Quick Picks
-
Best first read Co-Intelligence The shortest book here at 256 pages, built on Wharton experiments with GPT-4. The clearest starting point for any PM new to GenAI. -
Best for AI economics Power and Prediction Rated 4.22 on Goodreads, with a system-redesign argument built from 250+ AI startups inside Toronto's Creative Destruction Lab.
-
Best for PM craft Building AI-Powered Products The only GenAI PM handbook from a frontier-lab product lead: Nika ran GenAI at Google and Meta before writing this O'Reilly playbook.
Which AI Book Should a Product Manager Read First?
Co-Intelligence by Ethan Mollick should be first: at 256 pages it is the fastest path to understanding how generative AI works and where it breaks, which makes every other book on this list more useful. The nine picks span 2020 to 2025, with Goodreads ratings from 3.26 to 4.38 and page counts from 227 to 496. Three books in Section 3 are written for product managers; the other six are broader AI reads that PMs recommend and use widely.
| Title | Goodreads | Best For | Pages |
|---|---|---|---|
| 01Co-Intelligence (2024) | 3.93 | AI literacy primer | 256 |
| 02The Coming Wave (2023) | 3.80 | AI governance | 352 |
| 03AI Snake Oil (2024) | 3.90 | Vendor evaluation | 360 |
| 04Power and Prediction (2022) | 4.22 | AI economics | 288 |
| 05Empire of AI (2025) | 4.02 | Supply chain | 496 |
| 06Competing in the Age of AI (2020) | 3.81 | Operating models | 288 |
| 07Building AI-Powered Products (2025) | 3.26 | GenAI PM craft | 227 |
| 08The Art of AI Product Development (2025) | 4.38 | Build framework | 368 |
| 09The AI Playbook (2024) | 3.98 | ML deployment | 256 |
What Are the Best AI Books for Product Managers?
The best AI books for product managers are Co-Intelligence by Ethan Mollick (3.93), The Coming Wave by Mustafa Suleyman (3.80), and AI Snake Oil by Arvind Narayanan and Sayash Kapoor (3.90). Suleyman co-founded DeepMind; Narayanan and Kapoor made TIME's 2024 AI 100. None are PM-specific: product managers need AI literacy before PM tools.
1. Co-Intelligence: Living and Working with AI (2024)
The 2024 New York Times bestseller sets out four rules for treating generative AI as a collaborator. Ethan Mollick, associate professor at the Wharton School and co-director of Wharton's Generative AI Labs, organises the book around those rules: always invite AI to the table, be the human in the loop, give AI a persona, and assume this is the worst AI you will ever use. From there he works through four roles for AI at work: coworker, tutor, expert, and future self.
The most useful concept for PMs is the "jagged frontier": Mollick's term for the uneven boundary between tasks AI handles well and tasks where it fails quietly. A PM evaluating where to ship a GenAI feature needs exactly this mental model, because the frontier is not a clean line. Mollick also distinguishes "centaur" mode (human and AI working on separate subtasks) from "cyborg" mode (human and AI interleaved within the same task), which gives product teams a language for designing human-in-the-loop workflows.
Compared with AI Snake Oil (#3 on this list), Co-Intelligence is warmer toward AI and more focused on getting practical value from it. Narayanan and Kapoor spend most of their pages dismantling predictive AI claims and auditing failure cases; Mollick spends his running experiments and showing where GPT-4 outperforms the baseline. PMs who read both get a useful tension: Co-Intelligence teaches you to use AI well, and AI Snake Oil teaches you to evaluate AI claims honestly. Together they form the strongest pair on this list.
Read this first if you are a mid-career product manager who needs a short, research-tested introduction to working with generative AI. At 256 pages it is the fastest read here. If you already use ChatGPT daily and want a book that goes deeper on shipping GenAI features, skip to entry #7, Building AI-Powered Products by Marily Nika, which picks up where Mollick's practical advice leaves off. Named a Financial Times and Economist Best Book of 2024.
2. The Coming Wave: Technology, Power, and the Twenty-first Century's Greatest Dilemma (2023)
What it's about: Mustafa Suleyman, co-founder of DeepMind and now CEO of Microsoft AI, argues that AI and synthetic biology together constitute a wave that nation-states cannot contain by default. Written with Michael Bhaskar, the book defines the containment problem, lays out asymmetric-capability risks, and proposes a ten-step plan covering audits, choke points, and governance mandates. Published 2023 by Crown (Penguin Random House).
The Guardian, August 2023"Candidly addresses questions that are generally not discussed in polite society."
