I have read many these kind of books, and I have found none of them to be helpful. They are just mangling delicious-sounding sentences.
I have found that the most effective way to think is to write your own book, your own expedition of the matter at hand. When you write a sentence/paragraph, you notice how poor/ugly/erroneous your writing is, and then you rewrite it. I love being noticing how wrong I am, because at that point I have learned something. This way, you have iterated and learned the matter, and learnings are not just in your brain with you all the time, but you also have externalized it in writing, and the passing of time shows if it is timeless bulletproof understanding/thinking/learning/whatever.
I have physical copies of these books. They are really well quality material, well layouted, and the content is similar to what is found at fs.blog website. All the "mental models" for thinking they include are kind of p*rn for this kind of stuff, but it is nearly impossible to apply these mental models consciously in real life, especially ad-hoc real life situation. There is no time to think each different mental model for every situation/task/problem/matter. The most effective thinking is the real-life "quick math/no BS" style of thinking, and the real superpower is if you have made your own expedition to the matter, because then you can just recall the stuff you need from your own brain quicker and better than any LLM.
Thanks. This is brilliant and makes so much sense, but it's not something I would have searched for.
This also plays along very well with OP's question. I personally always thought that AI makes me think a lot more but maybe that's because I give it massive essays while other people use it to avoid writing essays.
Surprised no one has recommended it. It’s like an explanation of how we think, and how we should think. My only caution with this book is it represents “thinking” from a Westernised concept of rationality and it also represents western cognitive biases. A lot of the research done by Daniel Kahneman and Amos Tversky was only studied among westernised, educated, industrialised, rich, democratic populations which only actually makes up about 12% of the world population. It’s certainly not going to apply to every culture across humanity, but - it’s super helpful.
Having awareness of cognitive biases helps you recognise them in yourself, but also leverage them in other people and that’s power.
This is going to sound like nonsense, but the best books that really teach how to think, how to handle complex thoughts and logic, and how to reason multiple steps in advance, and expand that into all the possibilities and still resolve all of that mess into a logical course of action... this process is documented, in story narrative form over and over again in literature. This is how a novel gets nominated for a Nobel Literature award: they depict a situation, typically a dire tragedy, and the individuals facing that situation and how they logically navigate to success. The characters typically begin without the critical thought capacity to succeed in their situation, and the novel is the story of their piece-wise transformation into the critically aware individual operating with reasoning multiple steps ahead of themselves. These novels are not math, are not anything technical, yet they teach their readers a critical skill that math and anything technical requires to master: critical thinking in an unknown space where one needs to make predictions with unknown information, and how to manage that. It's literally a literature "formula".
You don't need books on how to think. You already know how to think. What you need is raw material (knowledge) to think with. This means reading books, lots of books, in a variety of fields. It also means shutting off your smartphone and reading.
Have you considered the opposite... writing a book? I think writing a book is going to help you learn to think better than reading ever would. Writing is basically refined thought.
- Think Again: The Power of Knowing What You Don't Know by Adam Grant
- Thinking, Fast and Slow by Daniel Kahneman
- The Daily Stoic: 366 Meditations on Wisdom, Perseverance, and the Art of Living Hardcover by Ryan Holiday
A succinct book which is a good primer for thinking through things is "Rulebook for Arguments" by Anthony Weston. Quick to get through and understandable.
Not directly answering your question; but this is the advice I wish someone had told me 30 years ago.
1. Logical arguments. No one wants to be interrogated by a courtroom lawyer. Sometimes it is necessary for very high-stakes scenarios where everything must be exactly right. But nobody likes it. This is the same reason why "waterfall" gets a bad name. If the specifications were correct, waterfall won't get such a bad reputation. The reason why specifications are never correct is that nobody actually sits through and takes the time to logically define in excruciating detail what they want. The only part of logical argumentation I have found useful is agreeing upon common terminology; in any non-trivial document, you should always have a terminology section where you specify this is what you mean by X. I have witnessed massive pushback against the common definitions like what a database is or what a system owner is.
2. People problems. I would argue 20% of tech problems are people problems which are really communication problems. e.g. business people don't know how things work and what is involved in performing Y. In their minds, what they are asking for is very simple. What they don't realize is they are asking you to integrate three different pieces of software created by different vendors that are completely incompatible with each other. When the developer tries to explain the technical complexity, they think you are lying.
3. Political problems. I would argue 80% of tech problems I have encountered at work is pure power politics. This is war in the non-violent sense. People have lied to me repeatedly and actively tried to sabotage my reputation and my projects. Not to mention to accuse me of wrongdoing for things they were actually responsible for. In this land of where everything is a lie, people go with their "gut". This is where psychology is useful. It turns out their "gut" is nothing more than past biases and cognitive short-cuts. The ability to read the room is the most critical ability here.
I have read many these kind of books, and I have found none of them to be helpful. They are just mangling delicious-sounding sentences.
