Absolute Beginner’s Guide To Crafting Great ChatGPT Prompts


As someone who’s passionate about conversational AI, I’m always eager to share tips on how to make better prompts for ChatGPT. Crafting effective prompts is essential to get the most out of these powerful language models, and it’s something I’ve become quite skilled at over time. And you can become skilled as well. Here’s how.

Crafting effective ChatGPT prompts involves understanding prompt types, avoiding mistakes, focusing on clarity, specificity, and context, and using advanced techniques like breaking prompts into steps and implementing constraints. To improve, provide real-life scenarios, experiment, and practice.

In this article, I’ll share some techniques I’ve discovered to help you improve your ChatGPT prompts. By following these tips, you’ll be able to generate more accurate and relevant responses for any questions you might have, as well as elevate your interactions with ChatGPT to the next level!

Table of Contents

Understanding ChatGPT Prompts

Chameleon GPT Putting On Many Hats

When it comes to making the most out of ChatGPT, I believe that understanding the different types of prompts and common mistakes is key. Let’s dive into these two areas to improve your prompt-writing skills.

Prompt Types

First, there are several types of prompts that can be used with ChatGPT. Some prompts simply ask questions, while others instruct the AI to generate content, such as essays or stories. You may also encounter prompts that direct the AI to adopt a specific role, like that of a tutor or an advisor, to provide more contextualized responses. The trick is to figure out the right combination of instruction and context for your specific need.

Common Mistakes

Now, let’s talk about common mistakes when crafting prompts. One common mistake is being too vague. ChatGPT might not be able to generate a suitable response if your prompt is unclear. To avoid this, use specific keywords and clear instructions when crafting your prompts.

On the other hand, overloading your prompt with too much information or unnecessary context can also confuse the AI. Keep your prompts concise and to the point to help ChatGPT better understand what you’re looking for. Additionally, it’s crucial to consider the audience you’re targeting, as this will help shape the AI’s responses in a way that resonates with your intended readers.

Crafting Effective Prompts: SPEED

SPEED: Simplicity, Precision, Economy, and Details

Creating better ChatGPT prompts involves paying attention to four key elements: clarity, specificity, briefness, and context. By keeping these in mind, I can help ensure that my ChatGPT model provides accurate and engaging responses.

Simplicity

Writing clear simple prompts is essential for successful communication with ChatGPT. I make sure to use simple language and avoid any ambiguity, so the model understands exactly what I’m asking for.

Precision

Being precise in my prompts is crucial. By providing detailed information about my request or topic, I help the ChatGPT model generate more useful and relevant responses. For example, instead of saying “Recommend a restaurant,” I would ask, “Can you recommend a vegetarian restaurant in Paris?”

Economy

While it’s important to be specific, I also strive to keep my prompts brief. This ensures that the model stays focused and avoids generating unnecessarily lengthy responses. A concise prompt helps me get the information I need, faster.

Details

Providing context is another crucial aspect of crafting effective prompts. I always give some background information when necessary, so the ChatGPT model understands the bigger picture. This helps generate responses that align with the desired topic or scenario.

Tips For Dealing With Problems

There are a lot of common errors that occur when crafting ChatGPT prompts. To solve the problem it’s important to identify it first and then to make adjustments for instance,

  • Trying to do too much in a single prompt. At times having a prompt that can do multiple things is very useful but for day-to-day use, Simple is better.
  • If you’re being too vague try to be more specific with your request.
  • If you’re putting too much information into the prompt it can confuse ChatGPT
  • Not providing enough context. You need to provide enough background information and your prompts so that the AI understands the big picture.

Keep in mind the SPEED technique, (Simplicity, Precision, Economy, and Details) to help you create more effective prompts.

Examples of Bad Prompts

Say What???

Take a look at each of these prompts and before you go to the explanation of why it’s bad use your new understanding to figure out where it’s lacking.

Bad Prompt: Tell me about technology.

What if I asked you to tell me about technology what would you say?

Personally, I have no idea what you would say. Why because the question is too vague you could tell me about literally everything from semiconductors to the wheel. This prompt is too vague it is not Precise. A better prompt would specify exactly what type of Technology you were interested in.

Bad Prompt: I want a recipe that I can make tonight and I have some chicken in the fridge and some vegetables. I want something that’s spicy but not too spicy. I need a recipe that I can make in about 30 minutes any suggestions?

Prompts like this tend to confuse ChatGPT. This prompt included lots of information that ChatGPT doesn’t need. It doesn’t need to understand that you’re making this chicken for tonight. It lacks Economy. All the additional detail is not part of the context that it needs. It’s not going to understand that you want something that’s spicy but not too spicy.

Take a minute to think about how would you rewrite this prompt to be better?

Better Prompt: Please give me a recipe for chicken and vegetables that is spicy and that I can cook within 30 minutes.

Moving on

Bad Prompt: Can you provide me with an in-depth analysis of the impact of climate change on global economies and political systems?