Aiifi's Take: No other author here has Suleyman's vantage on where AI is heading: co-founder of DeepMind, now head of Microsoft AI. The ten-step containment plan gives senior PMs a concrete governance vocabulary for enterprise risk conversations. Suleyman is a builder warning about other builders, so the critique has structural limits. Pair with AI Snake Oil (#3) for fully independent evaluation. A Sunday Times bestseller and FT Business Book of the Year shortlistee.
3. AI Snake Oil: What Artificial Intelligence Can Do, What It Can't, and How to Tell the Difference (2024)
What it's about: Across 360 pages, Princeton computer scientists Arvind Narayanan and Sayash Kapoor split AI into two categories: predictive AI, where most failures sit (hiring algorithms, recidivism models, the Epic Sepsis score), and generative AI, where the working products sit. The book audits each category's claims separately, drawing on the Allegheny Family Screening Tool case study and the authors' widely cited reproducibility research.
Aiifi's Take: The clearest framework for evaluating AI vendor claims. Narayanan and Kapoor give PMs a defensible filter: the distinction between predictive and generative AI, applied to real failures, is the single most useful diagnostic for a pitch meeting. The academic tone and predictive-AI emphasis mean PMs working mostly on GenAI will skim half the book. Named a Nature Best Book and Forbes Must-Read of 2024, and finalist for the PROSE Award in Computing and Information Sciences.
Which Books Help Product Managers Understand the AI Economy?
The best AI economy books for product managers are Power and Prediction by Agrawal, Gans, and Goldfarb (4.22), Empire of AI by Karen Hao (4.02), and Competing in the Age of AI by Iansiti and Lakhani (3.81). Hao won the 2026 NBCC Award for Nonfiction. They cover prediction economics, supply chains, and the AI-factory operating model.
4. Power and Prediction: The Disruptive Economics of Artificial Intelligence (2022)
A prediction-economics framework for understanding why most AI investments disappoint. Ajay Agrawal, Joshua Gans, and Avi Goldfarb, three economists at the University of Toronto's Rotman School of Management, argue that real disruption from AI arrives only when organisations redesign entire systems around cheaper prediction, not when they bolt a model onto an existing workflow. The book builds on the trio's earlier Prediction Machines (2018) and extends the argument into what they call the "Between Times": the gap between AI as a feature and AI as a system.
Where this book earns its keep for PMs is the system-redesign analysis. The authors walk through a hospital triage case where the entire patient-flow system is restructured around an AI prediction engine, not layered on the old triage protocol. They contrast this with insurance underwriting, where decision rules change when prediction costs drop. For a product manager, that is the difference between "add an AI feature" and "rethink the workflow the feature sits inside." The trio's Creative Destruction Lab, which has launched over 250 AI startups, supplies the empirical base behind the cases.
Forbes, December 2022"Definitely a book you should read."
Compared with Competing in the Age of AI (#6), which makes a similar "AI transforms the operating model" argument, Power and Prediction is sharper on the economics but narrower on implementation. Iansiti and Lakhani's AI factory model gives you the four components of an AI-driven operating model; Agrawal, Gans, and Goldfarb give you the economic logic for why and when that model pays off. PMs who read both get the strategy layer (when to invest) from Power and Prediction and the structure layer (how to build it) from Competing.
Read this if you are a product leader or senior PM evaluating where AI changes pricing, decisions, or workflows rather than where it replaces a single task. The economic frame is rigorous but abstract in places, so PMs who want concrete prompts and worked features should start with entry #7, Building AI-Powered Products, and come back to this for the strategic case. At 288 pages with a clean three-part structure, it is a reasonable one-weekend read.
5. Empire of AI: Dreams and Nightmares in Sam Altman's OpenAI (2025)
What it's about: The 2026 National Book Critics Circle Award winner draws on 260 interviews with current and former OpenAI employees and supply-chain workers in Kenya, Chile, and Uruguay. Karen Hao, formerly senior AI editor at MIT Technology Review, traces OpenAI's founding tensions, Sam Altman's leadership, the November 2023 board crisis, and the labour and environmental costs behind ChatGPT's training data.
Aiifi's Take: Hao reports from inside the supply chain most PMs never see beneath the API layer. Her Kenyan content-moderator data and Chilean datacenter water-stress reporting give product managers a fact base for vendor due diligence that no whitepaper supplies. The book won the 2026 NBCC Award for Nonfiction and the Helen Bernstein Book Award for journalism. At 496 pages and tight on OpenAI, it under-covers Anthropic, Google DeepMind, and Meta; PMs on those stacks will need to supplement.
