How to Use Visualization to Aid Thinking - Obsidian Practice (Book Notes on "Genius and Algorithms") - Minority Report#
Omnivore#
How to Use Visualization to Aid Thinking - Obsidian Practice (Book Notes on "Genius and Algorithms")
This article shares a practical case of using the Excalidraw plugin in Obsidian to create book notes.
The book being shared is "Genius and Algorithms," originally titled "The Creativity Code."
The author of this book is Marcus du Sautoy, a mathematician who has created popular mathematical concepts and is skilled at expressing complex mathematics and numerical concepts in a vivid and easily understandable language. In this book, he uses mathematical thinking to help us understand algorithms and explores the innovative applications of artificial intelligence algorithms in music, writing, art, science, and mathematics, telling us how AI challenges human "creativity" in all aspects.
This book was published in 2019, and now in 2023, from the emergence of ChatGPT to the recent OpenAI conference, we hear about new AI products every day, experiencing unprecedented challenges and opportunities. In this rapidly developing era of AI, everyone should read this book.
The video is based on the content of this book and is divided into two parts to introduce to everyone:
- In the first part, I will share the book notes of "Genius and Algorithms" and provide a detailed introduction to the content of this book.
- In the second part, I will share the entire process of creating book notes using Obsidian, providing some practical references.
Book Notes on "Genius and Algorithms"
Book Notes on "Genius and Algorithms"#
Mathematics is the basic logic of computer programming, and algorithms, as the core of mathematics, have existed for thousands of years. Marcus du Sautoy's background as a mathematician allows him to look at problems from the perspective of algorithms. He believes that technology is reconstructing the world order.
Chapters 1-3: Can Machines Be Creative?
The first three chapters of the book discuss the question "Can machines be creative?" The impulse to create is one of the key elements that distinguishes humans from other animals. Margaret Boden divides human creativity into three types: exploratory creativity, combinatorial creativity, and transformational creativity.
This book discusses the "limits" of the new generation of artificial intelligence (AI): Can it possess creativity similar to or even surpassing that of humans? Can AI learn to create and help us improve our own creativity?
The first programmer, Ada Lovelace, firmly believed that any creative behavior depends on the programmer (human), not the machine. However, the new generation of programmers does not think so. They believe that "code" can also perform creative work. In verifying the question "Do machines really have creativity?" the author did not use the "Turing test" but proposed the Lovelace test, which sets the standard that an algorithm needs to create an artwork that human programmers cannot explain the workings of, and the entire process is reproducible. Ada called this an "insurmountable challenge."
In the process of creating AI, Go was once considered a challenge to computer creativity. Scholars believe that the way of thinking in Chinese Go better reflects the creativity and intuition of mathematicians. Go can form intricate and complex reasoning under simple rules. The AlphaGo created by Demis Hassabis' team defeated Lee Sedol with a score of 5:1, completing this challenge and shocking the world. Hassabis designed a clever model for the algorithm, which is to write a universal "metaprogram."
A universal "metaprogram" is like the brain of a newborn baby, which does not have pre-set methods to deal with survival challenges, but they will strengthen themselves through continuous learning and make appropriate adjustments based on changes in the environment. From a certain perspective, deep learning algorithms extract feature information that humans cannot describe and express in language.
Machines can learn, provided that humans teach them the right things to learn, and what they learn is a skill that humans do not possess: analyzing massive amounts of data and discovering valuable information from it.
Chapters 4-6: The Evolution of Algorithms
Chapters 4-6 of the book discuss the evolution of algorithms. Algorithms use the patterns we use to solve problems to guide us in finding solutions to problems. As our lives depend more and more on algorithms, it becomes increasingly important to have a deep understanding of the principles and processes of algorithms.
One important piece of information here is that "data" has triggered the AI revolution.
The internet can now generate 1EB (10 to the power of 18) of data every day. The amount of data generated by humans in two days can rival the total amount of data generated from the "dawn of civilization" to 2003. Massive amounts of data are the most important catalyst for machine learning entering a new era. The current business war is a war to seize user data, and they obtain more user data through recommendation algorithms like "Guess what you like."
There is also a type of data generated by machines through self-learning. The training data of AlphaGo comes from all the chess game data left by humans for thousands of years, and all subsequent optimizations are based on the initial "human data." DeepMind has developed AlphaZero, which no longer learns from human chess books or moves, but relies entirely on self-play to quickly improve its chess skills, thus breaking free from the limitations and patterns of human understanding of Go. AlphaZero completed its self-training in just 3 days, with a total of 4.9 million self-played games. What humans took 3000 years to achieve, it accomplished in just 3 days. When playing against the version of AlphaGo that defeated Lee Sedol, AlphaZero achieved an overwhelming record of 100:0.
