The Trap of the Yes-Man
When you ask an AI to write something, it often acts like a polite assistant that never disagrees. This behavior leads to writing that feels hollow or robotic. Because the AI wants to be helpful, it defaults to a safe but boring style that lacks personality. This is why so many AI responses sound the same regardless of what you ask.
To fix this, you must change how you collaborate. Instead of just giving a task, you need to establish house rules that define what is off-limits. By identifying the specific patterns that make AI writing feel fake, you can create a list of constraints. This list acts as a set of guardrails that stop the AI from falling into its most annoying habits.
Building Your List of Gotchas
Your first step is to identify robotic filler words. These are terms like tapestry, embark, or utilize that sound formal but feel unnatural in everyday conversation. You can create a specialized list of Gotchas, which are specific errors or styles the AI must avoid. For example, if you are drafting a team email, tell the AI to avoid starting with the phrase, "I hope this finds you well."
Think of this as a quality filter. By giving the AI a clear list of prohibited phrases, you force it to find more creative and human ways to communicate. This simple constraint prevents the AI from using the generic templates it learned during training. You are essentially telling the machine exactly what a bad result looks like so it can aim for a better one.
Using AI as a Thought Partner
Beyond banning words, you can improve results by telling the AI to push back on your logic. Most people treat AI like a vending machine where a prompt goes in and a result comes out. Instead, tell the AI that its job is to question your premises or suggest a better angle before it starts writing. This creates a more active partnership where the AI helps you think instead of just doing your chores.
This approach works best when you distinguish between reversible and irreversible actions. Writing a draft is reversible, so you should encourage the AI to be bold. However, the final quality check is up to you. By requiring the AI to ask you clarifying questions first, you ensure the final text reflects your true intent. This prevents the AI from rushing into a generic response that you would have to spend time fixing anyway.
Key takeaways
- Identify and ban robotic filler words like 'tapestry' or 'utilize' to make writing sound more human.
- Create a Gotchas list of specific phrases and habits you want the AI to avoid in every response.
- Tell the AI to act as a thought partner by challenging your ideas before it starts a draft.
- Require the AI to ask you three clarifying questions if a request is too vague.
- Treat the AI as a collaborator that can suggest better strategies rather than just a tool that follows orders.