The Knowledge Revolution Is Not About Big Data, It's About Well-Connected Little Data

Designer Austin Yang's fanciful iTypewriter

Most often, tech companies are rewarded for giving their customers what they want. But there is a circularity to this notion that bears consideration. Users of apps, for instance, now mostly want what they have been trained to want by… app makers. There’s something wrong with this picture.
Let’s start again. You know how they say we use the new technology as a better version of the one that came before? Think “horseless carriage.” But then, at a certain point, humans realize what the new thing can uniquely do (i.e. drive 90 miles per hour for many hours straight) that the older one could not. We are now at that point with digital technology, but few of us have realized it.
But before we get to the utopian spiel—or the dread singularity—let’s also remember that what the new thing can uniquely do, done to excess, can be ruinous. Think all those cars driving at 90 mph. So, it’s not just that we need to discover what digital technology can uniquely do, we also have to see what it has been doing and decide if some sort of correction is in order.
This is a leading question. I think there is a correction to be made, and I am not alone in that notion.
So, let’s step back for a moment and see, on a global level, what we have been doing with all of this digital capacity. In simple terms, we have been storing and accessing information on a massive scale. Increasingly, our businesses are dependent on manipulating data streams to create value. And consumers have been trained to consume larger and larger quantities of that manipulated data. And even though the smartphone in many people’s hands is, in historical terms, an incredibly powerful computer that is capable of writing to the internet as well as reading from it, the flow remains highly asymmetrical.
On a certain level, the public has been treating the internet like a super-sized repository of the media they already know. We access text, pictures, audio, video and interactive graphics from this massive storehouse as if we were pulling books off a shelf or turning on the TV. Even if we are contributing to this global library through blogging, or podcasting or uploading music, video or Instagramed photos, we are just filling in boxes that others have made for us.
The internet is all of these things, but it is also (more importantly) the relationships between all of these things. And it is from the “glue” of these relationships that our collective knowledge emerges. Right now, for the most part, we only have the adhesives provided to us by the tech companies that have built these architectures in silicon. In order to really make your own individual connections, you need to write some code. If we rely, solely, on the code of others, we will unknowingly be manipulated by it.
This is the overriding theme that has run through my week, but it has been building for a while. A month ago, I first interviewed Lars Hård, the CTO of San Francisco “knowledge design tools” company, Expertmaker. I was intrigued by his concept of “desktop AI,” of a “Photoshop for knowledge.” But, I didn’t really understand what was involved. So I spent a month thinking about what it might be, until I was able to meet with Hård in person this week and have a demo of the software in action.
I am writing up my conversations with Hård in a separate post, but interacting with his software brought a number of things into focus for me in a way that I hope are somewhat unique. And, as the ideas moved through the boson field of contemporary culture, they attracted the mass of like-minded thoughts. Somewhat magically, two books (and their writers), popped up just as I needed them, Mind Amplifiers by Howard Rheingold and Program, or Be Programmed by Douglas Rushkoff.
Check back soon for posts about those books and my correspondence with their authors, but the point of both books is that most people are not engaging with the tools that already exist to create value from their own knowledge. Expertmaker is just this sort of tool.
On a fundamental level, our use of the web has become too much about things and not enough about the interactions between them. Useful here is a basic distinction in physics (thank you, Wikipedia) between fermions and bosons: “Fermions are sometimes said to be the constituents of matter, while bosons are said to be the particles that transmit interactions (force carriers), or the constituents of radiation.” Social media is potentially valuable because it is about interactions, about how things radiate. But here is where the “big data” of social media is of questionable use.

