You’re as rich as your database

Wednesday, July 25th, 2018

Ryan Holiday shares the 5-step research method he used for Tim Ferriss, Robert Greene, and Tucker Max:

Prepare long before gameday

In The Black Swan, Nassim Taleb proposes a test.

Someone walks into your house and sees your many books on your many bookshelves. Have you really read all these? they ask. This person does not understand knowledge. A good library is comprised in large part by books you haven’t read, making it something you can turn to when you don’t know something. He calls it: the Anti-Library.

When I met Robert Greene and he asked me to become his research assistant, the only thing that changed was that I started getting paid. I’d already marked and organized hundreds of books with interesting leads and material and many other relevant books I had yet to get to.

Learn to search (Google) like a pro

Take a tip I learned from Tim Ferriss. When he was researching for the secrets behind health and diet, he used a few specific hacks to drastically reduce the search area he needed to cover.

For example, if he was looking for a high level athlete, he’d go on Wikipedia and find the world’s first and second-best athletes in that sport from a decade ago. Why? It meant there would be plenty of available material and unlike athletes currently on top of the world, these would be willing and available to talk.

Me, I like to look for phrases, in order to find unlikely subjects.

Go down the rabbit hole (embrace serendipity)

In a book about media (and my own personal experiences) I somehow made my strongest citations from Ender’s Game, The Long Goodbye by Raymond Chandler, All the King’s Men by Robert Penn Warren, Status Anxiety by Alain de Botton, and a handful of other books.

Directly, these books had nothing to do with what I was writing about, but because my mind was primed to see connections, I found them in the most unusual places.

When in doubt, turn to the classics

Remember, there is nothing new under the sun. And if we’re all just saying the same things with new words, what quote is going to have more authority: one from Tacitus or some flavor-of-the-month blogger?

The Classics are “classic” for a reason. They’ve survived the test of time. Consider it the survivorship bias put to your advantage.

Keep a commonplace book

While researching an article I wrote for Tim Ferriss’ blog, I discovered (and subsequently borrowed) a tool used by the famous essayist and experimenter Montaigne.

Montaigne kept what he called a “common place book” — a book of quotes, sentences, metaphors and miscellany that he could use at a moment’s notice. I keep a more modern — but still analog — version of this book. I write everything down on 4×6 note cards, which I file in boxes. (You could do this digitally, I suppose, but the physical arrangement — being able to lay them out on a desk — is critical to my work). I picked up this system from Robert Greene, who also used it for his books.

This means marking everything you think is interesting, transcribing it and organizing it. As a researcher, you’re as rich as your database. Not only in being able to pull something out at a moment’s notice, but that that something gives you a starting point with which to make powerful connections. As cards about the same theme begin to accumulate, you’ll know you’re onto a big or important idea.


  1. Senexada says:

    This advice is excellent. I began doing it with a text file and I find myself consulting it frequently, including for comments here.

    Regarding the classics, a prominent 19th century Harvard man (Lowell) said it thus:

    Yes, there is a choice in books as in friends, and the mind sinks or rises to the level of its habitual society, is subdued, as Shakespeare says of the dyer’s hand, to what it works in. Cato’s advice, cum bonis ambula, consort with the good, is quite as true if we extend it to books, for they, too, insensibly give away their own nature to the mind that converses with them. They either beckon upwards or drag down.

    We are apt to wonder at the scholarship of the men of three centuries ago and at a certain dignity of phrase that characterizes them. They were scholars because they did not read so many things as we. They had fewer books, but these were of the best. Their speech was noble, because they lunched with Plutarch and supped with Plato.

    We spend as much time over print as they did, but instead of communing with the choice thoughts of choice spirits, and unconsciously acquiring the grand manner of that supreme society, we diligently inform ourselves, and cover the continent with a cobweb of telegraphs to inform us, of such inspiring facts as that a horse belonging to Mr. Smith ran away on Wednesday, seriously damaging a valuable carry-all ; that a son of Mr. Brown swallowed a hickory nut on Thursday ; and that a gravel bank caved in and buried Mr. Robinson alive on Friday. Alas, it is we ourselves that are getting buried alive under this avalanche of earthy impertinences.

  2. Sam J. says:

    I have a bunch of text files that just cover links and post I’ve made on different subject, hmmm…I have 357 of them. Then I have a folder called papers that has 273 folders with pages I’ve kept on different subjects. In some of those folders I have a text file where I have notes that I have made on the subject in the folder. It’s far from perfect but it helps me to organize things I find interesting.

    Being able to lay out cards on a table has great benefit but the time to write all these out would be prohibitive and I don’t make a living from it so it’s not worth the trouble.

