In this article, we will explore the Concepts and Use Cases of ChatGPT plugins and how it can transform the way businesses interact with their customers.
Table of Contents
ChatGPT Plugins
ChatGPT plugins are software extensions or add-ons that are designed to enhance the functionality and capabilities of ChatGPT, which is a language model developed by OpenAI. ChatGPT itself is a large language model trained using deep learning techniques, capable of generating human-like text responses based on input prompts.
You might also like
ChatGPT plugins can extend the capabilities of ChatGPT by providing additional features or functionalities that can be integrated into applications or platforms. These plugins can be developed by third-party developers or by OpenAI, and they typically leverage the capabilities of ChatGPT to generate rewritten or paraphrased text, provide suggestions, or enhance the performance of specific tasks, such as content creation, language translation, summarization, and more.
ChatGPT plugins can be used in a wide range of applications, including content creation, social media management, SEO optimization, chatbots, virtual assistants, writing assistance, and other use cases where generating text or paraphrasing text is required. These plugins aim to provide users with more flexibility, customization, and efficiency in using ChatGPT for their specific needs.
It’s important to note that ChatGPT plugins are separate entities from the core ChatGPT model and require installation and integration into specific platforms or applications to be utilized. Each plugin may have its own set of features, functionalities, and settings that can be customized according to the user’s requirements.
Examples of ChatGPT plugins
Until now, some plugins have been made. However, it’s clear that there will be many more ChatGPT plugins in the upcoming weeks and months.
In addition to third-party ChatGPT plugins, OpenAI suggests hosting two plugins internally to begin with: web browsing and a code interpreter. Finally, they’ve released the source code for a retrieval plugin.
Let’s take a closer look at each of these.
ChatGPT web browsing
A test model that understands when and how to use the internet.
OpenAI has previously done extensive research into integrating online surfing to LLMs, and since it’s such an important tool for connecting ChatGPT to the rest of the internet, it makes reasonable that they’re handling this one internally.
When using this plugin, you can see exactly what browsing steps the LLM is taking, including what it searches, clicks on, and reads, similar to the LangChain library.
As we can see, the plugin also includes links to and cites for the sources used in the answer.
They also explain that, for safety reasons, the plugin can only make GET queries, which means it can only get information and cannot execute “transactional” requests like filling out a form.
ChatGPT code interpreter
An experimental ChatGPT model that can manage uploads and downloads using Python
Previously, ChatGPT could understand and write code; now, it can interpret and run code in the browser.
For the time being, the models are only given a Python interpreter, but we expect them to add support for other coding languages in the future. It also allows you to upload files to the chat workspace and receive the results.
They highlight a few instances when the code interpreter comes in useful:
- Solving mathematical problem
- Data visualization and analysis
- Converting files between different formats
ChatGPT retrieval plugins
Natural language queries are used to give a versatile solution for semantic search and retrieval of personal or organizational materials.
In other words, the retrieval plugin allows you to discover relevant document snippets in files, notes, emails, and other similar sources. With this plugin, businesses may also make internal papers available to employees via ChatGPT.
The retrieval plugin, as predicted, computes the embeddings of each document chunk using the text-embedding-ada-002 embeddings model and saves them in a vector database.
Because it is open-source and self-hosted, developers can create and register their own retrieval plugins. They also support a variety of vector databases, including Pinecone, Weaviate, and others.
ChatGPT third-party plugins
An experimental model that understands when and how to employ plugins.
You may also utilize third-party plugins by reading the plugin descriptions and determining when and how to use them. A manifest file, for example, describes third-party plugins:
{
"schema_version": "v1",
"name_for_human": "TODO Manager",
"name_for_model": "todo_manager",
"description_for_human": "Manages your TODOs!",
"description_for_model": "An app for managing a user's TODOs",
"api": { "url": "/openapi.json" },
"auth": { "type": "none" },
"logo_url": "https://example.com/logo.png",
"legal_info_url": "http://example.com",
"contact_email": "[email protected]"
}How to create a ChatGPT plugin
- Create an API with endpoints for the LLM to use; these can be new APIs, existing APIs, or a wrapper around existing APIs designed specifically for ChatGPT and language models.
- Submit documentation for the OpenAPI specification as well as a manifest file that describes the API for both models and humans, as well as some other metadata.
After creating a ChatGPT plugin, users can select which plugin to enable, which can be accessed via the API documentation linked above.
This article is to help you learn about the ChatGPT plugin. We trust that it has been helpful to you. Please feel free to share your thoughts and feedback in the comment section below.
