Image from OpenAI GPT's main view. 
 

In our rapidly evolving digital world, artificial intelligence (AI) is not just a buzzword but a revolutionary force reshaping how we interact with technology.

Ever since ChatGPT was first launched, there’s not been a single week without a big leap forward in the AI field.

Just a week ago, I got my hands on OpenAI’s shiny new toy, GPTs (personalized ChatGPT versions), unveiled at the latest OpenAI DevDay.

Missed the event? No worries!

Here’s a quick scoop: GPTs are these cool, customizable ChatGPT versions that you can create without writing a single line of code.

So stay with me and let’s try to discover it all together.

 

1. Getting Started with OpenAI’s GPTs

 

I’m about to take you on a DIY journey to craft your own GPT, perfect for those eyeing taking advantage of AI to boost their productivity. 

For ChatGPT Plus subscribers, this feature’s rolling out as we speak.

The recent launch of customizable Generative Pre-trained Transformers (GPTs) by OpenAI marks a significant milestone in AI’s journey towards user-centric design.

With these developments, creating a personalized ChatGPT instance is no longer a complex task—it’s now accessible to everyone, opening a world of possibilities for tailored AI interactions.

Post-update, ChatGPT’s got a new look, packed with features like web browsing, DALL-E, and code interpreter, all under the GPT-4 umbrella.

The only feature that has been left out is plugins, that are still an option of their own.
 

Image by Author 
 

The new GPTs feature can be found within the “explore” button. 

So press the button and…

 

2. Let’s Get Building! 

 

Here you will find all your created GPT versions, the option to create a brand-new GPT with “Create a GPT” and with the OpenAI’s library Made by OpenAI. 

 

Image by Author 
 

Ready to make your GPT?

Then click “Create a GPT” in the “My GPTs” section, and voila—this unfolds a user-friendly editor, split into two parts: the GPT Builder and a live preview of your creation. 

This interactive setup allows you to chat directly with the GPT Builder, giving life to your AI vision in real time.

 

Image by Author 

 

3. Fine-Tuning Your AI Companion

 

Configuring your ChatGPT goes beyond just naming it. You’re provided with many options to personalize its behavior, purpose, and even appearance.

You have two main ways to do so:

1. Opt for the “Configure” option, where you can manually customize your GPT.
2. Engage in a direct conversation with the GPT Builder. This approach involves directly communicating your specific needs and requirements for your ChatGPT instance to the Builder, and letting the magic happen.

So… you should decide what to do next. To further understand both ways to proceed, I will explain both.

 

Process 1. Creating it from scratch

 

If we select the ‘Configure’ option, it leads to a page detailing the steps for creating your GPTs. The main sections to craft are: 

1. An avatar for this personalized GPT version. 
2. A Name and a Description to give your GPT an identity.
3. Instructions to outline its high-level behavior.
4. Conversation Starters to set the tone with initial prompts.
5. Knowledge to enhance with custom files.
6. Capabilities to add web browsing, DALL-E, or coding skills.
7. Actions to integrate external APIs or data. 

 

Image by Author 
 

So in this case, you can fill in all the information required step by step.

Of course, you can take advantage of the capabilities of ChatGPT to write you most of the previous information. 

And this leads us to the second process…

 

Process 2. Taking advantage of the GPT builder

 

Instead of writing manually all the required inputs in order to generate your personalized ChatGPT instance, you can directly chat with the GPT builder and let the magic happen. 

So in this case, imagine I want a tech advisor that can simplify complex data science concepts. I can describe this hypothetical GPT instance so that the GPT builder can generate a first version.

 

Image by Author 
 

And just like that… magic happens and the GPT builder makes it all for us!

 

Image by Author 

 

4. Beyond the basics

 

To elevate your GPT beyond a quick build and prepare it for market-readiness, focus on its uniqueness. This can be achieved by enriching it with specialized knowledge and actions, such as incorporating valuable resources or integrating APIs for access to real-time data.

 

Knowledge

 

GPTs present a knowledge feature that allows us to upload data files so they can “expand” their knowledge.

Add valuable files to your GPT, such as a CSV containing related information for their task to perform.

However, be aware that the contents of these files are visible during chats, so it’s important to be cautious of others who might copy your ideas.

 

Actions

 

When creating a GPT with the requirement of incorporating external data into user requests, it’s essential to integrate an API capable of supplying this data.

This integration involves defining specific endpoints, parameters, and instructions for how the model should utilize this information.

While the concept of actions may seem daunting at first, they are a significant tool that can enhance the uniqueness and profitability of your GPT. Given the complexity and importance of this feature, a dedicated guide will be developed shortly to offer more detailed insights.

 

5. Testing your new AI companion

 

As for prompt engineering, generating your own GPT instance is an iterative process. This means you have to test your brand-new model, check its behavior as expected, and try to enhance those errors you might find.

Experiment using the Preview option and, when you’re content with the results, proceed to Save or Update.

 

6. Bringing Your GPT to Life

 

Once you’ve molded your ChatGPT to your liking, it’s time to bring it to the digital world. 

With a simple save and publish action, your AI companion is ready to interact, learn, and evolve, offering a unique experience tailored just for you.

 

Image by Author 

 

The Dawn of Personalized AI

 

The brand-new launched GPTs by OpenAI are more than just a technological advancement; it’s a paradigm shift in how we perceive and interact with AI. 

This new tool not only enhances user experience but also heralds a new era of personalized digital assistants. 

Whether for business, education, or personal use, the ability to tailor AI to our specific needs is a monumental leap forward. 

So, why wait? 

Dive into the world of personalized ChatGPTs and experience the future of AI customization firsthand. 
 
 

Josep Ferrer is an analytics engineer from Barcelona. He graduated in physics engineering and is currently working in the data science field applied to human mobility. He is a part-time content creator focused on data science and technology. Josep writes on all things AI, covering the application of the ongoing explosion in the field.