While AI offers incredible speed, its default output often reads like a soulless corporate robot — until you learn how to effectively train it. All three prompt variations in this guide solve that exact problem by using few-shot prompting to clone your unique brand voice directly from your past content. If you just need a quick, authentic social media post, the beginner prompt delivers instant mimicry with zero friction. If you are looking to scale your operations with a reusable style guide, try the intermediate variation; or, if you need complex, long-form content that smoothly transitions from an educational tone to a promotional pitch, dive straight into the advanced prompt.

Every week, Ketelsen.ai runs the same prompt topic through three of the biggest AI platforms on the planet — ChatGPT, Gemini, and Claude — and publishes all three results side by side. Same topic, same template, same rules. The only variable is the AI doing the thinking. This is Week 2, and the topic hit a nerve that every AI user has felt but few know how to fix: role assignment. That one sentence you type before your actual question — "You are a senior marketing strategist" or "Act as a cybersecurity architect" — that turns out to be the single highest-leverage improvement most people never make. All three platforms took their shot at explaining why it works, how to do it well, and what happens when you push the technique to its limits. The scores were closer this week, but one platform still pulled ahead.

Every week, Ketelsen.ai runs the same prompt topic through three of the biggest AI platforms on the planet — ChatGPT, Gemini, and Claude — and publishes all three results side by side. Same topic, same template, same rules. The only variable is the AI doing the thinking. This is Week 2, and the topic hit a nerve that every AI user has felt but few know how to fix: role assignment. That one sentence you type before your actual question — "You are a senior marketing strategist" or "Act as a cybersecurity architect" — that turns out to be the single highest-leverage improvement most people never make. All three platforms took their shot at explaining why it works, how to do it well, and what happens when you push the technique to its limits. The scores were closer this week, but one platform still pulled ahead.

Have you ever asked an AI to draft an email, brainstorm a strategy, or solve a problem, only to receive a response that feels painfully robotic and generic? This happens when the AI is left in its default "helpful assistant" mode, forcing it to guess the context and expertise required for your task. By simply assigning the AI a specific role—like a "Senior Tech Copywriter" or a "Veteran Startup Advisor"—you completely rewire how the model processes your request. One sentence changes everything, instantly elevating the output from amateur to expert level.

Here's the challenge every AI user faces: you ask a perfectly reasonable question, and you get back something bland, vague, or oddly generic. The fix isn't longer questions—it's one simple sentence that changes everything. All three variations below are built around the same core idea: assigning the AI a clear role before giving it a task can dramatically improve focus, tone, and usefulness.

This post explores role assignment — the practice of telling an AI who it should be before asking it to do something — through three progressive variations, each building on the last. Role assignment is deceptively simple: one sentence can fundamentally shift the quality, depth, and usefulness of everything the AI produces.

This week's topic asked each AI platform to produce a complete blog post featuring three prompt variations — Beginner, Intermediate, and Advanced — all targeting the same problem: the handful of avoidable mistakes that cause the majority of weak AI output. The topic matters because it sits at the exact intersection of what every AI user needs and what almost nobody teaches well. Each platform had to produce usable prompts, explain the engineering behind them, provide industry examples, and give the reader something they could use immediately.

You typed a perfectly reasonable question into ChatGPT, Claude, or Gemini — and what came back was a bland, meandering wall of text that could have been written by a bored textbook committee. Sound familiar? Here is the thing most people never realize: the AI did not fail you. Your prompt did. The gap between a disappointing AI response and a jaw-dropping one almost always comes down to a handful of avoidable mistakes — five of them, to be exact. Fix these five, and you will wonder why AI ever felt unreliable in the first place.

We've all been there: you ask an AI for a brilliant strategy, and it hands you a robotic, generic wall of text that sounds like a 1990s textbook. It is incredibly frustrating, especially when you are pressed for time, but the truth is that the AI isn't broken—your prompt just fell into one of the classic traps. By understanding and fixing the five critical mistakes that cause 80% of bad AI output (lack of context, unclear goals, vague language, missing constraints, and missing formatting requirements), you can transform mediocre responses into high-value assets. This prompt acts as your personal AI editor, helping you instantly identify and fix these errors before you ever hit "generate."

Every week, Ketelsen.ai publishes prompt posts that work across ChatGPT, Claude, and Gemini — posts with structured breakdowns, real industry examples, and difficulty levels that scale from beginner to advanced. What most readers don't see is the system that makes that consistency possible. It's called CRAFT — Configurable Reusable AI Framework Technology — and it's the engine underneath everything you read on this site. This post pulls back the curtain — what CRAFT is, why it exists, how it powers every stage of the Ketelsen.ai weekly workflow, and why it's the reason Ketelsen.ai 2.0 can promise something most AI content sites can't: measurable improvement over time.

Ketelsen.ai has always been an experiment. From the beginning, the goal was to go beyond the obvious prompt ideas — past the "write me a blog post about X" suggestions that flood every AI tips list — and get to the prompts that are actually creative, actually useful, and actually new.

That mission hasn't changed. But how we get there has.

Welcome to Ketelsen.ai 2.0.

Bad AI output usually is not the AI being lazy, broken, or secretly plotting against your productivity. More often, it is the digital equivalent of asking a contractor to "fix the kitchen" with no budget, no style preference, and no timeline. This beginner prompt is designed to help new AI users spot the five most common mistakes that derail results before they waste time chasing a second-rate answer. It turns fuzzy prompting into a simple habit anyone can learn in one sitting. Official guidance from OpenAI, Anthropic, and Google consistently points to the same fundamentals: clarity, context, structure, and iteration matter.