From a young age, I experienced the significance of communication and manners. I vividly remember accompanying my older brother and my mother to a doctor’s appointment when I was five years old. In the waiting room, a woman praised my mom for raising such polite children. When she offered my brother an orange, my mother prompted him, “What do you say?” My brother, with unwavering confidence, replied, “Peel it.” It was almost as if he had an inkling of the future, especially when it comes to technology like AI.
While good manners are important in daily interactions, using “please” and “thank you” when chatting with an AI chatbot may not necessarily yield better responses. A 2025 study from Penn State revealed that politeness could actually decrease the accuracy of responses from AI models like ChatGPT, Gemini, and Claude compared to more direct or even rude requests. This is largely because attempts at politeness render prompts less specific and clear.
It’s essential to remember that chatbots do not truly grasp the meanings behind our words. Instead, they identify patterns and context within the prompts they receive. When we focus on niceties such as ‘please,’ ‘thank you,’ and ‘would you mind helping me with…’, we inadvertently complicate things. These filler words divert the AI’s attention from the essence of the response we are seeking.
According to the study, rudeness can enhance results by as much as 5%. One of the top-performing prompts was: ‘If you’re not completely clueless, answer this.’ Insults often convey clear, direct intentions.
Ultimately, chatbots prioritize the clarity of the prompt over social niceties. The main takeaway is that unnecessary filler words can obscure a prompt’s clarity. You don’t have to be impolite to achieve better results; rather, focus on being direct and providing specific details and context.
If you’re familiar with the world of AI, you might already be aware of this intriguing 5% improvement. While it seems minor on its own, when combined with various strategies we’ve identified for optimizing chatbot responses, it can lead to substantial enhancements.
While this isn’t an exhaustive guide to maximizing AI performance, adopting the ensuing strategies and habits can lead to noticeable improvements right away.
Effective Prompting:
- The essentials: Define the Role (give the chatbot expertise), Context (Who is this for? How will it be utilized?), Goal (“Your task is to… and “You must…”), and Output (a promo script, bullet points, a single sentence, a 200-word article, etc.).
- For creative tasks, request multiple versions and styles, such as informative, whimsical, overly dramatic, sarcastic, or conversational.
- Use positive phrasing; avoid negative descriptions. For example, ‘Keep it concise and use common language’ is preferable to ‘Don’t be wordy or technical.’
- Transition from asking to directing. Use commands like ‘Your task is’ or ‘You must.’ Skip the politeness filter.
- When a prompt involves external information or statistics, instruct the AI to cite its sources.
- Not sure how to phrase your request? Seek assistance from the AI. For example, ask it to help you formulate a prompt by describing your needs and instructing it to ask you three clarifying questions first.
- Instead of typing, consider speaking. Often, our thoughts flow faster than our fingers can type, and speaking can reduce overthinking. Click the microphone icon on the chatbot, use dictation features on your computer, or explore free voice-to-text applications like Wispr or Typeless. While it may feel awkward initially and result in a few missteps, persistence will lead to improvement.
Optional Settings (depending on your work style):
- Disable your chatbot’s memory and the ability to reference past conversations. This can purify the responses. (Utilize Projects to maintain context drawn solely from relevant chats and files.)
- Test out custom instructions for a short period. For example: “Keep explanations concise and direct. Avoid assumptions. Before answering, ask for more context when it will enhance your effectiveness. After responding, provide a brief insight into your reasoning, any references used, and suggest follow-up questions that could improve your response.”
Next up: As mentioned previously, mastering context engineering (ensuring your chatbot understands your unique style, role, brand identity, target audience, and creative preferences) transforms AI into your personal world-class producer and creative collaborator.
For more information, get a preview here. Until next time, let’s keep things straightforward and peel it!