    My papers folder is up to 133GB now. It has books, videos, links, etc.

    There’s software that can link links, save pages and link the ideas together like this. I found the idea intriguing but it’s not so portable as a bunch of folders and text files. Sometimes I have the same file in different folders and also have a text file telling me to look in a different folder for links inside a different folder.

  3. Harry Jones says:

    Indexing is the key to everything. It does you no good to have a piece of information if you can’t find it when you want it.

    If you can find it, but only with effort, that’s inefficient.

    I once did my own note taking software, so it would work exactly the way I want it to work. My only regret was I did it in Perl. It got to the point where I wanted to redesign parts of the user interface for better workflow but the code got in the way. I re-wrote it all in Python and got past that roadblock.

  4. Kirk says:

    I’ve often wondered why the hell nobody has come up with a software solution for this, one that you could start using in grade school, and continue on with throughout your life.

    Lotus Magellan used to have some utility for that sort of thing–I remember one of my old bosses who had that software set up on his system (back before Windows was even really a thing…), and he could make that program sit up and dance. Sadly, though, Lotus never did a damn thing with the concept much past the original release.

    This is an area that really makes me wonder if we’re ever going to make effective use of all this technology we’re running around with. The smartphone needs an information app and infrastructure built into it, such that organizing and managing information is essentially idiot-proof and intuitive. I can’t even begin to enumerate the times where I’ve run into a situation, and it’s been like “Oh, hell… I know I saw a solution that would fit here, a few months/years/decades ago… Where the hell was that, again…?”.

    Case in point: I had to go digging on a search site to try to remember what Lotus Magellan even was called…

    Really makes you wonder, it does. How long is it going to be before we really adapt to having these smartphones around, and really using them as more than mere toys and communications tools? It really should not be that hard for someone to do in software, and I’ll be damned if I can understand why there’s no real market for what could be termed “aide memoire” apps.

    Not to mention, why the hell isn’t personal networking built into the basic phone services? I should be able to just wave my phone near the five other parents on a field trip, and build an instant phone roster/comms list that enables the six of us to talk and text as a group. As well, why the hell aren’t impromptu interpersonal mesh networks a “thing”, anyway? If you’re out at a park, and some kid goes missing, there ought to be a way that every phone within geographic reach of that site gets informed, and you can then opt into search and rescue for that missing kid…

  5. Albion says:

    Years ago, when doing some teaching, I told my students that it was less important what you know than knowing where to find the information you need. It was also important to know that you didn’t know, so could find out.

    In this respect, organising and filing and being aware there are places you can go to for accessible information works very well.

  6. Harry Jones says:

    There’s software out there, but it’s not well known. I don’t even know of a name for this class of software, except it’s sometimes called “note taking software”.

    After I had begun rolling my won, I stumbled across “zim”, “kjots” and “nevernote”. There’s also Evernote from Microsoft. By this time I was too used to my own way of doing it.

    Maybe the problem is there are too few people who can appreciate the value, making it hard to market. Any PHB can see the use of a PIM, but to move beyond mere information management to managing knowledge and insight… that’s a niche market.

    The invention of hypertext was a major breakthrough, but the Web has not evolved in the most enlightened direction.

  7. lucklucky says:

    You just have to see how people run idiotic Office, LibreOffice spreadsheets as an sort of note taking software for all this. I do it too.
    It is dreadful.
    I often say that we are still with XX Century software but with XXI Century hardware.

  8. Kirk says:

    You have to wonder, sometimes… Just how long does it take before adaptation, real adaptation, to new technology really “takes”. I think we’re only at the dawn of beginning to make real use of the potential in things like smartphones, and we’re mostly using them for inane commentary, pictures of friggin’ cats, and taking pictures of our dicks to send women, who we ask for reciprocal nudes. Is this what it all culminates in?

    We really should be using the damn things as mind adjuncts and tools–Yet, half the damn time, I’m having to wade through a literal pestilential swamp of half-wit advertising and mindless, useless dreck.

