Reviews of books and films from my week with Jacob:
Films watched:
The Big Short
Review: Really fun. I liked certain elements of how it displays bad nash equilibria in finance (I love the scene with the woman from the ratings agency—it turns out she’s just making the best of her incentives too!).
Grade: B
Spirited Away
Review: Wow. A simple story, yet entirely lacking in cliche, and so seemingly original. No cliched characters, no cliched plot twists, no cliched humour, all entirely sincere and meaningful. Didn’t really notice that it was animated (while fantastical, it never really breaks the illusion of reality for me). The few parts that made me laugh, made me laugh harder than I have in ages.
There’s a small visual scene, unacknowledged by the ongoing dialogue, between the mouse-baby and the dust-sprites which is the funniest thing I’ve seen in ages, and I had to rewind for Jacob to notice it.
I liked how by the end, the team of characters are all a different order of magnitude in size.
A delightful, well-told story.
Grade: A+
Stranger Than Fiction
Review: This is now my go-to film of someone trying something original and just failing. Filled with new ideas, but none executed well, and overall just a flop, and it really phones-it-in for the last 20 minutes. It does make me notice the distinct lack of originality in most other films that I’ve seen though—most don’t even try to be original like this does. B+ for effort, but D for output.
Grade: D
I Love You, Daddy
Review: A great study of fatherhood, coming of age, and honesty. This was my second watch, and I found many new things that I didn’t find the first time -about what it means to grow up and have responsibility. One moment I absolutely loved is when the Charlie Day character (who was in my opinion representing the id), was brutally honest and totally rewarded for it. I might send this one to my mum, I think she’ll get a lot out of it.
Grade: A
My Dinner With Andre
Review: Very thought-provoking. Jacob and I discussed it for a while afterward. I hope to watch it again some day. I spent 25% of the movie thinking about my own response to what was being discussed, 25% imagining how I would create my version of this film (what the content of the conversation would be), and 50% actually paying close attention to the film.
Overall I felt that both characters were good representatives of their positions, and I liked how much they stuck to their observations over their theories (they talked about what they’d seen and experienced more than they made leaky abstractions and focused on those). The main variable that was not discussed, was technology. It is the agricultural and industrial revolutions that lead to humans feeling so out-of-sorts in the present day world, not any simple fact of how we socialise today, that can simply be fixed / gotten out of. Nonetheless, I expect that the algorithm that Andre is running will help him gain useful insights about how to behave in the modern world. But you do have to actually interface with it and be part of it to have a real cup of coffee waiting for you in the morning, or to lift millions out of poverty.
The last line of Roger Ebert’s review of this was great. Something like: “They’re both trying to get the other to wake up and smell the coffee. Only in Willy’s case, it’s real coffee.”
Grade: B+
Books read (only parts of):
Computability and Logic (3rd edition)
I always forget basic definitions of languages and models, so a bunch of time was spent doing that. Jacob and I read half of the chapter on the non-standard numbers, to see how the constructions worked, and I just have the basics down more clearly now. Eliezer’s writings about these numbers connects more strongly to my other learning about first order logic now.
Book is super readable given the subject matter, easy to reference the concepts back to other parts of the book, and all round excellent (though it was the hardest slog on this list). Look forward to reading some more.
Modern Principles: Microeconomics (by Cowen and Tabarrok)
I’ve never read much about supply and demand curves, so it was great to go over them in detail, and how the price equilibrium is reached. We resolved many confusions, that I might end up writing in a LW post. I especially liked learning how the equilibrium price maximises social good, but is not the maximum for either the supplier or the buyer.
It was very wordy and I’d like to read a textbook that had the goal of this level of intuitiveness, but aimed at readers with assumed strong math background. I don’t need paragraphs explaining how to read a 2-D graph each time one comes up.
Jacob made a good point about how the book failed to distinguish hypothesis versus empirical evidence, when presenting standard microeconomic theory. Just because you have the theory down doesn’t mean you should believe it corresponds to reality, but the book didn’t seem to notice the difference.
Overall pretty good. I don’t expect to read most chapters in this book, but we also looked through asymmetric information (some of which later tied into our watching of The Big Short), and there were a few others that looked exciting.
Thinking Physics
I am in love with this book. I remember picking it up when I was about 17 and not being able to handle it at all and just flicking through to the answers—but this time, especially with Jacob, we were both able to notice when we felt we really understood something and wanted to check the answer to confirm, versus when we’d said ‘reasonable’ things but which didn’t really bottom out in our experiences of the world.
“Well, if you draw the force vectors like this, there should be a normal force of this strength, which splits up into these two basis vectors and so the ball should roll down at this speed.” “Why do you get to assume a force along the normal?” “I don’t know.” “Why do you get to break it up into two vectors who sum to the initial vector?” “I don’t know.” “Then I think we haven’t answered the question yet. Let’s think some more about our experience of balls rolling down hills.”
