By this point, we all know what ChatGPT is.
If you aren’t aware for whatever reason, simply put, its a text-based generative AI platform that uses what is known as large-language models (or a large set of text data) to give you output of information based on a series of inputs (or ‘prompts’).
Despite only having been released publicly last November, in just a matter of months, ChatGPT has grown to be one of the fastest-growing platforms on the Web.
In February of this year,
- ChatGPT reached over 100 million users
- In just 5 days of its release, ChatGPT surpassed 1 million users
Whether you’re an avid user or just simply heard of its name and never gave it a second thought, there is no doubt that the release of ChatGPT and other similar platforms is a huge milestone in transforming how we interact with computers.
Just as how Google changed how we find information, ChatGPT and other generative AI chatbots have the potential to change how’ll inevitably get information, do certain jobs, and learn.
So why is it so impactful?
Imagine having a digital assistant that can write content, answer questions, provide recommendations, and engage in relatively intelligent conversations.
We can use ChatGPT to create chatbots, develop creative content, improve customer service, help with coding, and so much more.
But maybe you’ve used it, and thought, it wasn’t that great.
While it definitely still does have its limitations, you need to remember it still has been less than a year since its release. And with every new iteration that comes out, things are only going to get smarter.
What many people are missing about using ChatGPT is knowing exactly how to talk to it.
Honestly, there is just so much you can do with it — but only if you write the right way.
And to do this, you should first learn how to write prompts.
Learning to talk to ChatGPT
You might be thinking: what’s so hard about writing prompts? I just instruct it to tell me what I want and it’ll give me the answer, no?
Well, not necessarily.
What many people are missing about ChatGPT, and why there are loads of critics, is that the power of ChatGPT lies in conversation, not instruction.
Writing prompts effectively requires finesse.
A well-crafted prompt can pull out the information you genuinely seek and after enough back-and-forth, you learn to guide the AI model toward a desired outcome.
The most common misconception about the use of prompts and ChatGPT is that prompts are just sentence starters. A lot of users are unaware of the profound role their own words play in the direction of the response. ChatGPT is sometimes seen as some grand oracle that always produces amazing answers, regardless of what you put.
But that definitely is not the case.
Just as how we say things matter to whether a person is receptive to the thing we say, talking to an AI, is kind of the same.
Think about ChatGPT like a genie.
If a genie asks you for a wish, and you simply state ‘give me more money,’ it may grant the wish, but perhaps you might end up having a lot of money but you’re a notorious drug dealer wanted by the government. Or perhaps you have all the money in the world, but are dealing with terrible health, unable to spend the money itself.
So, it’s really a matter of specificity.
Why you should learn prompting
This is exactly why a field like prompt engineering exists.
For those who aren’t aware, prompt engineering is a recent field that has been popping up since this generative AI boom, with some jobs offering a starting salary of 300,000 USD.
It is the process of carefully designing and refining prompts or instructions given to ChatGPT. It empowers us to shape and lead to more accurate, contextually relevant, and meaningful responses.
Remember: crafting prompts is not just about providing sentence starters but skillfully guiding the AI to generate the desired outputs.
We all can speak, but its those who speak well that are considered great communicators and get their way in life. It’s no different with prompting — knowing how to control language and harness the true capabilities of AI can do wonders.
Now, I’m not saying that you need to get into prompt engineering. What I am saying is that knowing how to write prompts is going to be a useful skill to have.
The Current Untapped Market
In spite of ChatGPT gaining significant attention, prompt engineering represents a market that still has immense potential.
Despite having millions of users as mentioned earlier, many businesses and individuals are still unfamiliar with its potential. This means there’s a huge untapped opportunity to get ahead of the game and learn to position yourself at the forefront of this emerging field.
Why Learn Prompt Engineering Now
By mastering prompt techniques, you learn to optimize and fine-tune the model’s behavior to suit your specific needs, getting better results vs ones that are so-so.
According to the Gartner Hype Cycle, it typically takes about two years for new technologies to adapt and mature. ChatGPT and other generative AI have only been publicly around for less than a year.
