Journal Keeper
Using AI to Support Your Journaling Habit
Quick Access
For those who just want to jump in to try out this particular AI solution, here’s a quick link to access a GPT that I created using OpenAI’s GPT building tools: Journal Keeper GPT. As of writing this article, you will need an OpenAI ChatGPT Plus subscription to access this custom GPT.
Don’t Wait for Someone Else to Build It
The advent of large language models is opening the door to many new applications. But we don’t have to wait for new startups or established technology companies to create the applications for us to use. The flexibility of the underlying AI allows for each of us as everyday users to create our own personalized applications of the AI with just a few minutes of preparation using the existing AI tools already available on the market.
In the “Everyday Use of AI” newsletter, I will describe many possible ways in which you can use AI to help with daily activities in your life. For this particular entry in the newsletter, I will discuss how you can use AI to help you with you daily journaling habit.
Journal Keeper
Maybe you’ve already built a habit of daily journaling, or perhaps you are someone who wants to start this new habit. AI can provide you with a quick and easy way to do your journaling.
In this post, I will describe how to quickly prompt the AI to support your needs. In this example, I’ll be showing the steps using OpenAI’s ChatGPT Plus but you may find that this same approach can be used with other LLMs hosted by other companies.
The above picture is a snapshot of the experience. In this simple AI application, you can provide the system with your journal entry, and it will record it into a nicely formatted Word document that you can download at a later time. Every time you give it a new entry, it will append that into the same Word document along with the date and time thus allowing you to create a long-running journal without needing to boot up Word every time you want to add an entry to your journal.
Note that the AI will ask for your time zone at the start of the conversation - just once - to be sure it is writing the date and time correctly into the Word document. That’s a capability I prompted it to do as you’ll see in the prompt exploration below. Without this custom prompting, the AI would use the server’s system time - which isn’t quite what we want as users.
Downloading the journal document provides the following document content…
The Prompt
This AI application is really just a simple system prompt that I authored along with some configuration of OpenAI’s GPT capabilities.
Here’s a copy of the prompt:
Your role is to act as a journaling assistant, helping users record their journal entries into an ongoing Word document using your code interpreter capabilities. Whenever a user provides a message, evaluate whether it is intended to be a journal entry. If it is, add it to a Word document, creating one, if necessary, where all the user's journal entries will be stored for future download. If you're unsure whether a message is a journal entry, ask the user to confirm. If you're certain it's not a journal entry, respond to the user appropriately. Your behavior should prioritize understanding and accurately categorizing user inputs as journal entries or general queries and handling them accordingly. Ask the user once when first starting the conversation what their time zone is in order to accurately record the time of their entry. The first time you create a Word document, make sure you format it nicely to look like a personal journal.
You don’t need to create your own custom GPT in order to get similar behavior. You could use the above prompt when chatting with ChatGPT Plus. Just make sure you have the appropriate model selected that includes the “analysis” capabilities. That will ensure that the AI can leverage some core code-writing functionality that will be needed to create the Word document.
Wrapping It Up
As you can see, getting the AI to do a specialized set of tasks that support your everyday healthy living is a pretty easy process once you understand some of the basics of prompt engineering. Creating a good prompt just requires a little practice and knowhow. We’ll get into the details of prompt engineering more in future posts on “Everyday Use of AI” - your newsletter for pragmatic, useful, and healthy applications of AI.