Chapters 7-10: The Telescope of Mathematicians
Chapters 7-10 of the book discuss the view of machines, which the author calls the "telescope of mathematicians."
Mathematics is the science of discovering and explaining patterns. Mathematicians are essentially explorers and discoverers of patterns. The ability to discover patterns gives humans an advantage in negotiating with the natural world.
The author quotes Nietzsche's words that our writing tools are involved in the process of forming our thoughts. Some mathematicians believe, "We are in a period of transition between the old and the new: Although the development of mathematics is limited by the human brain, with the help of computers, our exploration of mathematics has far exceeded the scope of human thinking."
This section mentions a topic that I have always been interested in: "fractals."
The author believes that fractals are the code of nature. Nature uses fractal algorithms to create ferns, clouds, waves, mountains, etc. Scientists have discovered this "algorithm" and used computer-generated fractal images to simulate the natural world. This algorithm is now widely used in Hollywood movies.
Chapters 11-16: The Artistic Journey of Algorithms
Chapters 11-16 of the book discuss the creative use of algorithms in music, writing, painting, and other fields.
Style is an algorithm. The artistic style and symbols created by artists are the "human code" running in their brains. Humans have evolved to have a high sensitivity to the abstract structures that make up the natural world, and the behavior of artists depends to some extent on their own "algorithms" in response to the surrounding world.
Computers are powerful tools for extending human intelligence. When AI can learn our "human code" and have access to all "human data" in its "memory," it will sound the alarm for us to return to the essence of things: Why do we create? And how do we create next?
I hope that after reading this book, you can find some answers or gain some inspiration.
My personal knowledge management process is divided into [🌱Collect-🌳Organize-🌻Create-🌖Share].
Collect#
When collecting information from a book, I mainly focus on three aspects: the author, the outline, and the highlights of the book.
The author's background and writing time are what I am most interested in. I will look up the author's background, which helps me better understand the author's viewpoints.
The book outline is the logical structure of the story the author tells.
Finally, the highlights of the book are the viewpoints that inspire my thinking.
In the execution process, I have both the physical and electronic versions of this book. I read the physical version and underline interesting parts. I also insert a piece of paper with comments on the page at that time, and I use simple graphics on this paper to understand the information in the book.
The technique here is to record only one viewpoint on each sticky note, and I also record the original text information for easy reference in the future. This technique is derived from Luhmann's card box note-taking method. This process is mainly based on my own inspiration, not on quickly finishing a book. This technique is my principle of collecting information: "Collect quickly without interrupting thoughts."
Organize#
First, create a book note and record all the notes that need to be collected in one place. When reading an e-book, there are synchronization techniques or AI tools that can help with collection. Personally, I use the photo-taking function of WeChat and then extract the text from the photo. Here is a small tip: I will create a temporary template for a book that I find valuable, which helps maintain the order of my notes during organization.
The second step is to place this book note in other notes in my knowledge base. For example, if Demis Hassabis is mentioned, I will copy and paste it into the note on Demis Hassabis. If "human creativity" is mentioned, I will copy and paste it into the note on creative methods.
For this book, I chose to organize it after reading it in its entirety, as there are quite a few things to organize, but in the process of organizing, I read the key content again. When reading other books, you can organize them every few chapters according to your needs.
Create#
The third stage is also the most important stage in reading a book. Based on my own understanding, I will reorganize the content of the book. To think and understand better, I use the whiteboard function, which is the Excalidraw plugin. Depending on your situation, you can also choose to write articles or record videos for sharing.
Based on my understanding of this book, I will outline a new storyline and use visual graphics to depict abstract concepts, recording the images in my mind and assisting the abstract text to make it easier for future me to understand. Presenting complex stories or inspiring deep thinking through a combination of graphics and text is what I call visual note-taking. I use free resources, and now it is better achieved with the help of AI tools.
Storyline
Share#
After completing the creation of the notes, if you are a book note blogger or need to introduce this book to others, you can send the article or video to major platforms.
Create and Recreate#
Finishing a book is never the end.
In other thematic notes, there are also some viewpoints from this book. Based on the viewpoints I understand from this book, I will conceive new content on other channels, combining it with information collected from other sources to create a new idea. For example, I have another note called [🌠 Creativity - Humans and AI.excalidraw], where I place all the content related to this topic.
Deep Thinking - Visual Note-Taking#
My daily thinking, which is the "creation" stage of my workflow, is mainly done using a whiteboard. I try to use simple lines to outline the images in my mind, avoiding descriptions with only abstract text, and using a combination of graphics and text to present complex stories or inspire deep thinking.
All sudden inspirations are the result of long accumulation - Henri Poincare
Interested in studying PKM, interested in studying TFT, interested in thinking about how to stimulate creativity nebulapkm.com