There seem to be few reviews of this book, and almost no citations to it in Google Scholar. I found one review at http://www.bayesianinvestor.com/blog/index.php/2009/04/28/good-and-real/. Quoting from it:
He uses a concept which he calls a subjunctive relation, which is intermediate between a causal relation and a correlation, to explain why a choice that seems to happen after its goal has been achieved can be rational. That is the part of his argument that I find unconvincing. The subjunctive relation behaves a lot like a causal relation, and I can’t figure out why it should be treated as more than a correlation unless it’s equivalent to a causal relation.
I'm having trouble understanding it too. And it concerns me that neither the evidentialist camp nor the causalist camp seem to see a need to rebut or comment on Drescher's ideas.
"And it concerns me that neither the evidentialist camp nor the causalist camp seem to see a need to rebut or comment on Drescher's ideas."
Doesn't concern me in even the tiniest, most infinitesimal amount. Remind me to post on the rationalist virtue of zs'hanh at some point.
Difference between PD and one-shot Newcomb: Agree the incentives are different; agree that the logical structure of the problem is potentially more complicated because of that; suggest that the decision to expend cognitive resources searching for a way to defect could be treated as a defection or a probabilistic defection itself.
Drescher on subjunctives - I agree, this strikes me more as Drescher trying to make partial progress toward a solution than presenting something well-defined in a logical sense. I'm not sure Drescher would disagree with that.
I've spoken to Drescher at length and I think he's trying to derive way too much "ought" from TDT, to the point of thinking TDT yields morality itself.
That said, "Good and Real" is still the reductionist book for now.
You'll have to write that post and explain what heuristics you use to decide who to pay attention to. I think I'm actually relatively good at this (for example I've been following your career ever since "Staring into the Singularity" :) but I wasn't particularly impressed with Drescher until I met him in person.
I'd like to add to the list of requests:
A link to a page of Amazon reviews written by Everyman to Everyman doesn't have the same value as a comment from one of us to the rest of us.
Influence: The Psychology of Persuasion by Robert B. Cialdini
Influence is a great book for new rationalists because it highlights a lot of cognitive biases with real-life examples derived from case studies. These examples are placed in the context of how the so-called 'compliance professionals' explout these biases to persuade people, who might otherwise be not so forthcoming.
In my case, The Selfish Gene was absolutely essential. Its impact on me was enormous -- I spent a couple of years just to update myself after reading it. This book is the primary cause of my interest in AGI, FAI, naturalism, reductionism, science and rationality -- I definitely think it should be on the list.
Added: There may be another important effect to this book. When I was reading it, I had a strong feeling that this book is capable of utterly destroying religious and other non-naturalistic worldviews in readers who can reason more or less straight -- and that without explicitly mentioning religion or any gods at all. I never was a theist, but since religion keeps popping up here on LW, it apparently still is an important issue -- which is another reason why I support keeping the book on this list.
For deep understanding of how computers and especially computation works: "Structure and Interpretation of Computer Programs"
(available freely online: http://mitpress.mit.edu/sicp/)
An obvious one is Richard Feynman's "Cargo Cult Science", which is published in the book Surely You're Joking, Mr. Feynman! but can be found online pretty easily.
Edit: The reason why it's a good read is because - all due respect to Randall Munroe - science really does require a lot more than "let's try it and see". That's the basic idea which got the whole exercise started, but we've learned a lot as scientists since then, and Feynman explains a lot.
(Incidentally, the entire book - Surely You're Joking, Mr. Feynman! (Adventures of a Curious Character) - is a fun read, although somewhat less focused on rationality as a whole.)
Hofstadter & Dennet - "The Mind's I"
Series of essays on self, AI, consciousness, intelligence, ... with Dennet and Hofstadter commenting them. It gives an excellent introduction in the subjects, and a lot of food for thought.
Long time ago since I read that. Maybe I should re-read it and see if it's as good as I can remember...
Rational Choice in an Uncertain World: The Psychology of Judgment and Decision Making, by Reid Hastie & Robyn M. Dawes.
Judgment Under Uncertainty: Heuristics and Biases, collection edited by Daniel Kahneman, Amos Tversky and Paul Slovic
I've tried putting together a list on Goodreads of books recommended on Less Wrong, but it will only be as good as the people who contribute to it.
http://www.goodreads.com/list/show/10819.Less_Wrong_Recommended_Reading
Greg Egan's Axiomatic).
This is his first collection of short stories. Half of them are fictional explorations of what it would really be like, for the mind to be literally a physical process of the brain, usually with some near-future technology thrown in to sharpen the issues. The title story is about choosing your own utility function, "Learning To Be Me" is Philosophy of Uploading 101, "A Kidnapping" is about stealing a copy of someone's upload, and the others look at other aspects. "Seeing" uses no future science, only pre...
