Pages
225
Published
2024
A Practical Guide to Using AI Coding Tools and Productivity Assistants in Your Daily Workflow
Learn to use AI coding assistants, prompt engineering, and automation tools to write better code faster and ship more in less time.
AI-Assisted Programming by Tom Taulli is a hands-on guide to the AI tools that are already reshaping how software gets built. Covering GitHub Copilot, ChatGPT, and a range of other coding assistants, it shows you how to integrate these tools into real work: writing, reviewing, debugging, and documenting code. You will finish with a clear, practical framework for deciding when AI helps and when it gets in the way.
AI coding assistants are no longer experimental. Developers are using them every day to autocomplete functions, generate tests, explain unfamiliar code, and draft documentation. The question is no longer whether to use them, but how to use them well.
AI-Assisted Programming gives you that how. Tom Taulli walks through the tools that have the most traction in professional workflows, including GitHub Copilot, ChatGPT, and several others, and shows you exactly how each one fits into the stages of a real development cycle. This is not a survey of what AI can theoretically do. Every chapter focuses on concrete tasks you will recognize from your own work.
You will learn how to write prompts that produce useful output, not just plausible-looking code you have to fix anyway. You will see how to review AI-generated code critically, catch the subtle mistakes that tools introduce, and build the habit of treating AI output as a first draft, not a final answer. The book also covers how AI tools apply beyond code itself: generating and maintaining documentation, writing unit tests, and assisting with code reviews.
Taulli does not oversell. He is direct about the failure modes: hallucinated APIs, license and attribution risks, security vulnerabilities in generated code, and the productivity trap of accepting bad output too quickly. Understanding these limits is part of using the tools responsibly.
At 225 pages, this book is deliberately focused. You can read it in a weekend and start applying it on Monday. Whether you are a working developer looking to speed up routine tasks or a tech lead evaluating AI tools for your team, this is the practical reference you will keep returning to.
You get a clear map of the current AI coding tool ecosystem, covering what each major tool does and where it fits. This chapter establishes the mental model you will use throughout the book.
You install and configure GitHub Copilot inside your editor and run through your first real suggestions. The chapter covers settings, context window behavior, and how Copilot decides what to suggest.
You learn how to write prompts that produce useful, accurate code rather than plausible-looking noise. The chapter covers prompt structure, providing context, and iterating when the first result misses.
You work with ChatGPT and similar LLM interfaces for coding tasks including explanation, refactoring, and problem-solving. The chapter shows where chat-based tools outperform inline assistants and where they fall short.
You use AI tools to generate unit tests for existing functions and new code, then evaluate the output for correctness and coverage gaps. The chapter includes a process for reviewing generated tests before adding them to your suite.
You apply AI assistants to writing docstrings, README files, and inline comments, and use them to decode unfamiliar codebases. The chapter covers how to prompt for accurate documentation rather than generic filler.
You integrate AI tools into the code review process to catch issues, suggest improvements, and explain tradeoffs. The chapter also covers using AI to refactor legacy code safely.
You examine the concrete failure modes of AI coding tools: hallucinated APIs, security vulnerabilities, licensing risks, and over-reliance. This chapter gives you the criteria to decide when not to use AI output.
You pull together everything from previous chapters into a repeatable daily workflow. The chapter includes guidance for individuals and for teams adopting AI tools at scale.
The book covers concepts and tools that apply across languages, with examples primarily in Python and JavaScript. GitHub Copilot examples use VS Code, but the principles transfer to other supported editors.
No machine learning background is required. The book treats AI tools as software you configure and use, not systems you build. Solid general programming experience is all you need.
Published in April 2024, the book covers tools and practices current at that date. The core skills it teaches, prompt engineering, output validation, and workflow integration, remain applicable even as specific tool interfaces update.
Yes, the book includes code examples throughout. Check the publisher page at O'Reilly for any companion materials or errata associated with this edition.
This book is not for readers who want to build AI systems or train models. It is also not a beginner programming tutorial. It assumes you already write code professionally and want to work faster and more effectively with AI assistance.
New
The Disruptive Economics of Artificial Intelligence and What It Means for Business Decisions
by Ajay Agrawal
New
New
New
How AI and the Next Wave of Technology Will Transform Power, Nations, and Humanity