Hey there, fellow learners! Are you new to the world of prompt engineering? Feeling overwhelmed by the technical jargon and expert-level strategies? Fear not, my friend. I’m here to help you learn the ins and outs of crafting effective prompts.
Modifiers can improve the output of prompts to ChatGPT by providing additional context, specifying desired outcomes, or adjusting the tone of the prompt. Positive modifiers can encourage more creative and optimistic responses, while negative modifiers can limit the scope of responses.
Now, let’s dive into the world of prompt engineering and explore the different types of modifiers that can take your prompts from bland to brilliant. After crafting thousands of prompts (most of them bad), I know firsthand the importance of incorporating modifiers into your prompts. But don’t just take my word for it. Let me show you how it’s done!
Table of Contents
- What Are Prompts?
- Modifiers (The Secret Sauce)
- Modifiers For Creating Content
- Context Context Context
- Challenges to Look Out For
- The End, Well Almost
What Are Prompts?
Well, my dear reader, prompts are simply words or phrases that initiate a response. We use prompts every day in our lives without even realizing it. For example, think of the last time you asked Siri for directions or Alexa to play a song. You used a prompt to communicate with these devices.
But, when it comes to communicating with ChatGPT, prompts take on a whole new level of importance. ChatGPT is a powerful language model that can generate human-like responses to prompts. However, to get the best possible response, you need to craft your prompts carefully. That’s where prompt engineering comes in.
By incorporating modifiers into your prompts, you can elicit specific responses from ChatGPT. For example, using positive modifiers such as “exciting” or “engaging” can encourage ChatGPT to generate responses that are more upbeat and optimistic. Conversely, using negative modifiers such as “caution” or “warning” can steer ChatGPT towards more serious or somber responses.
So, are you ready to become a prompt-engineering pro? Let’s dive in and explore the world of modifiers and how to use them to communicate effectively with ChatGPT!
Modifiers (The Secret Sauce)
Alright, let’s talk about modifiers! In prompt engineering, modifiers are an essential tool to improve the effectiveness of your prompts. There are different types of modifiers, and each one can impact the outcome of your prompts in different ways.
First, there are “positive” and “negative” modifiers. Positive modifiers can enhance the intended meaning of your prompt, while negative modifiers can weaken or even change the meaning altogether. For example, consider the difference between “I want a red apple” and “I don’t want a red apple.” The positive modifier “red” specifies the type of apple desired, while the negative modifier “don’t” changes the focus of the sentence to what is not desired.
| Positive Modifiers | Negative Modifiers |
|---|---|
| always | never |
| best | worst |
| highest | lowest |
| most | least |
| greatest | smallest |
| excellent | terrible |
| outstanding | awful |
| superb | dreadful |
| fantastic | horrible |
| amazing | disgusting |
Another type of modifier is the “quantifier,” which adds specificity to the prompt by indicating the quantity or frequency of the desired outcome. For instance, “I want one red apple” is more specific than “I want a red apple,” and “I want a red apple every day” indicates a consistent desire for the outcome.
| Quantifier Modifiers | Example |
|---|---|
| All | All cats are mammals. |
| Most | Most people prefer dogs over cats. |
| Some | Some fruits are high in sugar. |
| Few | Few people enjoy eating liver. |
| None | None of the cookies were left after the party. |
| Many | Many people enjoy watching movies. |
| Several | Several books on the topic are available at the library. |
| Any | Can I borrow any of your pens? |
| Enough | There is enough food for everyone. |
| Plenty | There are plenty of seats available for the concert. |
Finally, there are “temporal” and “conditional” modifiers. Temporal modifiers specify when the desired outcome should occur, while conditional modifiers add a condition that must be met before the desired outcome can be achieved. For example, “I want a red apple now” is a temporal modifier, while “I want a red apple if it’s available” is a conditional modifier.
| Temporal and Conditional Modifiers | Example |
|---|---|
| Before | Before going to bed, brush your teeth. |
| After | After finishing your homework, you can watch TV. |
| While | While driving, keep your eyes on the road. |
| Since | Since the weather is nice, let’s go for a walk. |
| Until | You can play video games until it’s time for dinner. |
| Unless | You can’t come with us unless you finish your work. |
| If | If it rains, we’ll stay inside. |
| When | When you see the green light, you can cross the street. |
| Whenever | You can call me whenever you need help. |
| As soon as | I’ll start cooking dinner as soon as I finish this project |
Incorporating these modifiers into your prompts can help ChatGPT better understand your intent and generate more accurate and relevant responses. So, don’t be afraid to get creative with your modifiers and experiment with different combinations to see what works best. Take a look at a few examples and try to find the modifier in the prompt.
