Part 3: Working with the Major AI Platforms

Why Platform-Specific Knowledge Matters

Knowing how to prompt AI is a transferable skill. Understanding a specific platform is something different — and both matter.

General prompting principles carry across every AI tool: be specific, provide context, iterate, verify. But each platform has been built with distinct goals, trained on different data, optimized for different interaction patterns, and tuned with different values. A technique that works brilliantly in ChatGPT may produce mediocre results in Midjourney. A workflow that makes Claude shine may be awkward in GitHub Copilot. The reverse is equally true.

This is not a flaw in the AI landscape — it is a feature. Different platforms making different tradeoffs means practitioners can select tools that genuinely fit their work rather than forcing every task through one interface. But it does mean that becoming effective with AI requires platform-specific fluency alongside foundational skills.

Part 3 is where that fluency gets built.

The Landscape as of 2026

The AI tools landscape in early 2026 looks nothing like it did eighteen months ago, and eighteen months from now it will look different again. That said, a relatively stable set of major platforms has emerged across the most important domains of knowledge work:

Conversational AI: ChatGPT (OpenAI), Claude (Anthropic), and Gemini (Google) have established themselves as the dominant general-purpose AI assistants. Each has a distinct character, a distinct set of strengths, and a distinct failure profile. Practitioners who use all three strategically get considerably more from AI than those who rely on any single tool.

AI Code Assistants: GitHub Copilot, Cursor, Tabnine, and Amazon CodeWhisperer have transformed software development. These tools operate directly inside the development environment, making AI assistance ambient rather than conversational. The implications for workflow are significant.

Image Generation: Midjourney, DALL·E 3 (integrated into ChatGPT), and Stable Diffusion have made visual content creation accessible to non-designers while creating new workflows for professional creatives. The platforms differ dramatically in their interfaces, their aesthetic sensibilities, and their control systems.

Specialized and Domain-Specific Tools: Beneath the general-purpose giants, hundreds of tools have been built for specific domains — legal research, medical documentation, financial analysis, scientific literature synthesis, marketing copy, customer support, and more. Many professionals will find their most valuable AI tools in this category.

How to Use Part 3

Part 3 is designed to be read selectively. You do not need to read every chapter in sequence.

If you are a developer, Chapter 17 on GitHub Copilot is essential and Chapter 19's coverage of specialized tools is worth your time. Chapters 14 and 15 on ChatGPT and Claude will deepen your ability to use conversational AI alongside your coding tools. Chapter 18 on image generation may be less immediately relevant — or it may open new possibilities you had not considered.

If you are a marketer, Chapter 18 on image generation and Chapter 19 on marketing-specific tools are directly applicable. Chapters 14 and 15 will sharpen your use of conversational AI for strategy and copy work.

If you are a consultant or knowledge worker, Chapters 14, 15, and 19 are your highest priorities. Chapter 16 on Gemini's integration with Google Workspace is worth reading if you live in that ecosystem.

The chapters can be read in any order. Each is self-contained. Cross-references appear where concepts connect.

Each Platform Has Its Own Personality

This framing is intentional, not metaphorical.

Each major AI platform reflects specific choices by its creators: what data to train on, what behaviors to reinforce, what to optimize for, what guardrails to impose, what tone to adopt. These choices produce something that genuinely behaves differently from platform to platform — not just in capability, but in character.

ChatGPT tends toward helpfulness and comprehensiveness. It will usually give you something, even when it should perhaps tell you it cannot. Claude tends toward nuance and hedging. It is more likely to say "this is complicated" when it is. Gemini is deeply integrated with Google's data and services in ways the others are not. Copilot is contextual in a way that conversational AI is not — it knows what code you are looking at.

These personalities create distinct failure modes. A platform optimized for helpfulness will sometimes be helpful in ways that are confidently wrong. A platform that hedges may frustrate users who need direct answers. Understanding these tendencies is not a criticism of any tool — it is the foundation for using each one appropriately.

A Preview of Chapters 14 through 19

Chapter 14: Mastering ChatGPT covers the foundational conversational AI tool, GPT-4o capabilities, the Custom Instructions feature, memory, plugins, data analysis with Code Interpreter, and the full range of ChatGPT's increasingly capable feature set.

Chapter 15: Working with Claude examines Anthropic's Claude — its extended context window, its particular strengths in analysis and long-document work, its character and values, and how to use it most effectively for research-intensive and writing-intensive tasks.

Chapter 16: Google Gemini and the Google Ecosystem covers Gemini's deep integration with Google Workspace, its multimodal capabilities, and how it differs from OpenAI's offerings. Particular attention goes to Gemini for Gmail, Docs, Sheets, and Slides.

Chapter 17: GitHub Copilot and AI Code Assistants is the definitive chapter for developers. It covers Copilot's autocomplete and chat interfaces, the Cursor IDE, trust calibration for AI-generated code, security considerations, and a complete coding workflow combining multiple AI tools.

Chapter 18: Image Generation — Midjourney, DALL·E, and Stable Diffusion covers the three major platforms for AI image creation, platform-specific prompting techniques, practical workflows for visual professionals, and the copyright and attribution considerations that any practitioner should understand.

Chapter 19: Specialized and Domain-Specific AI Tools maps the landscape of purpose-built AI tools across legal, medical, financial, scientific, design, marketing, and productivity domains. It also provides a reusable framework for evaluating any new specialized tool — an essential skill as the landscape continues to expand.

A Note on Currency

The AI tools landscape changes faster than any book can track. Specific models are updated, pricing changes, features are added and deprecated, companies are acquired, and new entrants emerge that change the competitive picture. Rather than attempting to be a current buyer's guide — a losing game in print — Part 3 focuses on durable principles: how each category of tool works conceptually, what problems each platform is optimized to solve, what failure modes to watch for, and how to evaluate new tools as they appear.

The specific version numbers, pricing tiers, and feature sets mentioned throughout this part reflect the state of the landscape in early 2026. For the most current information on any specific platform, consult that platform's official documentation.

What will not change: the underlying logic of how to work with these tools. The skills you build here will transfer as the platforms evolve.

Part 3 begins with Chapter 14: Mastering ChatGPT.