The Knowledge Pipeline
Three Things #58: March 5, 2023
My consumption patterns have changed over the years, but what hasn’t changed is that I’m always hungry for knowledge and always consuming information. Sometimes this takes the form of hours lost reading random Wikipedia pages and following links. Sometimes it’s talks and conferences and watching recordings. Sometimes it’s conversations with people much smarter than I am. More often it’s podcasts, audiobooks, and random articles.
For years I’ve struggled with the question of what comes next. What to do with all of this content that I’ve so enthusiastically consumed? Because if you don’t do anything with the content you’ll lose most of it. Taking notes helps. Consuming the same content again helps. So does talking about it or sharing that information with other people.
I’ve recently begun to see content consumption as part of a longer, more interesting pipeline. While it’s fun to consume content, the act of consumption in and of itself is by definition not productive. So the real question is, what do you do with all of that content? Ultimately the goal is to do something productive with the content and knowledge we consume, isn’t it?
Since I spend so much time consuming content, and since that content is so important to me, I’m always looking for ways to optimize how I discover, organize, consume, process, and share content, and the ideas I’m discovering. Like a good engineer, I think it’s interesting to break down the process of knowledge accumulation and transfer into its sequential, constituent steps, to consider how they fit together and how we might optimize each step separately. This process can fundamentally be broken down into three pieces: consume, process, and share. Consumption involves discovering, organizing, and consuming content. Processing involves reflecting on that content, taking the time to understand it and integrate it into our existing knowledge system. Sharing involves transmitting both the original content as well as our own interpretation of that content, and engaging with others to present and test our understanding.
In theory this process is linear and clean. In practice, most of the time you need to cycle through consuming, processing, and sharing ideas. Expose yourself to new ideas. Noodle on them for a while, until you’ve developed or improved upon them, or built something in response to them. Share those things, gather feedback, rinse and repeat.
Thing #1: Consume
I’ll start by focusing on consumption, then look at what should come next. We live in a golden age of content, when good content and the tools to curate, discover, organize, consume, and share that content have proliferated beyond belief. Consumption is sort of the easy part, but not all content is created equal and there are doubtless strategies that are helpful. Here are a few that have worked well for me.
First off, focus on quality over quantity of content consumed. I don’t mind lingering, on a particular piece of content, a book, series, author or genre, for months or sometimes even years, if I find it compelling enough.
This leads us to another useful strategy: follow your heart. My consumption patterns don’t follow any particular theme or thread, other than the fact that I tend to consume lots of content related to things I care about like technology and cryptocurrency. The more interested you are in a topic, the more effectively you’ll consume the content, the more fun you’ll have with it, and the better you’ll remember it and put it into action.
Next, develop strategies for finding the signal in the noise. The biggest issue of living in a golden age of content is that, while there’s a lot of good content out there, there’s also heaps of bad content. Most of the stuff that people produce is garbage, probably 99% of it. Develop the ability to smell the garbage early and don’t be shy about moving onto something else quickly when you realize that something isn’t worth your time. Avoid sunk cost fallacy. (I’ve been guilty of this often enough, insisting on finishing a book just because I started it, even though I’m not enjoying it! Don’t be like me!)
It’s essential to have good tools and routines for organizing content: saving, sorting, filtering. I struggle with this because I don’t like most of the existing tools, and I think a lot about how to improve them, but I do find that certain tools like Pocket and Trello are helpful for tagging, prioritizing, and saving content to consume or refer to later. No tool and no productivity system are perfect. I suggest using generic tools, such as markdown files (Logseq is great for this, as is Github), folders, and tags, and developing a system that works well for you.
Finally, I recommend seeking information in less obvious places. Books, articles, and podcasts are all obvious places. Attending events, watching talks, and having face to face conversations are less obvious places to consume content, but they’re just as important and just as useful—and in some ways they’re even more useful than written or recorded content because of the social cues they inherently contain. It’s much easier to tell if an idea is bullshit, or if the person shilling it is a scam artist, in a face to face conversation than in an article!
Thing #2: Process
Just receiving information is never enough. Once you’ve acquired information you need to do something with it. If you don’t, you’ll lose it, or it’ll lie dormant. You need to integrate it into your knowledge system. There are several good ways to do this.
