How to Play-to-Earn with GPTs?
By OpenTaskAI profile image OpenTaskAI
7 min read

How to Play-to-Earn with GPTs?

You no longer need to be a tech expert to develop your own AI chatbot. OpenAI, the creators of ChatGPT, has revealed that users can now make their own custom versions of the widely popular AI chatbot. Ordinarily, creating a functional AI chatbot would require substantial knowledge in programming and

You no longer need to be a tech expert to develop your own AI chatbot.

OpenAI, the creators of ChatGPT, has revealed that users can now make their own custom versions of the widely popular AI chatbot.

Ordinarily, creating a functional AI chatbot would require substantial knowledge in programming and machine learning, as well as access to extensive training data sets.

1 What is GPTs?

At 02:00 AM Beijing time on November 7th, OpenAI hosted its first-ever DevDay event. Sam Altman presented for 45 minutes, unveiling several new products:

  • The new GPT-4 Turbo model, more powerful and affordable, supporting a 128K context window.
  • A new Assistants API, allowing developers to easily build AI assistant applications with goals and the ability to call models and tools.
  • New multimodal capabilities on the platform, including vision, image creation (DALL-E 3), and text-to-speech (TTS).

 Among these, GPTs are a tool that allows users to tailor their own AI assistants. Imagine being able to create a completely customized ChatGPT based on your needs and preferences. Whether you need an assistant to help sort through emails or a companion that provides creative inspiration, GPTs make it all possible.

This type of customized AI not only improves work efficiency but also brings more convenience to everyday life. Users don't need programming skills or a deep technical background. Through simple steps on the ChatGPT interface, they can create their own GPTs and use it in various scenarios.

The emergence of GPTs represents a significant advancement in AI technology, extending its application from the domain of experts to the daily lives of the general public. OpenAI has already launched several ready-to-use GPTs for everyone (as shown below), such as "The Negotiator", "Game Time", and more.

2 What are the functions of GPTs? Can everyone use the features of GPTs?

You might be wondering, what exactly can GPTs do?

Users can upload data to customize GPTs, creating AI assistants that better meet personal or professional needs. For example, you could build a robot focused on interview practice or upload extensive interview materials for specialized training. In addition to creating and using your own GPTs, you can also share them with others or use GPTs created by others.

Currently, the use of GPTs is limited to ChatGPT Plus users. Whether you want to create your own GPTs or use someone else's creation, you need to upgrade to ChatGPT Plus first.

Additionally, OpenAI plans to launch the GPT Store in the future, a platform similar to an app store that allows users to share their GPTs. There might even be a revenue-sharing system, creating new sources of income for creators and OpenAI.

3 How to create your own CTPs?

Step 1: Define Your GPTs Objective

First, consider what purpose your GPTs will serve. In our example, we will create an "English Resume Bot" focused on providing resume writing advice and rephrasing.

Step 2: Enter the GPTs Creation Interface

To use the GPTs feature, you need to be a ChatGPT Plus user. Here are the steps to access the GPTs creation interface:

Click on 'Explore' in the left column of ChatGPT.

Then select 'Create a GPT' to enter the GPTs creation interface. 

This interface is divided into two sides, left and right. The left side is mainly for bot building, with two tabs: 'Create' and 'Configure'. 

Under the 'Create' tab: Users begin a dialogue with the GPT Builder, which is the stage for defining the behavior and functions of the GPTs. Here, you would enter commands like "I want to create a resume bot," and the GPT Builder will set it up based on this information.

Switching to the 'Configure' tab: You will further refine your GPTs. Here, you can name your GPT, add a description, set specific commands and conversation starters, and upload relevant knowledge files, which will serve as the source of information for the GPT's responses. Additionally, you can set the capabilities of the GPTs, such as whether it can browse the web, use DALL·E to generate images, or parse code.

On the right side, the preview area, you can interact in real-time with the GPTs you are currently creating.

Step 3: Enter Requirements in the Dialogue Box

The process of creating GPTs is similar to having a conversation with ChatGPT. The GPT Builder will ask how you want your GPTs to function. For instance, when I enter the command "I want to create a resume bot" in the input box on the left, the GPT Builder then provides feedback based on the entered command.

After I input continue your talks, the GPT Builder begins to guide the bot's setup step by step, starting from suggesting a name and even designing a representative robot avatar.

Then, the GPT Builder will pose a series of questions to help further refine the bot. In the example from the article, it asks what aspects the "resume bot" should focus on. Through such interactions, you can continuously adjust and improve the design of your chatbot.

Moreover, during the process of building GPTs, you can use the dialogue interface on the right side to instantly test and view results. As I build the bot on the left side and input resume data on the right interface, the "resume bot" immediately responds, just like interacting with the original ChatGPT.

 

If you're not satisfied with the output from the GPT Builder during the construction process, you can make adjustments at any time by directing the GPT Builder on the left interface. After a series of inputs and adjustments, the GPT Builder will make improvements to better meet your needs. For example, as shown in the image below, I specified certain features a resume should have based on specific needs. The GPT Builder then makes adjustments, and after I test again on the right side, the results are more in line with my expectations.

Step 4: Custom Settings

In the process of creating personalized GPTs, the GPT Builder is not the only tool you have. You can also make more detailed customizations in the 'Configure' section (as shown below).

In the 'Configure' tab, you'll find several customizable fields, as listed below:

Name: Give the bot a unique name.

Description: Provide detailed information about the chatbot's purpose and features.

Instructions: Similar to the commands entered in ChatGPT. In fact, during the interaction with GPT Builder, it also adjusts the commands in 'Instructions' based on the requirements raised.

Conversation starters: Set preset text for when the chatbot starts a conversation.

Knowledge: Uploaded files become the knowledge base for the GPTs, and this content may be included in the conversations.

Capabilities: You can choose to enable three functions--Web Browsing, DALL·E Image Generation, and Code Interpreter.

Actions: This is an advanced feature. You can expand the GPT's functions to interact with external services or applications. Set up detailed API endpoints, parameters, and how the model should use this information to interact with external services or applications.

Step 5: Save and Share

The final step is to save your GPTs. Click the "Save" button in the top right corner of the screen, and you will see several options:

Only me: This means that only you can use this chatbot.

Anyone with a link: ChatGPT Plus users with a link can use this chatbot.

After selecting the appropriate option, click "Confirm" to complete the saving process. Now, you can start using the GPTs you've created.

On the "Explore" page, you can see all your GPTs, and if you need to make modifications or delete them, simply click the respective buttons.

To share it with others for their use, just click on the GPTs and choose "Copy Link" from the dropdown menu.

Step 6: Start Using Your GPT

As seen in the attached image, once back in the ChatGPT interface, the created "ExplainThis English Resume Bot" is ready to go. Simply input your resume experience in the chat box, and the bot will provide feedback based on the configured responses.

As shown in the example, after entering basic resume experience, the "ExplainThis English Resume Bot" can assist with translation, enhancement, and rephrasing of content. Once a conversation begins, the conversation history is saved. To initiate a new conversation, just click on the bot you want to chat with, similar to how you interact with the original ChatGPT.

This process demonstrates the convenience of using GPTs: the ability to create a personalized chatbot and share it with others without having to repeat the same commands. 

Conclusion:

In the future, the GPT Store will review GPTs published by developers. 

Once these curated GPT models are available for use, they can be searched and used, similar to downloading software from an app store. By listing their GPT models, developers not only gain exposure but also have the opportunity to earn income.

By OpenTaskAI profile image OpenTaskAI
Updated on
AI Tips