Pages
256
Year
2024
Level
beginner
Read time
7h
Ethan Mollick · WH Allen (Penguin UK) · 2024
Reviewed by Ashish Sheth · Updated April 2026
Co-Intelligence
Living and Working with AI
4.5 / 5
AMAZON · 4.0K RATINGS
ai society
SUBJECTS
What you'll come away with
01.
Always invite AI to the table — try it for every task before deciding it can't help
02.
Be the human in the loop — your judgment is the safety net
03.
Treat AI like a person, but remember it isn't one
04.
Assume this is the worst AI you'll ever use
05.
Practical mental models for delegating to LLMs without losing your own skill
06.
Why 'centaur' and 'cyborg' modes of working with AI both have their place
Strengths
+Most accessible AI book for non-technical readers — a NYT bestseller
+Practical examples grounded in actual classroom and workplace experiments
+Short and readable in a weekend
+Mollick's framing ('jagged frontier', 'four rules') has shaped how the industry talks about AI
Caveats
−Light on technical depth — software engineers may want more under-the-hood
−Some examples already feel dated as models improve fast
−Mostly observations, less actionable than the title suggests for builders
★ 4.5 FROM 4.0K READERS ON AMAZON
Check Price on Amazon →
Read this if
→Developers who want a conceptual frame before diving into AI engineering books
→Engineering managers thinking about how to roll out AI on their teams
→Anyone who feels they're behind on AI and needs a fast, smart on-ramp
Skip this if
—Engineers who already build LLM apps daily — most of this is intuition you have
—Readers wanting code, math, or implementation detail
—AI skeptics looking for a critical takedown — Mollick is broadly optimistic
Head-to-head comparisons
Co-Intelligence vs The Coming Wave → Co-Intelligence vs AI Engineering → Co-Intelligence vs The Hundred-Page Machine Learning Book → Frequently asked
Is Co-Intelligence too non-technical for software developers?
Most developers find it useful precisely because it isn't a coding book. It gives you the mental model for when AI helps and when it doesn't. Read it before AI Engineering, not instead of.
How is this different from The Coming Wave?
Co-Intelligence is practical and personal — how you, today, should use AI. The Coming Wave is strategic — what AI plus biotech does to society over decades. Read both.
Is Co-Intelligence still relevant in 2026 given how fast AI moves?
The specific model examples date quickly, but Mollick's frameworks (jagged frontier, four rules, centaur vs cyborg) have held up and are still cited across the industry.
Read this next
3 alternatives
Ready?
Check Price on Amazon →