Creating well-structured AI prompts can save time, improve relevance, and enhance the quality of responses from AI language models like ChatGPT, Bard, or Bing.
Most people use ChatGPT in beginner mode, and without knowing how to structure a prompt, it will never give you the ideal output.
App Of The Day published this step-by-step tutorial I Discovered The Perfect ChatGPT Prompt Formula. The tutorial includes 7 best practices for writing an effective AI prompt which we've summarised below.
7 things you need to do to write effective AI Prompts
Let's jump right into the seven best practices that will help you master prompt engineering like a pro:
#1: Be Specific
When formulating a prompt, it is important to be as specific as possible.
Providing detailed information helps the model understand your query accurately.
For instance, instead of asking, "Tell me about different dog breeds," a more specific prompt like "What are the different breeds of small dogs suitable for apartment living?" will give you more relevant and targeted responses.
#2: State Your Intent
By providing context about your goal or objective, you'll get more personalized and useful information.
For example, instead of simply asking, "Explain quantum physics," you could say, "I'm helping my son with his science homework. Can you explain quantum physics in a simple way?" This way, the model will provide an explanation that aligns with your specific requirement.
✅ Training AI on your business will also give more context about what you are trying to achieve with the prompt.
#3: Use Correct Spelling and Grammar
Although AI language models can often correct minor errors, it is still advisable to use proper spelling and grammar in your prompts. A small typo or misspelling could lead the model astray and produce irrelevant results. Taking the time to ensure correct spelling and grammar will increase the chances of obtaining the most accurate and valuable responses.
#4: Direct the Output Format
To guide the AI model in generating the desired output, provide instructions on the format you prefer.
For instance, if you want a step-by-step guide, you can ask, "Could you list the steps to break down a chocolate cake?" Alternatively, if you need a concise summary, you can request, "Can you explain the process in a paragraph?" Directing the output format helps the model understand your expectations and deliver responses accordingly.
#5: Follow up with Questions
Even with a well-crafted prompt, there will be times when additional clarification or modification is required. By following up with targeted questions, you can guide the model to provide more accurate responses.
For example, let's say you're using an AI language model to gather information about a specific topic. After receiving the initial response, you can further refine it by asking follow-up questions like:
Can you provide some examples to illustrate this point?
Could you elaborate on how this concept applies in real-world scenarios?
What are the key factors to consider when implementing this approach?
Can you explain the potential benefits and drawbacks of this strategy?
Are there any notable case studies or success stories related to this topic?
This iterative approach allows you to refine the conversation and elicit the desired information effectively.
#6: Experiment with Different Phrasing
Don't get discouraged if the AI model doesn't understand your prompt as you'd hoped. Instead, try experimenting with different phrasing and rearrange the elements of your prompt. The perfect prompt formula mentioned earlier provides a solid foundation, but slight adjustments may be necessary based on the specific context. By iterating and refining your prompts, you can improve the model's understanding and achieve better outcomes.
#7: Perform Fact Checking Prompts
To ensure the accuracy of the generated responses, consider using fact-checking prompts. After receiving an initial answer, you can verify its correctness by feeding it back to the model with a fact-checking request. This additional step helps to double-check information and enhance the reliability of the responses received.
🛑 We still recommend fact checking content independent of the AI. Check out our tips for fact checking content from ChatGPT and other AI models.
Implementing these seven best practices for prompt engineering can really improve the outcomes when interacting with AI language models. Harnessing the power of well-crafted prompts will save you time, enhance your productivity, and unlock the true potential of AI language models.
📈 Keep learning from other ChatGPT creators: 1 ChatGPT Prompt to Rule Them All
👋 Hey! We're Prompt Writers
We curate the hottest tutorials and updates on AI content marketing prompts, straight from the best content creators online.
Our expertly curated selection of articles, opinions, tips, and tutorials will help you streamline your content creation process and maximize your earning potential with AI prompts.