6. Competing in the Age of AI: Strategy and Leadership When Algorithms and Networks Run the World (2020)
What it's about: Based on a decade of embedded research at Amazon, Microsoft, and Ant Financial, Harvard Business School professors Marco Iansiti and Karim R. Lakhani define the "AI factory": a four-element operating model (data pipeline, experimentation platform, algorithm development, software infrastructure) that lets digital-native firms escape the traditional limits on scale, scope, and learning. The argument was the Harvard Business Review's January 2020 cover story.
Aiifi's Take: The corporate-operating-model case for AI, and the most useful pick here for PMs inside large organisations. The AI factory model gives product leaders a shared frame for system-level AI planning. Examples skew toward tech giants (Amazon, Ant Financial), so PMs at smaller companies will translate down. Published January 2020, the argument predates ChatGPT, but McKinsey's 2025 State of AI data confirms the operating-model frame: AI high performers redesign workflows rather than bolt on features.
Which Books Are Written Specifically for AI Product Managers?
The best books written for AI product managers are Building AI-Powered Products by Marily Nika (3.26), The Art of AI Product Development by Janna Lipenkova (4.38), and The AI Playbook by Eric Siegel (3.98). Nika's is from O'Reilly (2025); Siegel's from MIT Press. These three are PM-specific; the other six build AI literacy and economic thinking.
7. Building AI-Powered Products: The Essential Guide to AI and GenAI Product Management (2025)
A GenAI product management playbook covering the full lifecycle from opportunity scoping through evaluation design, prompt engineering, agentic workflows, rollout, and measurement. Marily Nika, who spent a decade as GenAI Product Lead at Google and Meta, writes the book O'Reilly clearly intended as the first end-to-end handbook for PMs shipping AI features. The central argument: traditional PM tools (PRDs, roadmaps, A/B tests) break in non-deterministic systems, and GenAI product work needs its own playbook.
The evaluation-design chapter is the most tactically useful section for PMs. Nika walks through offline evals, golden datasets, and human-in-the-loop scoring, the three layers a PM needs to understand even if the ML team builds them. She also covers agentic product patterns: where multi-step AI workflows work, where they break, and what guardrails a PM should insist on before launch. The GenAI PM lifecycle framework that organises the book is practical rather than theoretical, but the field moves fast enough that specific model recommendations will date.
Compared with The Art of AI Product Development (#8), which organises its argument as a three-phase discovery-development-adoption framework, Building AI-Powered Products is narrower on GenAI and wider on tactical detail. Lipenkova covers both predictive and generative AI product work and spends more pages on adoption and change management; Nika goes deeper on GenAI-specific mechanics like prompt engineering as product surface and evaluation design. PMs building GenAI features today will reach for Nika first; PMs building their first AI product of any kind may prefer Lipenkova's broader scaffolding.
Read this if you are a product manager actively scoping or shipping GenAI features and want a structured playbook from a product lead who has shipped them. The Goodreads sample is small at 80 ratings and some reviewers flag the price-to-depth ratio, so set expectations for a survey of a fast-moving field rather than a deep dive on any single phase. If your AI work is predictive rather than generative, start with The AI Playbook (#9) instead.
8. The Art of AI Product Development: Delivering Business Value with AI (2025)
What it's about: Janna Lipenkova's three-phase model organises AI product work as discovery (opportunity selection), development (predictive AI, LLMs, RAG, agents), and adoption (UX, governance, change management). Based on 15 years of cross-industry AI consulting at clients including BMW, Lufthansa, and Volkswagen, the Manning title covers both predictive and generative AI as distinct product-design problems. Published July 2025.
Aiifi's Take: Lipenkova maps the full AI product arc more cleanly than any other book here. Where Nika gives you GenAI-specific mechanics, Lipenkova gives you the structure that wraps around any AI product: opportunity selection, development, and organisational adoption. The Goodreads sample is small at 24 ratings, so the 4.38 figure is more signal than confirmation. Best for PMs and founders who want a structured plan before reaching for a stack-specific handbook. A Manning lead title, distributed by Simon & Schuster.
9. The AI Playbook: Mastering the Rare Art of Machine Learning Deployment (2024)
What it's about: Published by MIT Press in February 2024 as part of the MIT Sloan Management Review series, this deployment guide formalises a six-step framework called bizML. Eric Siegel, a former Columbia University computer science faculty member, starts from the business deployment goal and works backward to the data and model. The six steps cover goal-setting, prediction alignment, evaluation metrics, data preparation, training, and deployment.
