems (or their absence), liability frameworks, forensic audit trail standards, and governance/verification practices that prevent or respond to agent-caused harm. Distinct from ai-code-quality’s focus on code correctness/maintainability — this is about what happens when an agent does damage, and whether the industry has infrastructure to catch, attribute, and learn from it.

Big-picture architectural guidance for AI-assisted and AI-generated systems — system design, module boundaries, abstraction choices, when to let AI make structural decisions vs. human-owned architecture, and patterns for keeping AI-generated systems architecturally coherent as they scale. Emphasis on Claude Code and multi-agent systems, but covers AI coding assistants generally. Not line-level code quality — the macro-structure question.

The code review act itself where AI is involved on either side — AI-powered review tools and agents (automated PR review, LLM-based review bots), and the human practice of reviewing AI-generated code (what review workflows change, what reviewers systematically miss, review fatigue, and emerging review-specific tooling/standards). Distinct from ai-code-quality’s broader quality/maintainability scope, of which review is one practice among several.

The societal impact of AI — employment displacement, regulatory moves, public sentiment, the doom/acceleration debate, and institutional responses. The goal is mood capture and zeitgeist, not comprehensive reporting. What are people worried about? What’s actually happening? What’s the gap between fear and reality? Prioritise data-backed analysis and institutional reports over opinion pieces, but include opinion when it captures genuine public mood.

Concrete, real-world applications of AI coding in organisations — legacy system modernisation, citizen developer programmes, non-technical users building apps, enterprise adoption patterns, and governance challenges. The focus is on what organisations are actually doing with AI coding, not what tools exist. Case studies, adoption data, and institutional reports over product announcements.

Tools, plugins, and applications that integrate Claude (via the API or SDK) into domain-specific software: design tools, productivity apps, developer tooling, creative software, and specialised platforms. Focus on what’s being built with Claude, not on Claude Code itself.

Learnings, tips, behavioural approaches, and usage patterns for Claude Code (the CLI tool), Claude API, CLAUDE.md authoring, agent workflows, hooks, skills, and the broader Claude development ecosystem. Focus on practical techniques and real-world usage over announcements.

The legal and ethical battles over AI training data — copyright infringement lawsuits, fair use debates, opt-out mechanisms, synthetic data as an alternative, data licensing markets, and regulatory responses. This is foundational infrastructure: how these battles resolve will reshape what models can be trained on and who can train them. Focus on legal developments, regulatory proposals, and substantive analysis over opinion pieces.

A broad range of geopolitical, economic, demographic, and military analysis and forecasting — of the kind espoused by Peter Zeihan and Sarah Paine (tracked separately as Creators), but deliberately wider. The goal is a range of opinion on how the world is actually going, not confirmation of any single framework. This topic is a known filter-bubble risk, so it actively seeks out mainstream, skeptical, and contrarian counter-views alongside structural-collapse framings. Covers deglobalization, demographic decline, great-power conflict risk, trade wars, sovereign debt and currency shifts, and resource security.

The evolving conflict and interplay between open-source AI models (LLaMA, Mistral, DeepSeek, Qwen) and closed-source models (Anthropic, OpenAI, Google). Covers safety implications of open weights, licensing debates, competitive dynamics, access and democratisation arguments, innovation pace differences, and the regulatory dimension. Focus on substantive analysis of tradeoffs, not cheerleading for either side.

How teams and organisations adopt Claude collectively — shared CLAUDE.md conventions, hooks and skills at team scale, enterprise deployment patterns, coordination norms, productivity measurement, and multi-person workflow case studies. Focus is on org-level patterns and friction rather than individual technique (see claude-expertise for that).

The evolving landscape of AI-assisted “vibe coding” — techniques, tools, frameworks, and methodology. Includes IDE-based tools (Cursor, Windsurf, Copilot), agent frameworks (LangGraph, CrewAI, AutoGen), and emerging practices like spec coding, multi-agent orchestration, and prompt-driven development. Focus on genuine technique over tool roundups and marketing content.

Techniques, tools, and practices for producing and maintaining high-quality AI-generated code — correctness, maintainability, test coverage, review process, and preventing quality decay over time. Emphasis on Claude Code and Python, but covers AI coding assistants generally. Focus on concrete practice and evidence (benchmarks, case studies, postmortems) over tool marketing.