Learning a programming language traditionally begins with a Hello, World. example.
Today, we will follow that tradition and open the door to a conversation with a large language model.
Open the official product you use and enter this Prompt:
Hello, ChatGPT.
You may see a reply like this:
Hello! How can I assist you today?
Throughout this series, the content we send a model is called a Prompt, and the content the model returns is called its output.
If your reply is not exactly Hello! How can I assist you today?, do not worry.
Why?
Because a generative model does not always produce identical output. Even when the text you send looks the same, the model, context, and sampling process can change the wording.
That does not mean your Prompt is necessarily wrong.
You have now completed your first Prompt.
It looks simple, does it not?
Starting with this short exchange, the next chapters look at what happens after text enters a model and why the result can change from one run to another.
Note: Use public or fictional material while practicing. Do not submit passwords, API keys, trade secrets, identity documents, or unnecessary personal information.
Previous: What Is Prompt Engineering?
Next: What Is a Token?
Independent product builder · AI, agents and real workflows.