These 10 books will take you to understand the underlying logic of ChatGPT

Article source: People's Posts and Telecommunications Press

Since its release on November 30, 2022, ChatGPT has truly become popular all over the world: it has 10 million users in less than 40 days, while Instagram took 355 days; recently its daily activity has reached 10 million, This means that its users have exceeded 20 million.

▲ ChatGPT daily activity data statistics, source: ARK Venture Capital Company

The power of ChatGPT must have been seen or experienced by many friends: it can not only talk to us fluently, but also write poems, articles, and even code.

Because of this, many foreign universities and education systems have banned students from using ChatGPT because too many students use it to write homework and cheat on exams.

Many readers can't wait to learn about ChatGPT, but they have no way to start.

Today, Xiaoyou compiled a book list for everyone, so that everyone can understand the underlying logic of ChatGPT.

01 Popular Science

What exactly is ChatGPT? Will artificial intelligence conquer humans? Moss started layout? What is the development trend of artificial intelligence? The underlying logic of ChatGPT explosion?

This is ChatGPT

Author: [US] Stephen Wolfram
Translator: WOLFRAM Media Sinicization Team

Recommended language:

The author of this book, Wolfram , is an important innovator in the field of science and technology today, and the world's top AI scholar ; and this book can be said to be the best explanation of the principle of ChatGPT, and it is also strongly recommended , CEO of OpenAI and father of ChatGPT !

After reading this book, you will not only understand the success of ChatGPT, but also how and why it did it . This book will briefly introduce the internal mechanism of ChatGPT, and then explore why it can generate text that we think is meaningful.

As soon as the book was released, it was recognized by readers. In just one month, it was reprinted 4 times! If you also want to understand ChatGPT, use a book to understand the essence of ChatGPT, you can't go wrong with it!

A Brief History of Artificial Intelligence 2nd Edition

Author: [US] Nick

Recommended language:

Detailed explanation of the origin of artificial intelligence, neural network, deep learning, natural language processing, and genetic algorithm, won the China Excellent Publication Book Award, Wu Wenjun Artificial Intelligence Science and Technology Progress Award, was shortlisted for CCTV's "2017 China Good Book", and won the "Wenjin Book Award" Recommend books.

Hackers and Painters

Author: [US] Paul Graham

Recommended language:

Paul Graham, the godfather of entrepreneurship in Silicon Valley, believes that Sam Altman, the father of ChatGPT, is a very courageous leader and pioneer, and he also succeeded the president of YC. The influence of Paul and YC on him is also a factor in his ultimate success. The core content of this book is that an effective way of thinking is your most powerful weapon. Through the perspective of hackers and painters, you can talk to smart minds and understand why smart people think differently from us.

02 Introduction to Technology

Keep up with artificial intelligence and not be eliminated by society! Learn about artificial intelligence and play ChatGPT!

Graphical artificial intelligence

Author: [Japan] Tomoshi Tada

Recommended language:

This book collects the basic knowledge of artificial intelligence development in a graphical way, and the content involves popular technologies such as machine learning, deep learning, reinforcement learning, image and voice pattern recognition, natural language processing, and distributed computing.

The whole book is accompanied by pictures and texts, and it is an introductory textbook for artificial intelligence that takes into account both theory and technology. It is designed to help readers establish an overall impression of artificial intelligence technology and lay a foundation for in-depth exploration of this field in the future.

Introduction to Deep Learning : Theory and Implementation Based on Python

Author: Saito Yasuhiro

Recommended language:

The classic best-selling book on introductory deep learning in Japan, the original edition has been printed for less than 2 years and has reached 100,000 copies. For non-AI technical personnel, this book will greatly reduce the threshold for getting started with deep learning; for college students and graduate students, this book is a good textbook for learning deep learning; Readers of deep learning models can also gain new experience from this book.

Hands-on machine learning with Python

Author: [Japan] Makoto Ito

Recommended language:

This book is an introductory book for machine learning beginners. Starting from the construction of the learning environment, it introduces the knowledge of Python and mathematics required for learning machine learning with pictures and texts, and on this basis, combines mathematical formulas, sample programs, illustrations, etc. It generally introduces regression and classification in supervised learning, algorithms and applications of neural networks and deep learning, handwritten digit recognition, and algorithms for unsupervised learning. This book not only has graphics, codes, but also detailed mathematical derivation process, which greatly lowers the learning threshold of machine learning.

Getting Started with Natural Language Processing

Author: He Han

Recommended language:

The learning path of this book is clear, with both pictures and texts, algorithms, formulas, and codes are mutually confirmed, and both Java and Python are implemented. The author He Han ( @hankcs) has both first-line practical experience and hard-core academic background. The author of the natural language processing open source project HanLP, a doctoral student in computer science at Emory University , and his research direction is syntax analysis, semantic analysis and question answering system

After listing, it has received unanimous praise from experts in the industry: Zhou Ming, Li Hang , Liu Qun, Wang Bin, Yang Pan in the industry, Zong Chengqing, Liu Zhiyuan, Zhang Huaping in the academic circle.

The book comes with a large-scale mind map, source code download, GitHub Q&A, and teaching handout PPT for teachers.

03 Advanced Technology

How do technical people catch up with the wave of ChatGPT? How to trick ChatGPT? The Advancement of Artificial Intelligence!

Deep Learning with Python (2nd Edition)

Author: [US] François Chollet

Recommended language:

This book is written by Francois Cholet, the father of the popular deep learning framework Keras. It helps you build a deep learning knowledge system through intuitive explanations and rich examples. The author avoids mathematical notation and instead uses Python code to explain the core ideas of deep learning.

The book has a total of 14 chapters, which not only cover the basic principles of deep learning, but also reflect the important progress made in this rapidly developing field in recent years, including the principles and examples of the Transformer architecture.

After reading this book, you will be able to use Keras to solve many real-world problems ranging from computer vision to natural language processing, including image classification, image segmentation, time series prediction, text classification, machine translation, text generation, and more.

Machine learning formula derivation and code implementation

Author: Lu Wei

Recommended language:

As an applied discipline, machine learning is rooted in mathematical theory and implemented in code. This means that only by mastering formula derivation and code writing can we have a deeper understanding of the internal logic and operating mechanism of machine learning algorithms.

Based on the classification and sorting of all machine learning algorithms, this book provides detailed formula derivation and analysis of 26 classic algorithms in four categories: supervised learning single model, supervised learning integrated model, unsupervised learning model, and probability model. Code implementation, designed to help machine learning learners and researchers fully grasp algorithm details, implementation methods, and internal logic.

deep reinforcement learning

Author: Wang Shusen, Li Yujun, Zhang Zhihua

Recommended language:

This book is based on Wang Shusen's "Deep Reinforcement Learning" series of public video courses, which are highly praised by readers, and specifically solves the problem of "difficulty getting started with deep reinforcement learning".

The unique features of this book are: first, the knowledge is streamlined, and all unnecessary concepts and formulas are eliminated, making it easy to learn; Keep up with the latest technology. This book systematically explains the principles and implementation of deep reinforcement learning, but does not avoid mathematical formulas and various models. More than 100 original exquisite illustrations are displayed in full-color printing. Concise and clear language + vivid illustrations will help you remove any possible learning obstacles!

The content of this book is divided into five parts: basic knowledge, value learning, policy learning, multi-agent reinforcement learning, application and prospect, involving DQN, A3C, TRPO, DDPG, AlphaGo, etc.