Speed and Scale Are Dials · Using AI for Research

When researchers talk about AI, speed tends to take over. Then scale. Deployed without intention, AI becomes a shiny research assistant — useful for throughput, thin on understanding.

Choose the relationship before you open Copilot.

The useful question is not “How do I use AI in research?” It is: What relationship do I want with AI at this moment? Accelerate or add friction? Expand the field or go deeper into a defined space?

AI can compile, pattern-find, talk back, and reshape findings. It cannot decide what the work is for, own ethics of inclusion, replace fieldwork judgment, or guarantee that fluent synthesis is true.

Core principles

Widen aperture — accelerate and expand: take in more than the plan usually allows, then narrow on purpose.
Notice nuance — accelerate and deepen: surface patterns, tensions, and missing perspectives in what you have.
Challenge assumptions — slow and deepen: pressure-test plans, guides, and early insights.
Immerse in context — slow and expand: bring material to life so people can engage and act.
Before every session: Where in the process am I? What stays human? Which mode am I choosing?

Check yourself

What should you decide before opening Copilot for research?

Do this in Copilot

Pick one live research task. Write three lines: process stage, AI vs human split, starting mode. Keep that sticky for Lessons 2–5.

Paste into Copilot Chat. Answer the follow-ups, then save your three-line sticky.

Name the mode

I have a research task: [DESCRIBE IN 3–5 SENTENCES]. Ask me three short questions so we can name: (1) process stage, (2) what AI should do vs what must stay human, (3) which mode fits now — Widen aperture, Notice nuance, Challenge assumptions, or Immerse in context. Then recommend one starting mode and why. Do not draft findings yet.

Techniques: Mode selection

Open Copilot

Did you run this in Copilot? Mark complete when you have tried it.