Similar to The Prompt that we had before about technology they’re just too many possible ways to answer this question. It’s also very complicated so over complex as well. So what might be a better prompt if we were talking about the economies of developing countries?

Better Prompt: How has climate change affected the economies of developing countries?

The new prompt is Precise with more specific Details.

Bad Prompt: I want to know how to cook chicken.

Well, that’s nice but are you cooking a whole chicken, chicken breasts, do you want to fry it, or do you want to bake it? I think you see the problem. The prompt lacks Precision and Details.

Better Prompt: I want to learn how to marinate chicken breasts and cook them on the grill.

Moving on

Bad Prompt: What are some Cool things in Paris?

If somebody asked me, “What are some cool things in Paris?”, my first thought is are you going there? There are lots of cool things in Paris some of them are open to the public some of them are not. Some of them I would suggest to an older person some of them I wouldn’t. The definition of cool is very different between a teenager and an adult.

So that prompt Lacks important Details.

Better Prompt: My wife and I who are in our 30s are planning a trip to Paris what are some interesting things to see there?

If you can avoid these common mistakes you’ll get much better results from ChatGPT. I hope you enjoyed this exercise.

Advanced Techniques

Blast Off!

Now that we coved the basics I’m excited to share some advanced techniques that can help you craft even better ChatGPT prompts. These techniques can significantly improve the responses you get from the AI, and I’ll be explaining them step-by-step. Let’s dive in!

Step-by-Step Guidance

One of the most effective ways to get the desired output from ChatGPT is by breaking your prompt into smaller steps. This way, the AI can focus on each aspect of your query and give more accurate answers. I’ve found great success by using bullet points or numbered lists to present the steps to the AI, like:


1. Explain the concept of gravity.
2. Describe its effects on our daily lives.
3. Discuss its importance for space exploration.

ChatGPT will then respond to each step, which helps maintain clarity and focus.

Conditional Responses

Another useful technique I love using is to ask ChatGPT for different answers based on certain conditions. By doing this, you give the AI explicit instructions on how to approach your query, making the generated responses more useful and relevant.

For instance, you can prompt the AI with:

If the audience is beginners, explain the concept of artificial intelligence in simple terms. Otherwise, provide a more advanced explanation of AI and its current applications.

This way, the AI understands the context and adjusts its response accordingly.

Implementing Constraints

It’s important to set constraints so that ChatGPT produces answers within specific boundaries. By guiding the AI with limitations, the output is more likely to meet your expectations. You can do this by simply stating the constraints explicitly in the prompt.

An example of such a prompt can be:

Describe the process of photosynthesis, using only simple terms and without discussing any molecular or biochemical details.

With these constraints in place, ChatGPT will provide a concise and understandable explanation of the topic at hand.

Examples and Tips

Getting Real

Creating effective prompts for ChatGPT has a significant impact on the quality of the generated responses. I’ve put together some examples and tips to help you improve your ChatGPT prompting skills.

Real-Life Scenarios

Imagine you want to ask ChatGPT about the impact of deforestation on wildlife. Instead of simply writing, “What happens to wildlife due to deforestation?”, you could provide more context by writing, “As deforestation increases around the world, how does this affect wildlife in terms of habitat loss, food scarcity, and species extinction?”

Here’s another example: if you’re looking for information on effective project management strategies, you might write, “In a workplace setting, what are some tested project management strategies that can help teams boost productivity and achieve their goals in a timely manner?”

Best Practices

When crafting prompts for ChatGPT, keep these best practices in mind:

  • Be specific with your request, as vagueness can lead to less helpful responses. According to ReadyForAI, clear and concise prompts yield better results.
  • Provide context and background information. This helps ChatGPT understand your requirements better and generates more relevant responses. GPTBot also emphasizes the importance of providing context.
  • Experiment with different phrasings and approaches. If you’re not satisfied with ChatGPT’s initial response, try rephrasing your prompt or asking the question differently. This can lead to better results.

Remember, practice makes perfect. The more you experiment with ChatGPT prompt crafting, the better you’ll become at eliciting high-quality responses.

Conclusion And Additional Resources

It’s been quite an enjoyable journey sharing my knowledge with you. Crafting better ChatGPT prompts takes practice and persistence, but I’m confident you’ll excel in it as you experiment with different techniques.

Always remember to be clear and concise with your prompts. For even better results, try guiding the AI by providing context or data. And, if you feel your responses could benefit from additional organization, make use of the available Markdown output for a visually structured text.

Don’t forget to have fun, and let your interactions with ChatGPT be a learning experience. Happy prompting!

Chris

Chris Chenault trained as a physicist at NMSU and did his doctoral work in biophysics at Emory. After studying medicine but deciding not to pursue an MD at Emory medical school Chris started a successful online business. In the past 10 years Chris's interests and studies have been focused on AI as applied to search engines, and LLM models. He has spent more than a thousand hours studying ChatGPT, GPT 3.5, and GPT4. He is currently working on a research paper on AI hallucinations and reducing their effects in large language models.

Recent Posts