Positive Modifiers:
| Type of Modifier | Example Prompt |
|---|---|
| Adjective | “Write a heartwarming story about a loyal dog.” |
| Adverb | “Explain how to solve a math problem quickly and efficiently.” |
| Comparison | “Describe how the new smartphone model is better than the previous one.” |
Negative Modifiers:
| Type of Modifier | Example Prompt |
|---|---|
| Adjective | “Write a scary story about a haunted house.” |
| Adverb | “Explain why procrastination is harmful to productivity.” |
| Comparison | “Describe the disadvantages of using social media excessively.” |
Quantifier Modifiers:
| Type of Modifier | Example Prompt |
|---|---|
| Universal | “Explain why all humans need to sleep.” |
| Particular | “Give an example of a time when some people act irrationally in a group.” |
| Indefinite | “Describe a situation where someone may feel somewhat overwhelmed.” |
Temporal Modifiers:
| Type of Modifier | Example Prompt |
|---|---|
| Sequential | “Explain the steps to make a cup of tea.” |
| Durative | “Describe what it’s like to live in a foreign country for a year.” |
| Momentary | “Describe a moment when you felt proud of yourself.” |
Conditional Modifiers:
| Type of Modifier | Example Prompt |
|---|---|
| Hypothetical | “If you could time travel, where and when would you go and why?” |
| Causal | “Explain how social media affects mental health.” |
| Concessive | “Despite the challenges, describe why becoming a parent is rewarding.” |
To help illustrate the importance of modifiers, let’s consider a few examples. Imagine you want ChatGPT to write a recipe for apple pie. A prompt like “Write a recipe for apple pie” might generate a response that’s too general and unfocused. However, a prompt like “Write a recipe for a classic apple pie with a flaky crust and cinnamon-sugar topping” incorporates positive and quantifier modifiers to better specify the desired outcome (if that’s what you want).
You can even use another prompt engineering method, iteration. If you put the two prompts into ChatGPT you will find that the recipes are similar. The main addition is the topping. so you can now iterate this a little bit more by asking for instance ChatGPT to tell you what the best apple pie is
What’s the best apple pie?
Here are a few examples:
- “Mile High Apple Pie” by Rose Levy Beranbaum – This recipe features a tall, flaky crust and a filling of thinly sliced apples with a blend of cinnamon, nutmeg, and allspice.
- “Deep Dish Apple Pie” by America’s Test Kitchen – This recipe uses a double crust and a mixture of Granny Smith and Golden Delicious apples for a balance of tart and sweet flavors.
- “Salted Caramel Apple Pie” by Sally’s Baking Addiction – This recipe adds a layer of salted caramel sauce to the filling and incorporates a crumbly streusel topping for added texture.
So let’s change our prompt again.
Please write a recipe for an apple pie by combining the recipe for, “Salted Caramel Apple Pie” by Sally’s Baking Addiction” and “Mile High Apple Pie” by Rose Levy Beranbaum. The recipe should include the salted caramel sauce.
As you can see prompts can become quite long and involved to generate what you actually want.
Incorporating temporal and conditional modifiers can also improve the effectiveness of your prompts. For instance, if you want ChatGPT to generate a travel itinerary, a prompt like “Plan a trip to Paris” could result in a generic itinerary that doesn’t fit your specific needs or interests. However, a prompt like “Plan a week-long trip to Paris in June with a focus on art museums and local cuisine” provides more context and specificity for ChatGPT to work with.
So, in summary, using modifiers in prompt engineering can significantly improve the quality and relevance of ChatGPT’s responses. Experiment with different types of modifiers and see how they impact the outcomes of your prompts. And remember, specificity is key!
Modifiers For Creating Content
When communicating with others, we often adjust our language to fit the situation. We might speak differently with a close friend compared to a potential employer, for example. Similarly, when using ChatGPT, we can request that it write in a specific tone or for a particular audience. These modifiers can have a significant impact on the resulting output.
Expertise Modifiers
Asking ChatGPT to respond as an expert in a particular field is a very powerful modifier. You can read my article about it here.
| Expert Persona | Prompt |
| Great Mathematician | I want you to reply like one of the greatest mathematicians of all time. What is a prime and why are they useful? |
| Great Psychologist | I would like you to respond as one of the greatest psychologists, “What is REM sleep? |
| Great Email Marketer | I would like you to respond as one of the greatest email marketers, “What are some best practices for creating effective autoresponders?” |
Let’s consider tone. Tone refers to the attitude or emotion conveyed in writing. For example, you might request that ChatGPT writes in a formal tone for a business report or in a playful tone for a children’s book. By adjusting the tone, you can make the output more engaging and appropriate for the intended audience.
