Continue reading
Sign in to read this chapter.
Your reading progress and saved items will be available after sign-in.
Contents
- Front matter
- How to use this book
- Read this first
- Part A · For everyone who uses AI at work
- 1Understand what you are using
- 2What is actually happening when it answers
- 3Know where your data goes
- 4Judgement, overreliance and the fluency trap
- Part B · Deciding what to learn
- 5Where to start when everything is moving
- 6Choose what to know, use and master
- 7Build a roadmap you can revise
- 8Put the technology in its place
- 9What it costs and why
- Part C · Building AI into work
- 10Let the company problem drive the design
- 11Retrieve evidence with permission
- 12Test confidential retrieval end to end
- 13Give tools narrow authority
- 14Understand the MCP boundary
- 15Control an agent outside the model
- 16Walk through an attempted overreach
- 17Evaluate the task and its failure modes
- 18Decide what evidence permits release
- 19Design for hostile and unreliable input
- 20Give risk decisions an owner
- 21Assemble the production service
- 22Operate for failures and change
- Part D · Models, platforms and hardware
- 23Separate training from inference
- 24Choose model adaptation and deployment
- 25Read the seven-layer map correctly
- 26Understand GPU and memory constraints
- Practice labs
- 27Practise with evidence
- 28Practise actions and failure recovery
- Back matter
- Glossary
- References
- Technical references and representative tools