So by familiarizing yourself with some basic skills on how to write better prompts, you position yourself ahead of the curve.
7 prompt patterns to help you use ChatGPT
So with all that being said, let’s jump straight into 7 basic prompt patterns that you can uset o speak to ChatGPT more effectively.
Completion Patterns:
Completion patterns involve providing a partial sentence or phrase that the model needs to complete. It basically prompts the model to generate the remaining information based on a given context.
- How to use: Specify the initial part of a sentence or phrase you want the model to complete.
- When to use: Completion patterns are useful for tasks such as text generation, creative writing, or filling in missing information.
- Most useful for: Writers, content creators, creative thinkers.
- Example: “Once upon a time, in a land far, far away…”
Key tip: Its probably best to give some context first, before jumping straight into it.
Instructional Patterns:
Instructional patterns instruct the model on how to approach a given task or provide specific guidelines for getting a desired output.
- To do this well, you should clearly state the instructions or guidelines for the model to follow.
- When to use: Instructional patterns are effective when you need to direct the model on how to perform a specific task.
- Most useful for: Task-oriented prompts, guided writing.
- Example: “Write a persuasive essay arguing in favor of renewable energy.”
Comparison Patterns:
Comparison patterns ask the model to compare or contrast multiple options or concepts.
- How to use: Present the options or concepts to be compared and prompt the model for its analysis.
- When to use: Comparison patterns are valuable when you need the model to evaluate and discuss the similarities and differences between two or more subjects.
- Most useful for: Comparative analysis, pros and cons assessments.
- Example: “Compare the advantages and disadvantages of cloud computing versus on-premises servers.”
Question-Answering Patterns:
Question-answering patterns involve posing a specific question to the model, requiring it to generate a relevant and informative response.
- How to use: Ask a clear and specific question for the model to answer.
- When to use: Question-answering patterns are suitable when you need the model to provide concise answers to specific inquiries.
- Most useful for: Q&A platforms, information retrieval.
- Example: “What is the capital of France?”
Opinion or Argument Patterns:
Opinion or argument patterns prompt the model to express a subjective opinion or make a persuasive argument about a particular topic.
- How to use: Frame the prompt in a way that invites the model to share its perspective or justify a stance.
- When to use: Opinion or argument patterns are useful when you want the model to provide subjective viewpoints or engage in persuasive writing.
- Most useful for: Debates, opinion pieces, persuasive essays.
- Example: “Argue for or against the use of genetically modified organisms (GMOs) in agriculture.”
Scenario Patterns:
Scenario patterns involve setting up a specific situation or scenario for the model to respond to or generate a story around.
- How to use: Provide the context, characters, and relevant details of the scenario.
- When to use: Scenario patterns are effective when you want the model to create narratives, dialogues, or responses within a given context.
- Most useful for: Creative storytelling, fictional writing.
- Example: “You wake up one morning to find yourself in a mysterious, deserted town. Describe your experience and what you encounter.”
Command Patterns:
Command patterns direct the model to perform a specific action or task.
- How to use: Clearly state the command or action you want the model to execute.
- When to use: Command patterns are suitable when you need the model to follow specific instructions or perform a function.
- Most useful for: Task-oriented prompts, executing specific actions.
- Example: “Translate the following English text into French: ‘Hello, how are you?'”
Final thoughts
This video is in no way saying that prompt engineering will be a job or skill that will set you up for security in the future. With all the advancements of AI, anything and everything can go away in an instant.
However, whether it’s through Google’s Bard, Microsoft’s Bing Chat, or ChatGPT, one thing is for certain: generative AI is making its waves.
And whether its to help you learn a new skill faster, inspire you to write better, help you work smarter, or simply assist you in making money online, just knowing these simple tricks on writing prompts will definitely be helpful for the coming months or years.
While this video was more of a beginner overview to prompts, I’ll continue to release videos like this that go more in the context of marketing or design or making money online. So stay updated by subscribing to this channel so I can continue to help you navigate this digital ocean.