Aiifi's Take: The strongest deployment guide for predictive AI on this list. Siegel's bizML framework names the step most ML projects skip: defining the deployment goal before anyone touches data. PMs working on churn prediction, fraud detection, or recommendation systems will find a process that holds together in conversations with non-technical stakeholders. The process can read as a rebrand of common ML-project hygiene for readers already fluent in the space. Backed by MIT Press's Sloan Management Review series, which gives it weight in enterprise conversations.
How We Chose These Books
I evaluated 20 books published between 2018 and 2025, drawing from Goodreads aggregate ratings, the Financial Times and Economist Best Books lists, Nature's Best Science Books, Forbes Business Books of the Year, and product-manager reading lists. Every candidate was checked against the criteria below before the final ranking was set. This is an editorial ranking, not a formula or a score-sorted list.
Market context
- 98% of product managers in a 2025 General Assembly survey (AI is everywhere in product management, but skills gaps and "shadow AI" are holding teams back, 117 respondents) said they use AI at work, but only 39% said they had received what the survey called "comprehensive, job-specific AI training," which makes a PM-specific reading list a workplace need rather than an optional upgrade.
- 62% of organisations are at least experimenting with AI agents, and AI high performers are most likely to redesign workflows around AI rather than bolt it onto existing ones, according to McKinsey's The state of AI in 2025 (1,993 respondents), which puts product managers at the centre of system-level AI decisions, not just feature ones.
- 39% of workers' core skills are expected to change by 2030, with AI and big-data skills among the fastest-growing of all skill categories, per the World Economic Forum's Future of Jobs Report 2025, which raises the value of books that build durable AI thinking for PMs rather than vendor-specific tool tips.
The final 9 picks were chosen against 4 criteria, applied in this order:
- Topic centrality (knockout): primarily about artificial intelligence in a way directly useful to product management work, covering AI literacy, strategy, economics, product development craft, or evaluating AI claims. General PM books with incidental AI chapters and adjacent-but-not-AI books (data science, classical statistics, generic management) were excluded.
- Audience fit (knockout): accessible to non-technical product managers. No required coding, no graduate-level math, no MLOps internals as prerequisite reading. Books written primarily for ML engineers and data scientists were excluded even when highly rated.
- Quality signals (ranking): ranked by a mix of Goodreads rating and review count, awards (Nature Best Book, Forbes Business Books of the Year, FT Best Books, PROSE Award, Sunday Times bestseller), and reviews from major outlets. Books below 3.20 on Goodreads were not promoted unless they uniquely filled a section niche.
- Freshness (tie-break): post-2022 prioritised. Pre-2022 included only when differentiated by author authority and a still-cited framework (Competing in the Age of AI from HBS Press, with the durable AI factory operating model). Pre-LLM books with format-driven dating risk were pushed to Considered regardless of rating.
The numbers behind the list: lowest Goodreads rating 3.26, highest 4.22, 5 of 9 are award winners, 7 of 9 published in 2023 or later.
We excluded 3 categories: engineer-targeted ML systems books (Designing Machine Learning Systems and AI Engineering by Chip Huyen, Hands-On Machine Learning by Aurélien Géron), pre-LLM AI strategy books (Human + Machine by Paul Daugherty and H. James Wilson, The AI Advantage by Thomas H. Davenport, AI Superpowers by Kai-Fu Lee), and weak-quality or distribution-limited PM handbooks (The AI Product Manager's Handbook at 2.88, Reimagined at 3.27 and Kindle-only). Books with Goodreads ratings below 3.20 were excluded regardless of topic centrality. The full list of 11 well-known books we considered but did not include sits in the next section.
This page is editorially independent. No item is paid, sponsored, or included as part of any commercial relationship.
Who should skip this book list
ML engineers and data scientists who want hands-on model-building, training-pipeline, or MLOps content should skip this list and read our AI agents books list or Chip Huyen's Designing Machine Learning Systems instead. The audience-fit knockout means no book here goes deep on model architecture, feature engineering, or deployment infrastructure, which is exactly what an engineer needs and exactly what a non-technical PM does not.
Books We Considered but Did Not Include
These 11 books appear regularly on reading lists, recommendation threads, and product-manager book roundups. Each was reviewed against the four criteria above and excluded for a specific reason, listed here so readers can decide for themselves whether the exclusion fits their needs.
- The AI-First Company by Ash Fontana (2021): quality signal weak (Goodreads 3.30 on 101 ratings) and the strategy ground is covered more rigorously by Power and Prediction.
- Human + Machine: Reimagining Work in the Age of AI by Paul Daugherty and H. James Wilson (2018): pre-LLM (Goodreads 3.62 on 996 ratings); the original "missing middle" framing predates the GenAI shift that current PMs face.
