PART I — Learning How to Work With AI
Chapter 6
Conversation Beats Prompting
You Don't Have to Get It Right the First Time
One of the biggest changes in working with AI is realizing that your first question doesn't have to be perfect.
You can ask something, see the response, add information, correct a misunderstanding, ask for another explanation, or change direction.
That means you don't have to treat every interaction as a carefully constructed command. You can have a conversation.
AI Isn't an Automat
The old Philadelphia Automat offers a useful picture of a transaction: choose something, pay for it, and take it.
There is no conversation.
AI can work differently. You ask. AI responds. You react. You add context. AI adjusts. You discover something you hadn't considered, and that creates the next question.
The value may come less from the first answer than from what happens next.
Let the Answer Teach You What to Ask
Sometimes you don't know the next question until you see the current answer.
A new term may lead to a question you couldn't have asked five minutes earlier. A recommendation may reveal two different approaches. A problem you thought you understood may turn out to be a different problem altogether.
That is one of the most useful features of conversational AI:
The answer can change what you know, and what you know can change your next question.
You Can Change Your Mind
Your first question is allowed to be incomplete.
You may discover that you left something out. You may realize that your priorities have changed. You may discover that you were solving the wrong problem.
You don't have to defend the original question.
You can simply start again with what you now understand.
The book uses a personal example involving a collection of old photographic slides. What initially looked like a question about digitizing photographs became a much larger question about preserving, organizing, retrieving, and eventually finding those memories again. The conversation helped reveal the problem behind the problem.
The Goal Isn't the Perfect Prompt
There are situations where a carefully constructed prompt is useful. Large amounts of information, complicated requirements, specific formats, and repeatable tasks can benefit from detailed instructions.
But that is different from believing that successful AI use depends on discovering the perfect wording.
The goal isn't the perfect prompt. The goal is a useful result.
Sometimes the quickest route is a carefully written request. Sometimes it is a conversation. Often, it is both.
What You'll Discover
The book goes further into how conversation can:
help you discover what you really mean;
expose missing information;
reveal assumptions;
change the direction of a problem;
help you examine another point of view;
turn an ordinary question into a more useful investigation;
show when a detailed prompt really is worth the effort.
You will see why the conversation itself can become part of the thinking process.
Try It Yourself
Take a question you might normally ask an AI assistant.
Ask it naturally rather than spending time trying to construct the perfect wording.
Read the response. Think about what you learned from it. Then continue the conversation.
After several exchanges, look back at where you started.
You may discover that the question you began with wasn't quite the question you really needed to answer.
Why Read the Book?
Chapter 6 takes the idea from Chapter 5 one step further.
Chapter 5 asks you to ask better questions.
Chapter 6 shows you that you don't necessarily have to know the better question before you begin.
Understanding AI develops this idea into a practical way of working with AI: start somewhere, pay attention to what you learn, and let the conversation help shape what comes next.
You don't need the perfect prompt. You need enough of a starting point to begin a useful conversation.