    Real utility could be built into the damn things, but the people who designed the software and hardware were apparently unable to conceive of anything really utilitarian, like having a “public mode” near-field transfer function that would allow you to hold down a button, wave the phone near someone else’s handset, and then get all their public data for contact, network, and ad-hoc work groups. You should be able to build a workgroup on the fly, enabling ad-hoc creation of human networks to accomplish things, even such prosaic social activities as pub crawls. Yet, you want to get someone’s contact info? You have to have them text you their information–Whole thing is completely awkward, and the root functionality that should have been baked into the first smartphones hasn’t even been worked into the infrastructure. I see no effort on the part of either Apple or Google to do this, either. It’s like they’re more interested in the toy functions than the tool functions…

    Frustrating as hell, when you consider the potentialities. Why, for example, do we not have public opt-in work groups, such that you could link in to the nearest physical law-enforcement agent in an emergency? Instead, you have to call the 911 operator, and they’ll do everything in a top-down manner, which costs tremendous time in an emergency. The critical thing in most situations is getting the right information to the right people who are in the right places. Smartphones are able to do the rudiments of this, but we’re simply not even considering the potentials or trying to build the infrastructure in either physicality, law, or mindframes.

    As an example… Say there’s a child gone missing at a public venue, perhaps a park. Out of 200 people there, someone probably saw that kid wandering around and went “Huh… That’s odd… Oh, well, none of my business…”. Were the phones to be linked, voluntarily, into a location-sensitive network that could instantly link to everyone in that area that there was a missing small child wearing blue shorts and a red t-shirt…? Well, odds are pretty good that someone saw something, maybe a crying kid wearing that outfit and being led off by an adult. Under normal circumstances, that witness might never know they saw something of value to the search, at least until they went home and maybe saw a news report. With the right functionality, a spot report would be able to be sent by someone to everyone in that area at the moment that kid went missing, and then you’d have the eye-witness reporting what they saw almost immediately–”Kid in red shirt and blue shorts seen crying and being put into a silver gray minivan…”.

    Virtually everything we need to know in these situations, is known. The problem is that the right people aren’t able to ask the right questions of the people who know, and there’s no effective way to network these issues on the fly. If we were building smartphones intelligently, the ability to do these things would have been baked in from the beginning. Root problem is that nobody is looking at the issue from the standpoint of what’s possible, just building from what we’ve done before. The vision to see what the potentials are simply isn’t there… And, maybe never will be.

  9. Harry Jones says:

    For what it’s worth, some services allow people to swap contact info quickly via a QR code. One phone shows the code on its screen, and the other reads it via the camera.

    I wouldn’t want to share too much personal info automatically. Privacy concerns.

  10. Kirk says:

    Harry Jones,

    What I’m getting at is not something you have to go through multiple steps on each end to do, and which would be built into the damn architecture of the phones. Let’s put this into context of an actual real-world event, like organizing a set of parents to chaperone a school trip. Right now, there’s no easy way to build a damn phone tree so that those parents can talk to each other, nor is there an easy way to make an ad-hoc network you can group text or send alerts to.

    What should be basic functionality is this: You have an “open frame” that you put your basic information into, like name, email address and whatever else you feel you want to share with everyone. That’s in a sandbox on the phone; when the time comes and you need to build an ad-hoc network, everyone who is participating hits the “share info” button, and taps their phones together to initiate transfer. After all the parents on the trip literally “touch base” with each other, that’s put all the contact information into everyone’s phones, and you can start doing your thing without having to spend thirty minutes tracking down everyone’s phone numbers and contact information.

    Further functionality would be built into the apps using this feature, such that you could build your hierarchies and teams using the gathered data. Information could be transferred and transmitted as the day goes on: “Hey, who has Harry Jones with them? I’m missing a second-grader from my bus…” “Hey, the zoo has a really neat display right now of the new baby elephant… If you get here in the next half-hour, you can see the baby with its mother…”.

    Root problem is, nobody seems to see the capabilities these things enable, or see how those capabilities can enhance life for the average person. The modern smartphones are basically super-expensive toys for adults, more than tools for daily living.

    Further imagine a situation where phones and geographic data are being more effectively utilized. Let’s posit a system where first responders like cops, EMTs, and doctors are able to opt into a communication network that keeps track of where they are, and further tracks capabilities they have at the time. This would enable 911 to vector in whoever is in the area that could best help, like a cardiologist or off-duty cop.

    I have a friend who is only alive due to sheer coincidence–He had a massive coronary infarction, and he happened to have it literally within arm’s reach of a total stranger that turned out to be one of the best cardiac specialists in the state. My friend didn’t even realize he was having a problem until the cardiac guy walked up to him with a cell phone in his hand, calling the hospital and an ambulance for him. Sheer chance, and if the cardiac specialist hadn’t happened to glance over at him, see the mottled coloration of his face, and recognize the other symptoms…? He’d be dead. Cardiac specialist later commented to my friend that he’d been at a soccer game for one of his kids, years earlier, and someone had died of a heart attack that he was pretty sure he could have saved them from, if he’d only known about it–But, it happened out of his sight, and away from where he was sitting at the game. First he knew of it was when the ambulance showed up to cart the dead body off…

    Not so long ago, these little matters of organization and timing were things we could do nothing about. Today, we’re starting to build the tools and networks we need to have in order to do things about this stuff, only we’re just not bothering…

  11. Harry Jones says:

    I’m old enough to remember phone trees. I never liked them. Too prone to human error.