I have found that the most effective way to think is to write your own book, your own expedition of the matter at hand. When you write a sentence/paragraph, you notice how poor/ugly/erroneous your writing is, and then you rewrite it. I love being noticing how wrong I am, because at that point I have learned something. This way, you have iterated and learned the matter, and learnings are not just in your brain with you all the time, but you also have externalized it in writing, and the passing of time shows if it is timeless bulletproof understanding/thinking/learning/whatever.
Would you mind listing the ones you didn't find useful? I believe 90% of everything is crap, but I want to get to the good stuff faster.
Well, maybe the best example is this: https://fs.blog/tgmm/
I have physical copies of these books. They are really well quality material, well layouted, and the content is similar to what is found at fs.blog website. All the "mental models" for thinking they include are kind of p*rn for this kind of stuff, but it is nearly impossible to apply these mental models consciously in real life, especially ad-hoc real life situation. There is no time to think each different mental model for every situation/task/problem/matter. The most effective thinking is the real-life "quick math/no BS" style of thinking, and the real superpower is if you have made your own expedition to the matter, because then you can just recall the stuff you need from your own brain quicker and better than any LLM.
Thinking is writing and writing is thinking, and thus also we've been pretty intense about the topic for a long time:
https://hn.algolia.com/?q=how+to+write
Just two cents.
Thanks. This is brilliant and makes so much sense, but it's not something I would have searched for.
This also plays along very well with OP's question. I personally always thought that AI makes me think a lot more but maybe that's because I give it massive essays while other people use it to avoid writing essays.
“Thinking, Fast and slow” by Daniel Kahneman
Surprised no one has recommended it. It’s like an explanation of how we think, and how we should think. My only caution with this book is it represents “thinking” from a Westernised concept of rationality and it also represents western cognitive biases. A lot of the research done by Daniel Kahneman and Amos Tversky was only studied among westernised, educated, industrialised, rich, democratic populations which only actually makes up about 12% of the world population. It’s certainly not going to apply to every culture across humanity, but - it’s super helpful.
Having awareness of cognitive biases helps you recognise them in yourself, but also leverage them in other people and that’s power.
This is going to sound like nonsense, but the best books that really teach how to think, how to handle complex thoughts and logic, and how to reason multiple steps in advance, and expand that into all the possibilities and still resolve all of that mess into a logical course of action... this process is documented, in story narrative form over and over again in literature. This is how a novel gets nominated for a Nobel Literature award: they depict a situation, typically a dire tragedy, and the individuals facing that situation and how they logically navigate to success. The characters typically begin without the critical thought capacity to succeed in their situation, and the novel is the story of their piece-wise transformation into the critically aware individual operating with reasoning multiple steps ahead of themselves. These novels are not math, are not anything technical, yet they teach their readers a critical skill that math and anything technical requires to master: critical thinking in an unknown space where one needs to make predictions with unknown information, and how to manage that. It's literally a literature "formula".
You don't need books on how to think. You already know how to think. What you need is raw material (knowledge) to think with. This means reading books, lots of books, in a variety of fields. It also means shutting off your smartphone and reading.
Have you considered the opposite... writing a book? I think writing a book is going to help you learn to think better than reading ever would. Writing is basically refined thought.
It's math-centric, but How to Solve It (Polya) is a good book on how to think critically and approach problems methodically.
from previous thread, a lot of good suggestions.
Learning to Learn the coursera course is always recommended
A succinct book which is a good primer for thinking through things is "Rulebook for Arguments" by Anthony Weston. Quick to get through and understandable.
I really like The Philosopher's Toolkit: A Compendium of Philosophical Concepts and Methods by Baggini & Fosl.
Not directly answering your question; but this is the advice I wish someone had told me 30 years ago.
1. Logical arguments. No one wants to be interrogated by a courtroom lawyer. Sometimes it is necessary for very high-stakes scenarios where everything must be exactly right. But nobody likes it. This is the same reason why "waterfall" gets a bad name. If the specifications were correct, waterfall won't get such a bad reputation. The reason why specifications are never correct is that nobody actually sits through and takes the time to logically define in excruciating detail what they want. The only part of logical argumentation I have found useful is agreeing upon common terminology; in any non-trivial document, you should always have a terminology section where you specify this is what you mean by X. I have witnessed massive pushback against the common definitions like what a database is or what a system owner is.
2. People problems. I would argue 20% of tech problems are people problems which are really communication problems. e.g. business people don't know how things work and what is involved in performing Y. In their minds, what they are asking for is very simple. What they don't realize is they are asking you to integrate three different pieces of software created by different vendors that are completely incompatible with each other. When the developer tries to explain the technical complexity, they think you are lying.
3. Political problems. I would argue 80% of tech problems I have encountered at work is pure power politics. This is war in the non-violent sense. People have lied to me repeatedly and actively tried to sabotage my reputation and my projects. Not to mention to accuse me of wrongdoing for things they were actually responsible for. In this land of where everything is a lie, people go with their "gut". This is where psychology is useful. It turns out their "gut" is nothing more than past biases and cognitive short-cuts. The ability to read the room is the most critical ability here.
Philosophical Investigations
LLM’s are mechanisms simulating playing language games.