Book Image

Modern Generative AI with ChatGPT and OpenAI Models

By : Valentina Alto
4.9 (8)
Book Image

Modern Generative AI with ChatGPT and OpenAI Models

4.9 (8)
By: Valentina Alto

Overview of this book

Generative AI models and AI language models are becoming increasingly popular due to their unparalleled capabilities. This book will provide you with insights into the inner workings of the LLMs and guide you through creating your own language models. You’ll start with an introduction to the field of generative AI, helping you understand how these models are trained to generate new data. Next, you’ll explore use cases where ChatGPT can boost productivity and enhance creativity. You’ll learn how to get the best from your ChatGPT interactions by improving your prompt design and leveraging zero, one, and few-shots learning capabilities. The use cases are divided into clusters of marketers, researchers, and developers, which will help you apply what you learn in this book to your own challenges faster. You’ll also discover enterprise-level scenarios that leverage OpenAI models’ APIs available on Azure infrastructure; both generative models like GPT-3 and embedding models like Ada. For each scenario, you’ll find an end-to-end implementation with Python, using Streamlit as the frontend and the LangChain SDK to facilitate models' integration into your applications. By the end of this book, you’ll be well equipped to use the generative AI field and start using ChatGPT and OpenAI models’ APIs in your own projects.
Table of Contents (17 chapters)
1
Part 1: Fundamentals of Generative AI and GPT Models
4
Part 2: ChatGPT in Action
11
Part 3: OpenAI for Enterprises

Avoiding the risk of hidden bias and taking into account ethical considerations in ChatGPT

ChatGPT has been provided with the Moderator API so that it cannot engage in conversations that might be unsafe. The Moderator API is a classification model performed by a GPT model based on the following classes: violence, self-harm, hate, harassment, and sex. For this, OpenAI uses anonymized data and synthetic data (in zero-shot form) to create synthetic data.

The Moderation API is based on a more sophisticated version of the content filter model available among OpenAI APIs. We discussed this model in Chapter 1, where we saw how it is very conservative toward false positives rather than false negatives.

However, there is something we can refer to as hidden bias, which derives directly from the knowledge base the model has been trained on. For example, concerning the main chunk of training data of GPT-3, known as the Common Crawl, experts believe that it was written mainly by white males...