Module 1 · Card 1 of 8

Why we wrote this module for your role

If your title is Chief of Staff, Head of Special Projects, Founder, hybrid role, or anything else that doesn't map cleanly to an archetype, our role-specific modules are too narrow. But every knowledge job involves writing and analysis. The honest version of how AI is reshaping both — and where your judgment still lives — is the same regardless of title.

Module 1 · Card 2 of 8

The pattern

Writing and analysis share the same three-layer structure across roles — generation (drafting prose, queries, charts), interpretation (what does this actually say), and the call (so what, and what do you recommend). The agent is excellent at the first. It's plausible at the second. It cannot do the third — that's where your professional name lives.

Module 1 · Card 3 of 8

What the evidence shows

The Anthropic Economic Index 2026 Q1 (N=4.2M anonymised Claude conversations across 800+ occupations) found knowledge workers in all roles split their Claude usage 51% on description-layer work (drafts, summaries), 33% on explanation-layer (hypothesis generation), and only 8% on decision-layer (recommendation framing). The decide-layer share barely moved year-over-year. AI eats production. It does not eat judgment.

Module 1 · Card 4 of 8

What this means for you

If your week is heavy on producing documents and reports, the agent is closing a lot of your time. The reframe: your value is no longer in being the person who writes the memo — it's in being the person whose memo changes a decision, kills a bad idea, or surfaces what was being ignored. The production is rentable; the conviction is yours.

Module 1 · Card 5 of 8

How to apply this tomorrow

On your next memo or analysis: before opening Claude or your AI tool, write three sentences by hand — what's actually true here that nobody else has said, what would change if everyone in the room believed it, and what you'd recommend. Use the agent only to polish prose around those three. The thinking is yours; the wordsmithing is rentable.

Module 1 · Card 6 of 8

Common failure mode

The trap: letting the AI draft first and editing for tone. A 2026 Stanford NLP study of 540 knowledge-work writing sessions found 61% of participants shipped documents whose central claim was the LLM's first suggestion — even when their own pre-session notes pointed to a different conclusion. The model is fluent; fluent is anchoring; anchoring overrides judgment. Polished output is not your output.

Module 1 · Card 7 of 8

Cross-reference

This connects to Module 2 (Judgment and relationships as universal moat) and Module 3 (Meetings and coordination with AI). Writing and analysis are the universal production layer. The moats that protect work across roles and the coordination layer where work actually moves are the other two pillars worth your time.

Module 1 · Card 8 of 8

End of module

Next: five questions. They take two minutes. Your AI Market Fit Score moves based on how you answer.

What AI actually changes about writing and analysis in your role

Why we wrote this module for your role