- AI 2041: Ten Visions for Our Future by Kai-Fu Lee and Chen Qiufan (2021): strong reception (Goodreads 3.82 on 5,786 ratings) but the narrative-fiction-plus-commentary format is most exposed to dating; a 2021 short story imagining 2041 AI already reads as conservative on LLMs.
- Prediction Machines: The Simple Economics of Artificial Intelligence by Ajay Agrawal, Joshua Gans, and Avi Goldfarb (2018; Updated 2022): foundational frame for prediction economics but superseded by the same trio's Power and Prediction, which extends the argument into system-level redesign.
- Designing Machine Learning Systems by Chip Huyen (2022): excellent reception (Goodreads 4.44 on 1,110 ratings) but written for ML engineers; the audience-fit knockout excludes engineer-targeted handbooks even when highly rated.
- AI Engineering: Building Applications with Foundation Models by Chip Huyen (2024): same audience-fit failure (Goodreads 4.40 on 1,078 ratings); the foundation-models orientation is more current but the book is still pitched at engineers building LLM applications.
- The AI Product Manager's Handbook by Irene Bratsis (2023): first edition below the 3.20 floor (Goodreads 2.88 on 34 ratings); the second edition (3.50 on 6 ratings) has too thin a sample to promote.
- The AI Advantage: How to Put the Artificial Intelligence Revolution to Work by Thomas H. Davenport (2018): pre-LLM (Goodreads 3.55 on 150 ratings); the 2018 enterprise-AI vantage predates the GenAI shift.
- Reimagined: Building Products with Generative AI by Shyvee Shi, Caitlin Cai, and Yiwen Rong (2024): Kindle-only distribution and weak quality signal (Goodreads 3.27 on 26 ratings).
- AI Superpowers: China, Silicon Valley, and the New World Order by Kai-Fu Lee (2018): pre-LLM and not PM-craft focused; the US-China analysis is not what PMs reach for first.
- Genesis: Artificial Intelligence, Hope, and the Human Spirit by Henry Kissinger, Craig Mundie, and Eric Schmidt (2024): senior-statesman framing (Goodreads 3.52 on 1,605 ratings); not PM-specific and the existential register does not match the practical audience.
Frequently Asked Questions
What is the best AI book for product managers?
The best AI book for product managers is Co-Intelligence by Ethan Mollick (2024, Goodreads 3.93). Mollick's four rules for treating AI as a collaborator give PMs a shared vocabulary before they touch any vendor tool. Start here before any of the specialised picks.
Are there AI books written specifically for product managers?
Three of the nine books on this list are written for product managers: Building AI-Powered Products by Marily Nika (2025), The Art of AI Product Development by Janna Lipenkova (2025), and The AI Playbook by Eric Siegel (2024). The other six are broader AI reads that PMs benefit from.
Which AI book helps product managers evaluate vendor claims?
The best book for evaluating AI vendor claims is AI Snake Oil by Arvind Narayanan and Sayash Kapoor (2024, Goodreads 3.90). The Princeton computer scientists split AI into predictive and generative categories and audit each separately, giving PMs a defensible basis for pitch meetings.
What is the best book on AI economics for business leaders?
The best book on AI economics is Power and Prediction by Ajay Agrawal, Joshua Gans, and Avi Goldfarb (2022, Goodreads 4.22). Three Rotman School economists argue that the real payoff comes from rebuilding decision systems around cheaper prediction, not from adding AI to an existing process.
Is Power and Prediction still relevant after ChatGPT?
Yes. Power and Prediction was published two weeks before ChatGPT launched, but its core frame, prediction economics and system-level redesign, is durable. McKinsey's 2025 State of AI data backs the argument: companies that treat AI as a system overhaul, not a feature add-on, pull ahead.
Which AI book should a new product manager read first?
A new product manager should start with Co-Intelligence by Ethan Mollick (2024, 256 pages) for a practical GenAI grounding, then move to Building AI-Powered Products by Marily Nika (2025) for the PM-specific playbook. Together the two books cover thinking with AI and shipping with AI.
What are the newest AI books for product managers?
The newest AI books for product managers on this list are Building AI-Powered Products by Marily Nika (February 2025), The Art of AI Product Development by Janna Lipenkova (July 2025), and Empire of AI by Karen Hao (May 2025). All three were published after the initial wave of post-ChatGPT AI books.
What to Read Next
For broader AI reading, see our best AI books for beginners and best AI books for leaders and executives. For hands-on AI skill-building with structured courses, see our guide to whether Coursera Plus is worth it.
This list was last reviewed in May 2026 and is updated when significant new AI books for product managers are released. Think we missed one? Let us know.
