The A to Z map of the book

The core is 26 chapters, one per letter, from A to Z (p. 3-4). After every third letter, from C to U, there is a best practice guide: seven in total. A foreword by Pascal Bornet opens the book and a glossary of terms closes it (p. 8, p. 279).

The authors say it openly: some core ideas come back more than once (p. 4). The chapters are separate, so pick the letter that matches your problem today.

Most chapters follow the same three parts. First a short essay on one big topic: decisions in D, risks in F and R, prompt engineering in P, the future of work in O. Then a list of terms that start with that letter. Each term gets a plain definition, a business example and an example prompt you can adapt.

Last, productivity prompts for jobs that start with the same letter, from Accountant to Zoologist (p. 19, p. 277). It sounds like a game, but it makes a point. A hydrologist and a tax accountant can use the same tools. The prompts show what that looks like on a normal working day.

Between the letters sit seven longer guides. They are the most hands-on part of the book.

1. 1Artificial Intelligence Centre of Expertise essential rolesp. 36Who you need, from the executive sponsor to support engineers, with a RACI to say who does what.
2. 2How to build a generative AI business case and obtain ROIp. 73Value drivers, KPIs, direct and indirect costs, and how to tell the story to stakeholders.
3. 3How to select the right large language model(s) for your businessp. 105Model size, pre-training, fine-tuning needs, speed and cost, roadmap, licensing, and how to test.
4. 4Citizen innovators: AI powered superhumansp. 135How staff outside IT can build with low-code and generative tools, and the guardrails that keep it safe.
5. 5Key advice when implementing generative AIp. 174Assessment and discovery first, plan support and maintenance, stay tech agnostic.
6. 6Industry use cases for generative AIp. 210From drug discovery and fashion to customer service chatbots and code.
7. 7Exploring generative AI’s impact on public educationp. 238How teachers can win time back, with views from a teacher and two professors.

The longest, on citizen innovators, runs 15 pages. The shortest, on industry use cases, fits on two (p. 135-149, p. 210-211).

Each letter, its first page and the main topics. Page numbers follow the book’s table of contents (p. 3).

1. Ap. 12What generative AI is. “Act as” copywriting prompts. Summaries, translation, APIs.
2. Bp. 20AI and generative AI side by side. Biometrics, blockchain, big data.
3. Cp. 27Chain prompts. Copilot and Code Interpreter. Can ChatGPT replace software engineers?
4. Dp. 42Decision Insight. Eleven moves for leaders. Deep learning, domain-specific models.
5. Ep. 53Ethical AI. Explainable AI. Error handling.
6. Fp. 59Use cases by type of content. Fifteen risks and fifteen ways to reduce them. Fine-tuning, foundation models.
7. Gp. 81The Gartner framework for content. GANs and GPT. Generative design.
8. Hp. 89Six steps for leaders. Human-in-the-loop. Email productivity.
9. Ip. 98Prompts for your social profiles. Inference, interpretability, integration.
10. Jp. 111Prompts for job seekers. Jobs changed by AI. Rules that differ by country.
11. Kp. 118The generative organization. Knowledge bases and knowledge graphs. Key challenges.
12. Lp. 125Large language models. Large Action Models. Latent space.
13. Mp. 150What a model is. Moore’s law. A marketing prompt checklist. Multi-modal AI.
14. Np. 160Jobs and adaptation. Natural language processing. Neural networks, no-code.
15. Op. 167Outlook on the future of work. A boardroom assistant. Overcoming bias.
16. Pp. 181Prompt engineering. People and generative AI. Privacy-preserving AI.
17. Qp. 191Quality input and hallucinations. Quantum computing and AI.
18. Rp. 200Five ways to reduce AI risk. EU and US rules. Retrieval augmented generation.
19. Sp. 212Selling with ChatGPT. Social media. Sentiment analysis, supervised learning.
20. Tp. 221How teachers use ChatGPT. Code testing and creation. Transformers.
21. Up. 229Unsupervised learning. The ultimate prompt guide. Uncertainty.
22. Vp. 245The value of generative AI. Vector databases. Voice cloning.
23. Wp. 253Writing: tones, styles and reading levels. World models. Workflow automation.
24. Xp. 263eXplainable AI. eXtended reality content.
25. Yp. 268Yield optimization. Personalization. A creative companion.
26. Zp. 273Zero-shot learning. Zero-day exploit detection. Zombie model detection.

The introduction calls the book a user manual for the non-specialist line-of-business manager (p. 4). It is not a technical manual. The aim is to help a manager follow the talk with technical teams, ask better questions and see where AI can help (p. 10-11).

Two pages go deeper: use cases and prompts, and risks, ethics and adoption.