One of the best things about doing it with Jacob was that I often had cached answers to problems (both from studying mechanics in high school and having read the book 4 years ago), but instead on reading a problem I would give Jacob time to get confused about it, perhaps by supplying useful questions. Then eventually I’d propose my “Well isn’t it obviously X” answer, and Jacob would be able to point out the parts I hadn’t justified from first principles, helping me notice them. There’s a problem in discussing difficult ideas where if people have been taught the passwords, and especially if the passwords have a certain amount of structure that feels like understanding, that it’s hard to notice the gaps. Jacob helped me notice those, and then I could later come up with real answers, that were correct for the right reasons.
The least good thing about this book is the answers to the problems. Often Jacob and I would come up with an answer, then scrap it and build up a first-principles model that predicted it based in our experiences that we were very confident in, and then also deconstruct the initial false intuition some. Then we’d check the answer, and we were right, but the answer didn’t really address the intuitions in either direction, just gave a (correct) argument for the (correct) solution.
I think it might be really valuable to fully deconstruct the intuition behind why people expect a heavier object to fall faster. I’ve made some progress, but it feels like this is a neglected problem of learning a new field—explaining not only what intuitions you should have, but understanding why you assumed something different.
But the value of the book isn’t the answers—it’s the problems. I’ve never experienced such a coherent set of problems, where you can solve each from first principles (and building off what you’ve learned from the previous problems). With most good books, the more you put in the more you get out, but never have I seen a book where you can get this much out of it by putting so much in (most books normally hit a plateau earlier than this one).
Anyway, we got maybe 1/10th through the book. I can’t wait to work through this more the next time I see Jacob.
It’s already affected our discussions of other topics, how well we notice what we do and don’t understand, and what sorts of explanations we look for.
I’m also tempted, for other things I study, to spend less time writing up the insights and instead spend that time coming up with a problem set that you can solve from first principles.
This book made me think that the natural state of learning isn’t ‘reading’ but ‘play’. Playing with ideas, equations, problems, rather than reading and checking understanding.
Jacob and I now have a ritual of continuing the tradition of trying to understand the world, by going to places in Oxford where great thinkers have learned about the universe, and then solving a problem in this book. We visited a square in Magdalen College where Schroedinger worked on his great works, and solved some problems there.
Reviews of books and films from my week with Jacob:
Films watched:
The Big Short
Review: Really fun. I liked certain elements of how it displays bad nash equilibria in finance (I love the scene with the woman from the ratings agency—it turns out she’s just making the best of her incentives too!).
Grade: B
Spirited Away
Review: Wow. A simple story, yet entirely lacking in cliche, and so seemingly original. No cliched characters, no cliched plot twists, no cliched humour, all entirely sincere and meaningful. Didn’t really notice that it was animated (while fantastical, it never really breaks the illusion of reality for me). The few parts that made me laugh, made me laugh harder than I have in ages.
There’s a small visual scene, unacknowledged by the ongoing dialogue, between the mouse-baby and the dust-sprites which is the funniest thing I’ve seen in ages, and I had to rewind for Jacob to notice it.
I liked how by the end, the team of characters are all a different order of magnitude in size.
A delightful, well-told story.
Grade: A+
Stranger Than Fiction
Review: This is now my go-to film of someone trying something original and just failing. Filled with new ideas, but none executed well, and overall just a flop, and it really phones-it-in for the last 20 minutes. It does make me notice the distinct lack of originality in most other films that I’ve seen though—most don’t even try to be original like this does. B+ for effort, but D for output.
Grade: D
I Love You, Daddy
Review: A great study of fatherhood, coming of age, and honesty. This was my second watch, and I found many new things that I didn’t find the first time -about what it means to grow up and have responsibility. One moment I absolutely loved is when the Charlie Day character (who was in my opinion representing the id), was brutally honest and totally rewarded for it. I might send this one to my mum, I think she’ll get a lot out of it.
Grade: A
My Dinner With Andre
Review: Very thought-provoking. Jacob and I discussed it for a while afterward. I hope to watch it again some day. I spent 25% of the movie thinking about my own response to what was being discussed, 25% imagining how I would create my version of this film (what the content of the conversation would be), and 50% actually paying close attention to the film.
Overall I felt that both characters were good representatives of their positions, and I liked how much they stuck to their observations over their theories (they talked about what they’d seen and experienced more than they made leaky abstractions and focused on those). The main variable that was not discussed, was technology. It is the agricultural and industrial revolutions that lead to humans feeling so out-of-sorts in the present day world, not any simple fact of how we socialise today, that can simply be fixed / gotten out of. Nonetheless, I expect that the algorithm that Andre is running will help him gain useful insights about how to behave in the modern world. But you do have to actually interface with it and be part of it to have a real cup of coffee waiting for you in the morning, or to lift millions out of poverty.
