Free ChatGPT Crash Course for Beginners
Are you looking for an ultimate resource to learn how to use ChatGPT properly? Well, you've come to the right place.
In this free ChatGPT crash course, we'll learn everything there is to know about this revolutionary tool. And when you're done reading this course, you can call yourself a ChatGPT power user.
But before starting with all the exciting stuff, let's take a step back and look at how ChatGPT came into our world.
Introduction to Generative AI
As we know, artificial intelligence (AI) is a field that uses data, and machine learning to solve complex problems.
This course focuses on Generative AI (or Gen AI for short), which is a subset of AI.
How to treat Gen AI
Gen AI can make new things as humans do. It can write, solve problems, and even make art. But, in order to do so, it needs clear instructions written by humans. You can think of Gen AI as an intern who needs constant guidance.
We interact with Gen AI using prompts. And, prompts can be in any form, like text, picture, sound, or any other medium. If it's starting to get a little bit complicated, don't worry, it'll all make sense when we start looking at real-life examples.
Why is Gen AI Important?
Gen AI can do creative jobs that had to be done by humans before. This simply means you can save time and money if you can explain your job to Gen AI.
Some ways people are using generative AI are:
As you can imagine, possibilities are literally endless. And as I like to say, the only limitation of Gen AI is yourself.
How to Start With Generative AI
If you're interested in the field of generative AI and you'd like to learn more, you came to the right place. Because in this crash course, we will teach you all the basics you need to know to use generative AI properly.
Just so you know
During this course, we'll be using a tool called ChatGPT to learn and demonstrate how generative AI works.
Even if you're just a beginner, or you're a proficient ChatGPT user already, there'll be knowledge in this course for everyone! So without any further ado, let's dive right in!
Prompting With ChatGPT
In the previous section, we discussed Gen AI and how humans can instruct it to perform various tasks. This process is called prompting, and in this section, we will delve deep into the realm of ChatGPT prompting.
Brace yourself!
In the following sections, we will explore how to utilize ChatGPT to its fullest potential.
To kickstart your journey with ChatGPT, follow these easy steps:
Basic ChatGPT Prompts
Now that you are all set up, let's dive into the fascinating world of prompts and put ChatGPT to the test!
Use ChatGPT to Summarize a Book
Imagine you were given a task to read a book and describe what it's about. But you're not really in the mood to read it. By using ChatGPT, you can effortlessly get a summary.
Solve Math Problems
ChatGPT can also solve math problems! Let's demonstrate it on a simple problem.
As you can see, ChatGPT can help with simple tasks, and the applications are rather straightforward. You simply define a task, and ChatGPT returns the result. But as we'll find out in the next section, it doesn't always work like this.
Advanced ChatGPT Prompts
Sometimes, ChatGPT gives an incorrect response. It can forget to add details, provide incorrect information, or simply return the response you're not satisfied with.
Let's demonstrate this with an example. Imagine we want to use ChatGPT to write a humorous tweet about politics in America.
The response certainly satisfies the prompt that we gave to ChatGPT, but it's not humorous, and it doesn't mention anything about America. Let's try to change the original prompt a bit.
This is much better, it's funny, it mentions America, and on top of that, ChatGPT included hilarious hashtags.
You Should Remember
The process of refining prompt over time is known as prompt engineering.
You will never write the perfect prompt on your first try, so it is important to get good at refining your prompt. Being good at refining the prompt mostly comes from lots of practice, and trial and error.
Prompting Techniques
Now that we've covered some of the basic and more advanced prompts, let's take it one step further. In this section, we'll dive deeper into the art of prompting. You'll get to know different types of prompting techniques and their applications.
Let's start with the first one!
Instructions Prompting
Instruction prompting allows us to give specific directions (instructions) to ChatGPT, and let it do the task for us. Let's demonstrate that with practical examples.
Format Data
Data entry and manipulation are all around us, and it takes a lot of human labor to do so. Let's see if ChatGPT can handle it. Imagine we want to parse unformatted data into an Excel sheet.
In this example, the prompt takes raw data in the form of plain text and turns it into an Excel sheet.
Remove Personal Information
Prior to sharing private documents, organizations often manually hide sensitive information. With ChatGPT, we can automate this process and eliminate the need for labor-intensive human effort.