    A private chat group, perhaps? It needs some advance prep, but that would be part of the prep for the outing as a whole. An outing shouldn’t just be thrown together anyway.

    “Ad hoc network.” Are you using that phrase in the technical sense? You’re concerned about functioning where there’s no service? In that case, “first responders” would be beyond reach regardless. If you use the wifi in ad hoc mode, the range will be so limited you may as well yell for each other.

    Now if they were to build walkie-talkie functionality in, that would be handy.

    My concern about making things too easy is security. I just don’t trust Big Data that much. But keeping the range limited will mollify that concern. Walkie-talkie is analog, and we can make it low power, so that everyone who has no business listening will probably be out of range.

  12. Kirk says:


    I’m using the term “ad-hoc network” in the sense of human organization. Smartphones should be easily capable of doing the legwork for building phone trees and group chat functions on the fly, and the fact that they can’t, along with the fact we didn’t build that sort of secure capability in from the beginning…? Indicative. Highly indicative. The smartphone is, in my opinion, potentially as big a change as the shift from oral communications to writing–And, we’re operating in complete obliviousness to that fact. It’s as if the whole of early written work consisted of pornographic graffiti, or something equally inane.

    Mesh networking with stronger radios would be cool, but that’s a technical thing that I’m not really trying to get at–The “human software” side of things, the cultural uses and norms surrounding what these devices are capable of are what we need to be looking at and working on developing.

    Right now, you have an emergency, or other situation going on, what you have to do is call a central control node, ask for help, and wait for an official response. Smartphones could, in theory, enable much more effective responses by linking you to the nearest person with the capability to deal with your problem. Having a heart attack…? Let the phone vector in the nearest heart specialist or emergency medical technician. Someone’s breaking into your home, while you’re there? The phone allows you to find the nearest person who can help, who may not even be affiliated with official law enforcement.

    What I’m visualizing is a world where you don’t have the hierarchy and the bureaucracy for things like emergency services. Instead, you use the capabilities built into the phone to self-organize and get help for your problems from other people who are not beholden to the “system”. Time was, that was how law enforcement worked in this country–There were no police in much of the countryside. If there was a problem, then the hue and cry went out, and the citizenry responded. In terms of effectiveness, I think this is a better model than a lot of what we’re doing, because it puts the responsibility right where it belongs–On the individual citizen. Of course, a lot of folks are going to look at that way of doing things, and say that it won’t work, or that it would encourage vigilantism, but my point in response would be that the current system of official police forces and so forth ain’t really answering the mail, either. In truth, the final arbiter of whether or not the law is going to be followed is the individual citizen, not the force of the state. Convince the majority of the public that the police forces are illegitimate and/or ineffective, and “Poof”, there goes your public order. The reality is that our current system can only keep the lid on things so long as the public-in-general consents and believes in the system hard enough. Damage that consensus reality enough, and you’re not going to be able to police the community with anything short of an occupying military force that has superiority in numbers and weapons. The current ratio of police-to-policed ain’t gonna cut it. They’d evaporate overnight in the face of real social disorder, and I think we all know that. The system I’d advocate would be far more stable, in that there wouldn’t be a single point of failure.

    Fact is, when the shit really hits the fan, you’re only going to be able to rely on yourself and your neighbor. The “forces of public order” are going to be too damn busy trying to survive themselves. Picture the likely consequences of “the big one” hitting the Pacific Northwest–Expectations are that just about everything west of the I-5 corridor stretching from about Eugene north to Vancouver is going to experience the vast majority of its road communications being completely wrecked. The situation in most of those communities is going to be one of incredible isolation from traditional “authority”, and the public had better get used to the idea of self-organizing and relying on oneself and ones neighbors.

    Smartphones with mesh network capabilities would go a long way towards ameliorating the whole situation, and enabling recovery. What would go even further would be the mental conditioning for self-reliance and self-organization in the face of such disasters…

  13. Must I? says:

    “I once did my own note taking software, so it would work exactly the way I want it to work. My only regret was I did it in Perl. It got to the point where I wanted to redesign parts of the user interface for better workflow, but the code got in the way. I re-wrote it all in Python and got past that roadblock.”

    Any chance you could describe your software somewhat? Thanks.

Leave a Reply