The last line of Roger Ebert’s review of this was great. Something like: “They’re both trying to get the other to wake up and smell the coffee. Only in Willy’s case, it’s real coffee.”
Grade: B+
Books read (only parts of):
Computability and Logic (3rd edition)
I always forget basic definitions of languages and models, so a bunch of time was spent doing that. Jacob and I read half of the chapter on the non-standard numbers, to see how the constructions worked, and I just have the basics down more clearly now. Eliezer’s writings about these numbers connects more strongly to my other learning about first order logic now.
Book is super readable given the subject matter, easy to reference the concepts back to other parts of the book, and all round excellent (though it was the hardest slog on this list). Look forward to reading some more.
Modern Principles: Microeconomics (by Cowen and Tabarrok)
I’ve never read much about supply and demand curves, so it was great to go over them in detail, and how the price equilibrium is reached. We resolved many confusions, that I might end up writing in a LW post. I especially liked learning how the equilibrium price maximises social good, but is not the maximum for either the supplier or the buyer.
It was very wordy and I’d like to read a textbook that had the goal of this level of intuitiveness, but aimed at readers with assumed strong math background. I don’t need paragraphs explaining how to read a 2-D graph each time one comes up.
Jacob made a good point about how the book failed to distinguish hypothesis versus empirical evidence, when presenting standard microeconomic theory. Just because you have the theory down doesn’t mean you should believe it corresponds to reality, but the book didn’t seem to notice the difference.
Overall pretty good. I don’t expect to read most chapters in this book, but we also looked through asymmetric information (some of which later tied into our watching of The Big Short), and there were a few others that looked exciting.
Thinking Physics
I am in love with this book. I remember picking it up when I was about 17 and not being able to handle it at all and just flicking through to the answers—but this time, especially with Jacob, we were both able to notice when we felt we really understood something and wanted to check the answer to confirm, versus when we’d said ‘reasonable’ things but which didn’t really bottom out in our experiences of the world.
“Well, if you draw the force vectors like this, there should be a normal force of this strength, which splits up into these two basis vectors and so the ball should roll down at this speed.” “Why do you get to assume a force along the normal?” “I don’t know.” “Why do you get to break it up into two vectors who sum to the initial vector?” “I don’t know.” “Then I think we haven’t answered the question yet. Let’s think some more about our experience of balls rolling down hills.”
One of the best things about doing it with Jacob was that I often had cached answers to problems (both from studying mechanics in high school and having read the book 4 years ago), but instead on reading a problem I would give Jacob time to get confused about it, perhaps by supplying useful questions. Then eventually I’d propose my “Well isn’t it obviously X” answer, and Jacob would be able to point out the parts I hadn’t justified from first principles, helping me notice them. There’s a problem in discussing difficult ideas where if people have been taught the passwords, and especially if the passwords have a certain amount of structure that feels like understanding, that it’s hard to notice the gaps. Jacob helped me notice those, and then I could later come up with real answers, that were correct for the right reasons.
The least good thing about this book is the answers to the problems. Often Jacob and I would come up with an answer, then scrap it and build up a first-principles model that predicted it based in our experiences that we were very confident in, and then also deconstruct the initial false intuition some. Then we’d check the answer, and we were right, but the answer didn’t really address the intuitions in either direction, just gave a (correct) argument for the (correct) solution.
I think it might be really valuable to fully deconstruct the intuition behind why people expect a heavier object to fall faster. I’ve made some progress, but it feels like this is a neglected problem of learning a new field—explaining not only what intuitions you should have, but understanding why you assumed something different.
But the value of the book isn’t the answers—it’s the problems. I’ve never experienced such a coherent set of problems, where you can solve each from first principles (and building off what you’ve learned from the previous problems). With most good books, the more you put in the more you get out, but never have I seen a book where you can get this much out of it by putting so much in (most books normally hit a plateau earlier than this one).
Anyway, we got maybe 1/10th through the book. I can’t wait to work through this more the next time I see Jacob.
It’s already affected our discussions of other topics, how well we notice what we do and don’t understand, and what sorts of explanations we look for.
I’m also tempted, for other things I study, to spend less time writing up the insights and instead spend that time coming up with a problem set that you can solve from first principles.
This book made me think that the natural state of learning isn’t ‘reading’ but ‘play’. Playing with ideas, equations, problems, rather than reading and checking understanding.
Jacob and I now have a ritual of continuing the tradition of trying to understand the world, by going to places in Oxford where great thinkers have learned about the universe, and then solving a problem in this book. We visited a square in Magdalen College where Schroedinger worked on his great works, and solved some problems there.
You only get to read this book once. Use it well.
Hanging out with Jacob:
Grade: A++, would do again in a heartbeat.