Okay, I have to ask: what exactly is so great about GEB? I see it get highly praised, and Eliezer_Yudkowsky goes overboard with praise for it, but I don't understand what's so great. (Yes, the page warns the content may be obsolete, but I think he still stands by that part.)
I've read almost all of it, and while it was enjoyable reading, I don't understand how it's useful as rationalist reading, or for AI. It's just a bunch of neat observations strung together, and a long (but helpful) explanation of Goedel's Theorem. In talking about AI, all I found were ideas that seem quaint now and were bad ideas even at the time, like using semantic nets to attempt to solve visual analogy problems. (ETA: There's also no mention of Bayesian inference or anything like it.)
So, could anyone who agrees with this recommendation, please explain what is good about GEB from a rationalist or AI perspective? Be as specific as you can.
The Simple Truth seems like another Yudkowsky essay worthy of reading.
Edit: The primary reason to peruse it is to get an understanding of why people like the correspondence theory of truth, as well as to show off a lot of the confusions people have when they try to refute - or even support - the theory.
Satan, Cantor, and Infinity by Raymond Smullyan
Smullyan's books are the best introductions to formal logic I know. They are witty, entertaining, and make you think - without it being work.
"The Structure of Scientific Revolutions" by TS. Kuhn.
Enormously influential book length essay about how science progresses. Kuhn describes the idea of "normal science" - the everyday activities by which scientists take incremental steps forward. For normal science to be fruitful, it must be carried out within the context of a paradigm - a theoretical framework and set of shared commitments held by a scientific community. If no paradigm exists, or if the current paradigm is flawed, the incremental steps add up to nothing and no progress...
Falling Free, by Lois McMaster Bujold.
It's a great story, but there's one scene in it that permanently changed my understanding of rationality: Leo Graf's first lecture to the engineering class where he discusses the relationship between engineering and ethics. The argument applies to all science and ways of applying scientific knowledge - really, to any and all attempts to interact with reality.
Don't Believe Everything You Think: The 6 Basic Mistakes We Make in Thinking, by Thomas E. Kida.
Understanding Uncertainty, by Dennis V. Lindley.
It's a friendly and practical introduction to the bare minimum the author thinks everybody should understand of probability and decision theory as they pertain to uncertainty and reasoning under uncertainty.
Before you pick up anything in this thread you would be well advised to peruse How to Read a Book by Mortimer J. Adler and Charles Van Doren. This is doubly recommended if reading is one of your primary ways of acquiring knowledge.
The book was published some time ago, but books, and the reading habits associated with them, haven't changed all that much. The authors make the point that most people, even college graduates, read at an elementary level, and that many educational institutions make no effort to improve this. Elementary reading is characterized b...
While I no longer agree with Heinlein's characterisation of people who cannot grasp mathematics, I do think mathematics is a good thing to know. But on occasions when I have tried to explain what mathematics is all about, or to recommend a book which might do the job, I find myself at a loss. "Mathematics: A Very Short Introduction" didn't do it for one person I suggested it to. Does anyone else have suggestions, especially suggestions that worked for themselves or others?
Reaching the specified limit: The Recursive Universe: Cosmic Complexity and the Limits of Scientific Knowledge by William Poundstone.
This is another book which comes at things from lots of angles - it talks about Conway's Life, Maxwell's Demon (in several variations), self-reproduction ... it clears up a number of apparent paradoxes in an engaging and understandable way and explains a lot of really cool concepts on the way to understanding how patterns can arise from noise.
The Tyranny of Words by Stuart Chase--I haven't reread it since I was a teenager, so this is a little tentative, but it's an introduction to General Semantics, and having it hammered that the word is not the thing, and it's not just that there are exceptions, but the thing behind the word changes over time, were both very worthwhile for me to learn.
By the way, if your idea of General Semantics comes from van Vogt, that's a case of the word really not being the thing.
And why isn't my html working?
A belated, supernumerary entry (many thanks to taw's recent post for inspiration): How to Lie with Statistics. It is, in essence, a catalog of Dark Arts (and common mistakes) in that field, with some closing remarks on defenses against them.
This is not so much a recommendation as a request. Recently I stumbled on my well-thumbed copy of Carl Sagan's Cosmos. It was given to me when I was pretty young, and was, by and large, the first real book about science I ever read. It covers a great range of topics, from evolution and planetary chemistry to astrophysics and relativity, provides comprehensive historical background, and is written in a very personable style. It led me to Sagan's other work, which led to a broader interest in science and, along with a textbook on modern philosophical thought...
