The Difference Between Using AI and Working With AI
There are two fundamentally different ways that professionals interact with AI right now.
The first is the way most people do it: ask a question, get an answer, use or ignore the answer, move on. Open a new conversation tomorrow and start over. The interaction is transactional. The AI is a search engine that writes in full sentences.
The second is different: treat each session as a working relationship. Provide context. Build on prior exchanges. Direct, redirect, and refine. End with something that would have taken significantly longer to produce any other way.
The difference in output between these two approaches is not small. It is roughly the difference between asking a knowledgeable stranger a question in a hallway and having a working session with a capable colleague who has been fully briefed on your situation.
Why the Search Engine Model Underperforms
The search engine model of AI produces mediocre results for predictable reasons.
When you ask a question without context, AI gives an answer without context. The answer is calibrated to a generic version of whoever might ask that question, not to your specific situation, your specific expertise, or the specific constraints you are working within.
A commercial litigator who asks "how should I structure a deposition for a hostile witness" will get a competent general answer. That answer will be correct in a general sense and almost certainly less useful than what she would develop herself, because it does not know anything about her witness, her case theory, her jurisdiction, or the specific facts she needs to establish.
The question is not wrong. The framing is wrong. Asking AI questions is a limited use of the tool. Briefing AI on a situation and asking it to help you work through it is something different.
What Briefing Actually Looks Like
Briefing is not complicated. It is the same thing you do when you bring a new colleague up to speed on a matter.
You tell them what the situation is. What has happened so far. What you are trying to accomplish. What the constraints are. Who the relevant parties are and what their interests are. What you have already tried. What you are not sure about.
When you do this with Claude before asking your question, the quality of the output changes noticeably. The response is calibrated to your actual situation rather than a generic version of it. The follow-up questions are more useful. The output can be directed rather than accepted or discarded.
A corporate strategist who says "I'm working on a situation where a portfolio company in the industrial sector needs to decide between two growth paths — one organic, one acquisitive — and I need help structuring the analysis" will get something more useful than someone who asks "how do you evaluate organic versus acquisitive growth."
The information in that brief — industrial sector, portfolio company context, the fact that two specific paths are already defined, the strategic decision framing — shapes every part of the response.
The Ongoing Session
The second major difference between using AI and working with AI is what happens across the conversation.
Most people treat each AI exchange as complete in itself. They get an answer, they leave, they come back later with a new question in a new conversation. Each conversation starts from zero.
Working with AI means building within a conversation. You ask a question, get an answer, push back on the parts that miss, ask for more specificity, provide additional context, redirect, and iterate. The conversation becomes a work session where the output at the end is substantially better than the output from the first exchange.
A tax attorney working through the structure of a complex estate plan does not ask one question and stop. She describes the situation, reviews the initial analysis, challenges the parts that miss nuances she knows from the client relationship, provides additional constraints, and keeps working until the analysis is genuinely useful. The final output looks nothing like the first.
This is not a technique. It is a mindset shift. You are not consuming AI output. You are producing work with AI as a collaborator.
The Role of Specificity
Experienced professionals often underestimate how much their expertise adds to the quality of an AI interaction simply through specificity.
Generic requests produce generic answers. Specific requests — the kind that only someone with domain expertise could formulate — produce responses that are calibrated to that expertise.
A wealth advisor who asks "how do I handle a client who wants to take on too much risk" gets general advice. A wealth advisor who asks "I have a client who is seventy-two years old, has roughly sixty percent of his liquid assets in employer stock from a company he sold fifteen years ago, is resistant to discussing diversification because the stock has performed well, and I need to help him understand concentration risk without triggering a defensive reaction" gets something actually useful.
The specificity in that second request is not AI expertise. It is wealth management expertise. The AI cannot produce that brief. The professional brings it. That is the leverage point.
When the Shift Happens
Most professionals who go from using AI to working with AI describe a recognizable moment.
They are working on something — a document, an analysis, a difficult email — and instead of asking a question, they describe the whole situation. They include the context, the constraints, what they are trying to accomplish, who the audience is, what is making it difficult. They write more in the prompt than they usually do.
And the output that comes back is noticeably more useful. Not because the AI got smarter. Because they gave it what it needed to be useful.
After that, the pattern changes. They start briefing instead of querying. They start building within conversations instead of starting over. They start thinking about what context to provide before they open the tool.
That shift — from querying to briefing — is the actual skill. It does not require technical knowledge. It requires knowing what matters about your situation, which is exactly the kind of knowledge experienced professionals already have.
Practical Examples
An architect, 17 years in practice: Used to ask Claude to "write a client proposal for a residential renovation." Started instead briefing it on the client relationship, the project scope, the client's communication style, the budget sensitivity, and what the proposal needs to accomplish beyond just describing the work. The proposals now require half the revision time.
A senior HR director, 22 years: Stopped asking "what are best practices for managing a performance conversation." Started saying "I have a direct report who has been with the company for eight years, has a strong technical record but has been increasingly disengaged since a reorganization three months ago, and I need to have a conversation that surfaces what is actually happening without putting him on the defensive." The approach that comes back is calibrated to the situation, not to a generic performance conversation.
An independent consultant, 19 years in supply chain: Briefs Claude on every engagement at the start of each work session — client background, project context, what has been agreed, what tensions exist. Uses the conversation as a thinking partner throughout. Says the biggest change is that she articulates her own thinking more clearly by having to explain it.
Frequently Asked Questions
Does this mean I need to write long prompts every time?
Not every interaction requires a full brief. Routine tasks that are self-contained — draft this email, summarize this document — work fine with short prompts once you have provided the relevant context once. The brief investment pays dividends over a sustained conversation.
Is this different from prompt engineering?
Prompt engineering as typically taught focuses on syntax and formatting techniques. Briefing is about content — what information your situation contains that AI needs to give useful output. Most professionals do not need prompt engineering. They need to stop treating AI like a search bar.
Does this work across different AI tools, or just Claude?
The principle of providing context works across AI tools. Claude tends to be strong at sustained, complex conversations and at working through professional and analytical tasks. The briefing approach is particularly effective there.
What's the fastest way to start getting better results?
Before your next AI interaction, write two or three sentences of context before you write your request. Who are you? What are you working on? What does the output need to accomplish? That small addition changes most conversations meaningfully.
The Leverage Starter course ($199) is built around exactly this transition — from querying to working with AI, using your specific professional context as the input. If you want to build this into a systematic workflow across your practice, Leveraged Associate ($395) covers the broader design.
Where this goes next
Want the guided, build-it-this-week version of this? See The Leverage Starter — or Turn Experience Into Income with Claude if you want the broader path.
Related reading from The Briefing
- The Tasks You Should Stop Doing Yourself (And How to Transfer Them)
- Why I Ignored AI for Two Years and What Changed
- How to Learn a New AI Tool Without Losing Three Days to Tutorials
Not sure which program fits where you are? take the 2-minute course-fit quiz, or browse the full TLY course catalog.