T@b1e 0f C0ntent5
Let’s Get This Show Started!
Have you ever asked ChatGPT a question or requested a rephrasing or summary of a text, but the response was too long, and you only needed it in 3 - 4 lines? Then, you try reentering a prompt to shorten it, but the result comes as bullet points instead of a numbered list, and you have to prompt again to fix it. Isn't that time-consuming? Doesn't AI automation aim to minimize repetitive tasks and help you focus on the core functionality?
Hello their I’m semo, and this is my fourth technical article, and in it, I aim to explain how we can control ChatGPT's responses to align with our desired results and output using tags.
Brain Upgrade: ChatGPT Gets a Memory Makeover!
If you've ever prompted ChatGPT with a question, you might have noticed it mentioning something called update memory. But what exactly does that mean?
Update memoryletsChatGPTsave and adjust details about you or topics. This helps it give more accurate and consistent responses, based on each user preference.
Data on a Diet: Slimming Down the Memory for a Faster You!
So the question that arises here is: how can we assist ChatGPT with that? To do this, click on your profile icon in the upper-right corner, go to the settings section, navigate to the personalization section, and click on the manage button.
You'll find a lot of data there, representing the memories ChatGPT has saved about you to improve its responses. You can review this data, delete it entirely, or remove specific items based on your needs. However, in this article, I'll focus on how to create tags that align with specific prompts, eliminating the need to rewrite the entire prompt each time.
Tag, You’re It: Data's New Favorite Game!
I will open a new ChatGPT session and input the following prompts, you can update them later, but for now use the underline prompts as they are provided.
- Prompt One:
Update Memory:
When I use the tag [red] with the [text], you should summarize it in the following format: Headers Bullet points Use backticks (`) for quotes instead of quotation marks Replace > with →
- Prompt Two:
Update Memory:
When I use the tag [green] with the [text], you should summarize it in a maximum of 4 lines in bullet points.
- Prompt Three:
Update Memory:
When I use the tag [yellow] with the [text], you should rephrase it into a 6-line paragraph and provide a maximum of two example lines.
And so on. Now that you have an idea of how to update ChatGPT’s memory through the prompting section and use tags to influence its output behavior, you can apply the same approach to customize results based on your field of study, job role, or any other specific details. Be creative in how you tailor the responses, whether in general or with more specific details.
Now, returning to the personalization section we've explored, let's review the memory updates we've made:
As we can observe, we've reviewed the three related memory updates that we’ve made and saved within the updated memory section.
Lab Rats in Action: Testing and Tweaking!
I’m going to input the below text as an example into a new ChatGPT session, using this format: [tag name] [text provided], and then do the same for [green] and [yellow] to check the corresponding results we have.
- Used Text as an example:
- Prompting with tagging system formula:
Cyber Threat Intelligence (CTI) plays a crucial role in enhancing cybersecurity by providing actionable insights. It involves the collection, analysis, and sharing of data related to potential threats. CTI helps organizations understand emerging attack patterns and tactics used by cybercriminals. With real-time information, security teams can identify and respond to threats before they escalate. By incorporating CTI into defense strategies, businesses improve their ability to predict and mitigate risks. This intelligence can be integrated into security tools like firewalls and intrusion detection systems for more effective monitoring. Threat intelligence sharing among organizations further strengthens collective security efforts. Ultimately, CTI is vital for staying ahead of evolving cyber threats in an increasingly connected world.
[tag name] [text provided]Now, we will check the related results for each tagging name we provided with the previous text. Feel free to use any other text you have as well.
[red]tag answer:[green]tag answer:[yellow]tag answer:
We have implemented a double automation process. The first part uses AI to summarize the text we provide, and the second part automates the prompts that we send to ChatGPT.
The End... Or Is It?
If you’ve reached this section, thank you for reading the article. I hope we can connect on LinkedIn linkedin.com , and feel free to send me the automated prompts you’ve created.
Written By: Sameer Fakhoury aka.semo