PS - Is there a "New Readers Start Here" page, or something similar (aside from "About")? I seem to remember someone talking about one, but I can't find it.
This reading list isn't explicitly for our purpose, but many of the books would be appropriate (several are already recommended here). To pick one off the list that I have read: Daniel Dennett's Consciousness Explained is largely an exercise in speculation, but impressive for its density of reference to scientific literature of the time. If nothing else, it's an explicit demonstration of the non-mysteriousness of an all-too-common choice of mysterious phenomena, and the warnings about the Cartesian Theatre are wise to bear in mind.
"The Structure of Scientific Revolutions" by TS. Kuhn.
Enormously influential book length essay about how science progresses. Kuhn describes the idea of "normal science" - the everyday activities by which scientists take incremental steps forward. For normal science to be fruitful, it must be carried out within the framework of a paradigm - a theoretical framework and set of shared commitments held by a scientific community. If no paradigm exists, or if the current paradigm is flawed, the incremental steps add up to nothing and no progre...
A lot of people coming from a background of traditional rationality identify themselves as falsificationists, unfortuneately. For an entertaining demolition of the work Popper and his intellectual descendants, see Popper and After: Four Modern Irrationalists[1] by the inimitable David Stove.
[1]Also known as: Anything Goes: Origins of the Cult of Scientific Irrationalism and Scientific Irrationalism: Origins of a Postmodern Cult
Thinking Strategically by Dixit and Nalebuff. A introduction to game theory. I would have preferred more math, but the case studies are great ways to apply rational thinking to win.
I am a student, and have recently found this website through a friend. I am intrigued by the art that you are promoting here, and it strikes a chord with my own dedication to learning and becoming a better person. In the past, I have often preferred autodidactic learning; however, this particular subject seems especially dense and in some cases even dangerous if learned incorrectly. I am thus somewhat apprehensive at proceeding along my own course.
What is a good starting point for someone like me? Is there a single grand summary of the work being done here, or some similar statement of purpose and general principles?
Edward Tufte's books on clear, data-rich information designs, especially The Visual Display of Quantitative Information.
In a different direction than (I would guess) most of the posts we'll see here: I Remember Townsend... by Lizbeth Marcs.
Edit: This is a strange one to recommend, here - it feels more like a reminiscence and political argument than a textbook - but I do so because it gets at the importance of the truth from an angle which isn't common among us thought nerds. It isn't an instruction manual, it's a rousing speech.
I recommend any collection of essays by Stephen Jay Gould.
I know, I know. Yudowsky hates him, John Maynard Smith thinks everyone else hates him, Dawkins gave him his due but disagreed, etc.
But, if you read his essays with an eye toward the workings of the mind, specifically how humans think when they theorize (which I consider his main topic) you will find useful things there that you would be hard pressed to find anywhere else.
There seem to be few reviews of this book, and almost no citations to it in Google Scholar. I found one review at http://www.bayesianinvestor.com/blog/index.php/2009/04/28/good-and-real/. Quoting from it:
He uses a concept which he calls a subjunctive relation, which is intermediate between a causal relation and a correlation, to explain why a choice that seems to happen after its goal has been achieved can be rational. That is the part of his argument that I find unconvincing. The subjunctive relation behaves a lot like a causal relation, and I can’t figure out why it should be treated as more than a correlation unless it’s equivalent to a causal relation.
I'm having trouble understanding it too. And it concerns me that neither the evidentialist camp nor the causalist camp seem to see a need to rebut or comment on Drescher's ideas.
Also, chapter 7, "Deriving Ought From Is", doesn't take into account an important difference between Newcomb's Problem and One-Shot Prisoner's Dilemma that I pointed out at http://lesswrong.com/lw/6r/newcombs_problem_vs_oneshot_prisoners_dilemma/. I think this is a fatal flaw for Drescher's argument.
This has been discussed in passing several times, but I thought it might be worthwhile to collect a list of recommended reading for new members and/or aspiring rationalists. There's probably going to be plenty of overlap with the SingInst reading list, but I think the purposes of the two are sufficiently distinct that a separate list is appropriate.
Some requests:
Happy posting!
PS - Is there a "New Readers Start Here" page, or something similar (aside from "About")? I seem to remember someone talking about one, but I can't find it.
1"Everything Eliezer has ever written (since 2001)... twice!" while likely a highly beneficial suggestion for every single human being in existence, is not an acceptable entry. A Technical Explanation of Technical Explanation is fine. If you're not sure whether to classify something as "an essay" or "a blog post", there is a little-known trick to distinguish the two: essays contain small nuggets of vanadium ore, and blog posts contain shreds of palladium. Alternatively, just use your best judgement.