AI for Busy Professionals:
The Useful Mental Model.
By James Chia·16 Sep 2026·⏱️ 90s Read
Most professionals fail to get consistent value from AI because they adopt the wrong mental model. They either treat it as an all-knowing oracle, or discard it as an untrustworthy toy after their first hallucination.
The most reliable framework is simpler: treat an LLM as an exceptionally fast, probabilistic intern who has read the internet, but has zero institutional context and zero personal accountability.
The Three Rules of the "Intern" Framework
- 1. Never ask it to remember:Always supply the facts, data, or source text in the prompt. If you don't provide the briefing packet, it will politely invent what sounds plausible.
- 2. Assign tasks, not broad goals:Instead of "analyze our operations," ask it to "extract all dates and dollar amounts from these three vendor contracts into a JSON table."
- 3. You sign off on the work:You wouldn't send an intern's first draft straight to the board without reviewing it. Never wire an LLM output directly into customer or financial actions without an approval gate.
Where the Leverage Actually Lives
The real productivity breakthrough isn't asking conversational questions; it's using models for transformation, synthesis, and scaffolding:
High Leverage (Use AI)Summarizing meeting transcripts, drafting test cases, writing schema boilerplate, comparing two complex documents.
Low Leverage (Avoid AI)Deterministic calculations, critical compliance decisions, database transactions, blind trade executions.
When you give the model the right briefing materials and hold yourself responsible for the final output, AI stops being a novelty and becomes the highest-leverage amplifier in your daily workflow.