If you have used AI for anything practical — writing, analysis, advice, planning — you have probably had this experience: you ask a reasonable question and receive a reasonable answer that does not quite apply to your situation. It is accurate in a general sense. It covers the right territory. It just doesn’t specifically help you. You read it and think “yes, but my situation is different,” and then you try to explain how it is different, and AI tries to adjust, and after three or four rounds you are somewhere closer to useful but you have spent twenty minutes getting there.
This is not AI being bad at its job. It is AI doing exactly what it was asked. The generic question produced the generic answer. The specific answer requires a specific question.
Why AI Defaults to the Average
When you ask AI a question without context, it calibrates to the center of the distribution of people who might plausibly ask that question. “How should I structure a cover letter?” might be asked by a new graduate, a senior executive, someone changing industries, someone applying within their field, someone applying in a formal industry, and someone applying in a creative one. AI has no way to know which you are unless you say. So it produces an answer calibrated to the average person asking that question — which is accurate for the center of the distribution and less useful the further you are from it.
This is not a malfunction. It is a logical response to ambiguity. The system optimizes for what is most likely to be useful given what it knows, and what it knows is the question — which, without context, is everything it has to work with.
The One Thing That Changes This
Context. Specifically: who you are relative to this question, what your actual situation is, and what you are trying to accomplish.
The cover letter question transforms when you add: “I’m a 12-year project manager in construction transitioning to SaaS sales. My main challenge is that I have no SaaS experience but strong client-facing, quota-carrying experience in a different industry. I’m applying to a mid-market sales role at a B2B logistics software company.” That is no longer a generic cover letter question. It is a specific question about a specific person in a specific situation, and the answer to it will be specific in return.
The context does not need to be long. It needs to contain the things that make your situation different from the average person asking this question. Your role, your goal, your constraint, your specific version of the problem.
The Briefing Instinct vs. the Asking Instinct
Most people approach AI with an asking instinct: they have a question and they ask it. The asking instinct is natural and it works well for factual questions with a single right answer. “What is the capital of New Zealand?” does not require context. “What should I say in this performance review?” requires substantial context before any answer is useful.
The instinct that produces useful AI output is a briefing instinct: before the question is asked, a brief is given. Not the question first, then the context when the answer comes back generic — the context first, establishing who is asking, what their situation is, and what a useful answer would accomplish. This reverses the usual sequence. It feels slower because you are doing more work before the response. It is faster in practice because it produces useful output on the first attempt rather than after multiple rounds of adjustment.
What Changes When You Brief Instead of Ask
A brief establishes four things that a generic question omits:
**Who you are** in relation to this question — not demographically, but situationally. What is your relevant context, background, or position?
**What you have already** — what do you know, what have you tried, what has failed? This prevents AI from offering solutions you have already ruled out.
**What the constraints are** — time, resources, specific requirements, things that are non-negotiable. The answer to a question without constraints is often unusable because it doesn’t account for what you are actually working within.
**What a useful answer looks like** — not “help me with this” but “the output I need is X, in Y format, for Z purpose.” This is the most neglected part. AI can produce many formats and levels of depth for any given response — specifying what you actually need changes what it produces.
Generic Output Is Not AI’s Fault
The most important reframe for getting more useful AI output is to stop attributing generic responses to AI’s limitations and start attributing them to the absence of a brief. In almost every case where AI gives you something that doesn’t quite apply, you can trace the failure directly to information that wasn’t provided. The AI was accurate. It was accurate about the wrong person — a person who is a bit like you but not specifically you.
The brief is how you tell AI who you specifically are. It is the information that moves the response from the center of the distribution to your specific situation. It is the difference between AI writing a cover letter and AI writing your cover letter.
Briefing Fox is built to help you build that brief — capturing your specific situation, your goal, and your constraints before any AI task begins. Try it free at www.briefingfox.com.
The Next Time AI Gives You Something Generic
Before you add context after the fact or start a new conversation, ask one question: what did I leave out of the original request that would have told AI who I specifically am and what I specifically need? The answer to that question is what belongs in the brief. Next time, put it in first.
Try Briefing Fox free at www.briefingfox.com.
Because it calibrates to the center of the distribution of people who might ask that question. Without context about who specifically is asking and what their situation is, AI produces the answer that’s most likely to be useful for the average person — which is accurate in general and less useful the further you are from average.
Add context before the question — who you are in relation to this task, what your specific situation is, and what a useful answer would accomplish. The more specific the context, the more specific the response. The adjustment takes less than two minutes and eliminates multiple rounds of correction.
Your role or situation relative to the task, what you’re specifically trying to accomplish (the outcome, not just the task), and any constraints that limit what a useful answer looks like. Those three inputs move AI from the generic case to your specific case.
Almost always the prompts. The same AI that produces a generic answer to a vague question produces a specific, accurate, useful answer to a well-briefed question. The capability is there; the context is what activates it. Generic output is almost always traceable to information that wasn’t provided.