| Tone | Prompt |
|---|---|
| Formal | Write a formal press release announcing the launch of our new product. |
| Informal | Write a casual email to a friend inviting them to your birthday party. |
| Friendly | Write a friendly message to a coworker congratulating them on their promotion. |
| Authoritative | Write an authoritative essay discussing the effects of climate change on marine ecosystems. |
| Persuasive | Write a persuasive speech advocating for the importance of arts education in schools. |
| Humorous | Write a humorous short story about a talking animal. |
| Sarcastic | Write a sarcastic review of a terrible movie. |
| Sad | Write a letter from a character in a novel expressing their sadness over the death of a loved one. |
| Angry | Write a social media post expressing your anger over a recent political decision. |
| Hopeful | Write a blog post discussing the positive impact of community gardens on urban environments. |
Next, we have education level modifiers. These modifiers adjust the complexity and technicality of the language used in the output. For instance, you might ask ChatGPT to write in a simplified manner for a general audience or in a technical manner for a specialized audience. These modifiers can ensure that the output is understandable and informative for the intended readers.
| Education Level | Prompt |
|---|---|
| Elementary School | Write a short paragraph explaining how plants grow for elementary-level students using simple words and short sentences. |
| Middle School | Write a brief summary of the causes and effects of World War II for middle school students using appropriate vocabulary and examples. |
| High School | Write an essay analyzing the symbolism in “The Great Gatsby” for high school students using literary terms and critical analysis. |
| College | Write a research paper exploring the impact of climate change on global food production for college students using current scientific research and statistical data. |
| Postgraduate | Write a dissertation on the ethical implications of artificial intelligence for postgraduate students using in-depth analysis and extensive research. |
Lastly, audience modifiers can significantly impact the output. These modifiers take into account the specific group of people that the output is intended for. For example, you might ask ChatGPT to write for a specific age group, geographic location, or profession. These modifiers can help ensure that the output is relevant and relatable for the intended audience.
| Type of Modifier | Prompt Examples |
|---|---|
| Audience | Write a short article explaining the importance of recycling for a group of environmentally conscious teenagers. |
| Age Group | Create a 10-minute educational video on basic math concepts for 1st graders. |
| Geographic Area | Write a travel blog post about the top 10 attractions in New York City for tourists visiting from Europe. |
| Profession | Write a technical report on the latest advancements in artificial intelligence for a group of software engineers. |
Overall, by requesting ChatGPT to write with specific tone, education level, and audience modifiers, we can create output that is engaging, understandable, and relevant for the intended readers. So don’t be afraid to experiment and see how these modifiers can enhance the effectiveness of your prompts!
Context Context Context
Before we dive into the differences between novice and expert prompt engineers, it’s important to understand the role that context plays in the use of modifiers. Context refers to the situation or environment in which something occurs. When it comes to using modifiers in prompts, the context can greatly affect which modifiers are appropriate and effective.
Except for Beelzebub, he’s still a bit of a jerk.”
For example, let’s say you’re writing a prompt for a medical article. In this context, using technical language and complex modifiers would be appropriate and effective, as the audience is likely to have a high level of medical knowledge. However, if you were writing a prompt for a general interest article about health, simpler language and more neutral modifiers would be more appropriate and effective, as the audience is likely to have a lower level of medical knowledge.
Examples: Novice Prompt Engineers vs. Expert
| Prompt | Novice Prompt Engineer | Expert Prompt Engineer |
|---|---|---|
| Write a blog post about the benefits of exercise. | Simply write a blog post about exercise without any specific goals or audience in mind. | Tailor the blog post to a specific audience, such as middle-aged adults or beginners, and focus on specific benefits that are most relevant to that group. Use engaging and compelling language to keep readers interested. |
| Create an email to promote a new product to customers. | Write a basic email introducing the product and providing basic details about it. | Use persuasive language and highlight the unique features and benefits of the product. Include a call to action that encourages customers to take a specific action, such as making a purchase or signing up for a free trial. |
| Develop a survey to gather feedback from employees. | Create a survey with basic questions that ask for simple answers. | Tailor the survey to specific areas of concern, such as employee satisfaction, engagement, or opinions about new policies. Use a mix of open-ended and multiple choice questions to gather both quantitative and qualitative data. |
| Write a social media post for a new restaurant opening. | Create a basic post that simply announces the opening and provides the location and hours of operation. | Use engaging language and highlight unique features of the restaurant, such as signature dishes or a particular theme. Include photos or videos to showcase the atmosphere and menu items. |
| Develop a presentation on a new marketing strategy. | Create a basic PowerPoint presentation that outlines the new strategy and provides some basic data to support it. | Tailor the presentation to the specific audience, such as the marketing team or company executives, and use engaging visuals and language to effectively communicate the strategy and its potential impact. Include specific goals and metrics for measuring success. |
One key difference between novice and expert prompt engineers is their use of modifiers. Novice prompt engineers may overuse modifiers, using too many or overly complex modifiers that can make the prompt confusing or difficult to understand. They may also struggle to identify which modifiers are appropriate for different contexts and audiences.