The most straightforward option is simply to reflect on it by yourself. This step is necessary but not sufficient to full integration of information. Reflection allows you to examine a thing more fully than you can do when you first receive it. Pick it up, flip it around, see what it looks and smells like. Consider how it fits into the other things you already know. Update your priors on the basis of this new information. While this step is necessary and important and is a good initial step after receiving new information, it’s not likely to yield anything especially exciting or novel because in my experience good ideas don’t appear in a vacuum, i.e., during private contemplation.
Processing information doesn’t have to be done strictly by oneself. Others can be involved in the process. There’s admittedly a fine line between “mutual processing” of information and sharing, which is the next Thing, but I draw the line at how polished the presentation is. Sharing rough ideas that one doesn’t yet fully understand and that aren’t fully integrated is mutual processing; polishing those ideas and presenting them formally as a paper or a talk is sharing.
Mutual processing, otherwise known as conversation, is more likely to yield novel, exciting results. Present the idea in an ad hoc fashion, engage in dialog and debate about it, and see how others respond. Encourage other people to challenge you and poke holes, both in the original idea and in your interpretation and presentation.
My personal favorite way of processing information is tinkering and building. Perhaps not every idea is one that’s capable of being built upon or represented in code, but most are in one fashion or another. Try to build something on the basis of an idea to test if the idea is sound and whether it really works. Poking theoretical holes is one thing; trying to turn an abstract idea into a concrete implementation is one of the best ways to understand it and its strengths and weaknesses. It also gives you something concrete to share with others (which brings us to the next thing). Tinkering and building something is a fantastic way to integrate knowledge because building literally means connecting things; the process of doing so in a medium like code mimics the process of connections forming in your brain!
Thing #3: Share
Knowledge and experience aren’t useful if we keep them to ourselves. They’re examples of goods that are more valuable the more we share them. In the same way that insights don’t appear out of a vacuum and instead tend to appear through dialog and interaction with others, we need to contribute our insights back to the overall social dialog, to let people challenge our ideas and build on top of them. This process of give and take is how society advances.
The notion of “sharing” has been so overemphasized in the age of social media that the very term is trite and overused to the point of meaninglessness. We’re constantly “sharing” things: posting photos and videos, posting comments on social media, sharing updates in group chats, etc. But there’s a deeper form of sharing that ties into the knowledge transfer pipeline I described above. Just as we deeply, thoughtfully consume and consider content and new ideas, we should thoughtfully express and share those ideas with others once we’ve had a chance to process, understand, and draw insights from them.
The most obvious, most straightforward way of doing this is through writing: indeed, one of my main goals for this newsletter is to share some of the new ideas that I’m exposed to and find interesting, to advance (or in some cases, start) the dialog about them. But that writing doesn’t necessarily need to be short-form, non fiction, or in the form of a newsletter.
It doesn’t need to be writing at all. There are lots of creative ways to express ourselves and share ideas, from fiction to poetry to visual art to music to software to podcasts. Another excellent way to share ideas is through speaking: one on one, in a small group, or to an audience. The act of expressing ideas verbally or creatively isn’t just for the benefit of others! It’s also a fantastic exercise that forces you to understand an idea well in order to explain it to someone else. Personally, I find that the best way to truly grok an idea is to teach it to others.
If this pipeline I’m describing sounds anything like a technical process, that’s because it’s not unlike how computers receive, process, and transmit information. Blockchain nodes do this. They receive new information (in the form of blocks or transactions), they validate that information (make sure that signatures are valid, that coins aren’t being double spent, etc.), then pass it along to their peers. Of course, they’re not doing any meaningful analysis or synthesis of the information, but I still find the metaphor useful. If we think of ourselves as nodes in a global network of humans responsible for receiving, validating, and transmitting valuable ideas and memes, it makes sense. The network effects are strong: it doesn’t behoove a node in a blockchain network to keep transaction data for itself, nor does it behoove us humans to keep good ideas to ourselves.
This metaphor also suggests ways we might measure our productivity a little differently. Rather than attempting to measure our work output or dollars earned, we might consider counting the number of novel ideas generated (through the process of combining and considering received information), considered, and spread to others through the social network.
It doesn’t matter how you choose to express yourself and transmit the knowledge and ideas you’ve been exposed to. The important thing is that you don’t keep them to yourself. Because this process isn’t freestanding, it’s recursive and cyclical. The ideas you transmit to others will be incorporated into their own loops, and their feedback will in turn feed back into your own design and thought loops. We’re all connected, and we’re weaving this enormous social fabric together. Generating, considering, and sharing novel ideas is by definition the only way human society advances.