In this example, the prompt successfully replaces the names, numbers, and emails, with appropriate placeholders. And all of that by simply describing a task itself.
Not limited to simple tasks!
Instruction prompting can also be used for complex evaluations.
Evaluate Essays
ChatGPT can evaluate and provide feedback on essays based on various criteria such as grammar, clarity, coherence, argument quality, and evidence usage.
By prompting the AI with an essay excerpt and requesting feedback, you can gain valuable insights for improving the essay's quality.
In conclusion, instruction prompting offers an incredible approach to harnessing the capabilities of Gen AI. By giving specific instructions, you can accomplish complex tasks that would require human labor. And you can do it in a matter of seconds.
Role Prompting
Role prompting is a fascinating technique that allows us to control and shape the style and accuracy of AI-generated text. Whether you're writing a review, drafting an email, or solving a math problem, role prompting can help you achieve more tailored and targeted results.
Use role prompting to style the response!
This involves asking the AI to pretend to be a certain person, or act in a certain way, thus modifying how it writes based on the assigned role.
Restaurant Review Example
Imagine you're writing a review of a pizza place on Google Reviews. You had a nice experience in a restaurant, but you're not in the mood to write a long review, so you simply ask ChatGPT to do so.
The result is pretty good already, but let's see what happens when the AI assumes the role of an experienced food critic.
As you can see, the result is completely different. It adds more details, uses different language, and seems way more professional.
Facebook Ad Example
In this example, we're trying to use ChatGPT to write a Facebook ad copy for us. Let's see how the results will differ by using different roles.
In publishing, advertising and related fields, copy refers to written material, in contrast to photographs or other elements of layout, in books, magazines, newspapers and advertising. (source: Wikipedia)
As you can see, the Facebook ad contains a lot of funny elements because we instructed the model to act as a comedian. Now, let's see how the response will change when the model acts as an experienced copywriter.
It's important to say that both results generated by ChatGPT were good quality. However, there was a clear difference in the style of the results. All because of the role prompting.
In conclusion, role prompting is a powerful strategy for shaping generated AI text. It allows us to control the style, tone, and depth of the generated text, making it more suitable for specific contexts and audiences.
Whether it's writing reviews, drafting emails, or solving math problems, role prompting can significantly enhance the quality and accuracy of the results.
Tip!
As we explore the capabilities of AI, role prompting will continue to be a key technique in prompt engineering.
Few-Shot Prompting
Few-shot prompting, also known as "in-context learning", allows models to learn from a small set of examples or shots. Unlike zero-shot prompting, which relies solely on a provided prompt without any examples, and one-shot prompting, which involves a single example, few-shot prompting leverages the power of multiple examples to guide the AI model.
To see the power of few-shot prompting in action, let's consider a practical example.
Classify Reviews Examples
Suppose we want to classify customer feedback as positive or negative. Using few-shot prompting, we can show the model 2 examples of positive and negative feedback, followed by new unclassified feedback.
The model, after observing the classification of the initial examples, can utilize this information to classify the new feedback properly.
Example strucute is important
Because we structured the classification in a single-word negative or positive, the model also generates a single-word response, rather than a complete sentence such as this review is positive.
Few-shot prompting is a powerful and effective strategy for AI models to generate accurate and structured responses.
By leveraging the input of multiple examples, few-shot prompting enables models to comprehend the desired output format more reliably. Compared to role prompting, or instructions prompting, few-shot prompting stands out as a superior approach in most cases.
Combining Prompting Techniques
In previous sections, we explored different types of prompting techniques. They ranged from providing context and instructions to including multiple input-output examples.
However, to create even more effective and impactful prompts, we can combine all these prompting strategies together. And by doing so, we can unlock the full potential of AI-generated responses. Let's demonstrate this in an example.
AI Tweet Writer Example
Let's say we want to use AI to write a short once-sentence tweet on a topic of our choice.
Concluding Thoughts
You've made it!
In this free ChatGPT crash course, we introduced the topic of generative AI and explained why it's important to adopt this technology.
You also learned that we use instructions called prompts to interact with the AI. We also went over the different prompting techniques, and we used examples to show how to use each technique properly.
With all the knowledge you just gained, you're ready to embark on an exciting journey into the realm of Generative AI!