đ§ How to Use AI Effectively: A Practical Guide for Thinkers, Writers, and Policymakers
Most people treat AI like Google. The best treat it like a research assistant, strategist, and second brain.
After months of daily use across tasksâstrategy, research, policy design, writingâIâve come to one simple conclusion: thereâs a widening gap between people who just use AI, and those who use it well. This post is about closing that gap.
âď¸ The AI Tools You Should Knowâand What Theyâre Good At
Not all models are built alike. Hereâs how I use each toolâand what I avoid them for:
đ§ How to Write Prompts That Actually Work
You donât need to âprompt engineerâ like a developer. But you do need to be deliberate. Here are 5 patterns I use every day:
đ Using AI for Deep Research: Think Like a Manager, Not a User
If you're using AI like a search engine, youâre missing 90% of its value. Hereâs what I mean by Deep Research:
Ask structured, layered questions
Request competing perspectives
Synthesize evidence across time and place
Iterate until the output teaches you something new
Example Prompt Chain:
âBreak down the drivers of opposition to carbon taxes in the US based on CCAM data.â
âNow compare this to European countries using academic studies since 2015.â
âWhat policy design tweaks could increase acceptability? Provide real-world examples.â
This is how you move from AI as toy â tool â thinking partner.
đ How the Experts Use AI (and What You Can Steal From Them)
The best users arenât engineers. Theyâre curious thinkers who experiment aggressively. Hereâs what Iâve learned from some of the sharpest minds on this:
đ§Ş Ethan Mollick
âAlways use AI twiceâonce to explore, once to refine.â
Uses GPT-4 for everything from literature reviews to prototyping slide decks.
đď¸ One Useful Thing â Substack
đ§ Tyler Cowen
Sees AI as a superforecasting partner.
Encourages intellectual augmentation, not just automation.
âď¸ David Perell
Uses AI to enhance, not replace, writing (Humans have the two Es-Experience and Expertise to enhance your writing).
Focuses on authenticity, structure, and creativity.
đ perell.com
đ Karen Hao & Eliezer Yudkowsky
Remind us to stay critical. Donât trust. Verify.
Especially vital when using AI in domains like climate, policy, or ethics.
đ ď¸ Resources You Should Bookmark Right Now
đ§Š Final Take: This Isnât About Tech. Itâs About Thinking.
AI is not a replacement for thinking. Itâs a catalyst for it.
The best users donât ask it to do their work. They ask it to challenge their assumptions, speed up iteration, and give them a fresh lens.
If youâre thoughtful, curious, and structuredâyouâll win in this AI era.
đ I'd love to hear from you:
How are you using AI in your research or work?
Whatâs workingâand whatâs confusing?
Hit reply or comment below. Iâll respond to as many as I can.
đ Feel free to share this post with colleagues or teams exploring how to use AI more rigorously.