Samuel Arbesman, writing in the Boston Globe this morning, takes on this fallacy in his essay, “Big data, Mind the gaps.” “Big Data might be deep,” he writes,”but it’s not wide.” We know a lot about narrow slices of reality. We turn out to know a lot, for instance, about certain species of dinosaurs, he points out, not be cause they are the most important or exceptional, but because their fossils have been easy to collect.
When it comes to the commercial web, this problem is compounded. The data that is being collected is both what is easy to collect and also what web companies consider worth collecting. Not only do consumers have almost no say in what is done with their data, but they also have little say in what kinds of data are collected.
So where do we get the full width of human experience? From humans, of course. And this is where what I am referring to as the “knowledge revolution” comes in. From Howard Rheingold writing about “convivial technologies” to Douglas Rushkoff advocating for programming literacy to Lars Hård building knowledge design tools, it is the same impulse. To take advantage of the technology we already have, we should stop being so obsessed with (and intimidated by) “big data,” and start from where we are (to paraphrase Buddhist writer Pema Chodron), with our own “little data.”
As an interesting comparison, Google assumes that there is something we need to know that it can provide us with, whereas Expertmaker assumes that we know something that it can help us understand better. But even Google is a perfect example of a “mind amplifier,” and of something we can program. We may not realize it, but we are writing code all day as we google. When you type words into a Google search field, you are constructing a search query of Google’s database. The excellant Google Guide (not affiliated with Google) has pulled together everything you need to know (in exhaustive detail) about using search queries effectively, whether you are a novice or advanced user. The point is that even in asking you can encapsulate a lot of what you already know in the structure of you query through the use of operators or the advanced search form. On the other end of the query, we apply what Rheingold calls “crap detection” to sort out the results into the most useful information.
But this is all about how we interact with the knowledge of others. The starting point for the knowledge revolution is what each of us uniquely knows. If you imagining “big data,” as Arbesman does, as “a series of deep wells, each one plumbing the depths of certain topics,” our own “little data” are the settlements of local knowledge that dot the landscape between the huge well sites. Tools like Expertmaker will help us with what Rheingold refers to as “metacognition,” our awareness of our own thought. As I saw in my demo of the software, using Expertmaker requires, first of all, an openness to your own thinking. You can build a model of what you know by entering examples from your personal experience and categorizing them according to parameters that you think may be important distinctions between those examples. Once you’ve done some amount of this (less than you might think) you can use various visualization and correlation tools to see the impact of what you have done and see if you agree with yourself.
So, conviviality starts at home, it starts with you. To make use of your knowledge you have to begin to have a conversation with yourself about what you know and how you know what you know. Epistemology sounds complicated, in the abstract, but what digital technology has enabled are platforms through which we can build mental models and play with them. Once engaged, this playfulness can extend to what other people know. It is when we begin to map what we know into what others know (or vice versa) that truly powerful and revolutionary things can start to happen. But we only get there by, on some level, writing our own code.
I used that same phrase last week, to refer to the pop musician Gotye, and the connection between what I am trying to describe here and music is strong. For whatever reason, musicians are the individual artists that the public is most familiar with. What artists do, and what each of us have to do more of, is to develop a practice, a method of working, that encapsulates who they are and what they know. What makes artistis artists, is that they can’t not do this. Hip hop and electronic music are contemporary forms most relevant to what is going on with technology because they make this process very explicit. Artists working in these forms make extensive use of sampling, both of the music of others (exclusively, in the case of a purist like Girl Talk’s Gregg Gillis) and (notably in Gotye’s case) of the sounds of physical instruments, abstracted from their source. These samples are really like analogs to the kind of knowledge models I’m talking about. And the way musicians manipulate these samples, the way they create relationships between them, and turn them into something that (at least to their ears) sounds good is the way that they participate in the world of music and in the larger cultural landscape. Their music is the way they encapsulate who they are and what they know in the world. We can all do our own version of that, with what we know.
Just as each technology emerges in relation to what preceded it, each technology also poses a social challenge. As with writing and printing and broadcasting, the challenge with digital media is to resist the totalitarian impulse. New technologies mostly start off as closed systems, because they can evolve more rapidly that way (think cell walls.) But once a technology becomes universally available the dynamic shifts, and open standards allow the growth to be consolidated with other, often competing, technologies. The big tech companies that have gotten us here, have done so pursuing their own interests. And that commercial motivation has succeeded in spreading computers, mobile devices and the web to almost every corner of the earth. But the fact is that if all hardware development stopped today, it would be enough. It’s cool, but we don’t really need the next iPhone. The battles of the titanic tech companies are just about market share, they’re not really about what we humans need. But in the absence of humans making those needs clear through making stuff with all of this available digital technology, battle away they will.
And as musicians and other artists show, there are many ways to be a maker, not just through writing JavaScript (though that’s a good start). Open source projects require the writing of code, sure, but also developing documentation and training methods, managing and researching communities of users, testing the quality of existing code and designing the interfaces and interactions of user experience. The more you are conversant with what you know, the easier it is to develop your own projects and find your niche in the projects of others.
There are many ways to begin, whether by learning to recognize the code around you through Mozilla’s Webmaker tools, to program through CodeAcademy, to launch a startup through General Assembly or to model your own knowledge through Expertmaker. The important thing is to start. The big story about the knowledge revolution is not the big data of big companies, but the little data of individuals.