Expert prompt engineers, on the other hand, have a more nuanced understanding of which modifiers are appropriate and effective for different contexts and audiences. They can quickly identify the most effective modifiers for a given prompt, and use them in a way that enhances the overall quality and effectiveness of the prompt.
The path between novice and expert is clear, it amounts to experience and testing. There are ways to get ChatGPT to help you write very effective prompts. I’ll go into that in another article, but to get the best prompts that you still need I have your own experience.
Challenges to Look Out For
Alright, let’s talk about some of the challenges you might encounter when working with modifiers and ChatGPT. While modifiers can greatly enhance the quality and specificity of your prompts, they can also cause some complications if not used properly. Here are a few things to watch out for:
- Over-reliance on modifiers: It can be tempting to try and modify every aspect of your prompt to make it as specific as possible, but overusing modifiers can actually hinder ChatGPT’s ability to generate coherent and accurate responses. Try to strike a balance between using enough modifiers to guide ChatGPT, but not so many that it becomes overwhelming.
- Ambiguity in modifiers: Not all modifiers are created equal, and some can be more ambiguous than others. For example, the modifier “large” can mean different things depending on context – is a large car the same as a large elephant? Make sure your modifiers are clear and specific enough to avoid confusion.
- Incorrect use of modifiers: Using modifiers incorrectly can lead to inaccurate or nonsensical responses from ChatGPT. For example, if you use the modifier “not” in a way that contradicts the meaning of the rest of the prompt, it can cause ChatGPT to generate a response that doesn’t make sense.
- Lack of diversity in modifiers: Using the same modifiers repeatedly can lead to repetitive and uninteresting prompts. Try to mix things up and use a variety of modifiers to keep your prompts fresh and engaging.
Remember, using modifiers effectively is all about balance and specificity. Keep these challenges in mind and you’ll be well on your way to creating high-quality prompts that generate accurate and interesting responses from ChatGPT.
What NOT to Do
| Prompt | Problem |
|---|---|
| Write a blog post on “The Best Ways to Stay Healthy” for all ages | Over-reliance on modifiers: the prompt is too broad and lacks specificity, making it difficult to provide targeted and useful information |
| Create a social media post promoting a new product that is “Better than the Competition” | Ambiguity and modifiers: the use of the word “better” is subjective and could mean different things to different people, leading to confusion and potentially misleading information |
| Write a press release announcing a fantastic groundbreaking new partnership with a technology company | Incorrect use of modifiers: using too many modifiers, such as “revolutionary” or “game-changing,” without providing clear evidence or explanation can come across as hyperbolic or untrustworthy Additional context needs to be added to the prompt. |
| Create a flyer advertising a job opening for a “Young and Energetic Sales Associate” | Lack of diversity and modifiers: using words like “young” can create implicit bias and exclude qualified candidates who may not fit a certain age or demographic. Additionally, the use of the word “energetic” can perpetuate ableist attitudes by assuming that all individuals with disabilities are not able to perform certain tasks or have a certain level of energy. This can cause EOC problems for companies. |
The End, Well Almost
Alright, let’s wrap things up!
.Recap of Main Points
Throughout this article, we’ve covered the importance of modifiers in ChatGPT prompt engineering. We discussed how modifiers can add clarity and specificity to prompts, making them more effective in communicating information to ChatGPT. We also looked at common challenges to be aware of when incorporating modifiers, such as over-reliance, ambiguity, incorrect use, and lack of diversity.
Significance of Modifiers in ChatGPT Prompt Engineering
As we’ve seen, modifiers play a crucial role in crafting effective prompts for ChatGPT. They help ensure that the language model understands exactly what information is being requested and can provide accurate and relevant responses. This is especially important in fields like customer service, education, and healthcare, where prompt accuracy is critical for achieving positive outcomes.
Future Implications of ChatGPT Prompt Engineering with Modifiers
As ChatGPT and other language models continue to evolve and become more sophisticated, the use of modifiers will become increasingly important in prompt engineering. With natural language processing becoming more human-like, modifiers will help ensure that ChatGPT understands the nuances of language and can provide the most relevant and accurate responses. It’s an exciting time for language technology, and the future of ChatGPT prompt engineering looks bright!
Thank you for reading 🙂