Month: January 2025

Our “Age of Anxiety” is, in great part, the result of trying to do today’s job with yesterday’s tools – with yesterday’s concepts.
– Marshall McLuhan –

I love books and have since I was a child. Lately, however, I’ve become frustrated by their limitations as I run into roadblocks trying to extract ideas from them. They also limit my ability to make connections between those ideas and those of other writers (including myself). This has led me to scrutinize the utility of the book as a tool and how commerce and emotional attachments impede the free flow of ideas.

While I certainly appreciate the emotional value some of us hold for books, they are at root containers for ideas. Books are subject to deconstruction, and it may be time to liberate those ideas differently than we are used to.

Books are the foundation of any institution of learning. The modern university emerged from physical books. The reason that the British still refer to their education as having “read” at Oxford is because before the printing press, that was the only location those books existed.

Until recently, books were relatively scarce and valuable. My parent’s generation viewed them through a lens of scarcity. This is not the case anymore as I seem to be in a constant battle to economize the number of paper books I store in my house.

Don’t get me wrong. I attach great personal value to my library of books. Perhaps this was something I inherited from my book-loving parents, who probably had 20-30,000 volumes in their house when we cleaned it out.

Books still form valuable connections to my past, both personal and generational. However, that is not an ideational function, it is a talismanic function.

I thought digital books would change my relationship with reading and books, but the design of platforms like Kindle very much mirrors a traditional reading experience (only worse). There is the advantage of portability and no need for physical storage space, but the ideas in them are even more locked than a print book.

Books are containers of ideas. Ideas strategically connect pieces of information. With our traditional conception of the book, the only way these ideas interact is through human brains.

Humans train for years to remember and analyze connections between ideas within books. However, even those of us who have spent a lifetime analyzing writing and ideas are in a constant struggle with what we don’t know. I cannot synthesize ideas I’ve never read.

In college, I developed a system of small Post-it notes to mark interesting passages in books that I was reading or consuming. The system worked reasonably well. I could write limited notes on those posts and could even color code them.

If I knew the book well, the system worked. However, there are always passages that I glossed over when I first read the volume, but suddenly became important in retrospect. I hadn’t marked these. When I tried to retrieve key information, I couldn’t find it, even assuming I knew which book a particular idea was in.

I have wasted countless hours chasing down information and ideas in this way. Initially, I found the highlighting and search functions in a Kindle book were superior to my post-it note method. However, I had to exchange that functionality for the commercial limitations Amazon put on the content. (You don’t own Kindle books, you just rent them.)

As an author, I am far more excited about the prospect of sharing my ideas than I am about making a quick buck. Unfortunately, the iron grip of the publishing industry interferes with the exchange of ideas.

Ideas in books connect to ideas in other books, but this happens through an inefficient process of various citation methods (hence education’s obsession with Strunk & White). Rich citations are very much the exception to the rule. They also interfere with the flow of the narrative when used to excess.

As an author myself, I am honest about the fact that there are many ideas in my books whose provenance I have long forgotten. My books are collections of ideas, reading, and conversations that stretch back for decades.

Most books also suffer from the limitation of text. Text is a powerful tool, but it provides us with a linear narrative that is hard to break out of. Ideas are rarely linear and as an author, I’ve always struggled to organize my thoughts linearly. This is good discipline, but it also limits what I can express.

I have the same problem organizing my book collection. When you have hundreds or thousands of books, logically placing them on the shelves is a constant challenge.

I prefer to organize my books by subject, but I see connections between “unrelated” books. For instance, I put my design books with my constructivism books. Constructivism is a design process and vice versa. But there are many more intertwingled ideas that I cannot accommodate on a one-dimensional set of shelves.

A lot of our struggles to comprehend the world comes from an imposed need for informational silos created by the physical existence of books. Capitalism interferes with this as well. In the current publication/copyright model, a book can only be in one place at one time.

This is not an accurate reflection of the ideas in books, however. This visualization by the Open Syllabus Project shows how the intertwingled nature of books connect different academic disciplines by showing us where they’re being used in college syllabi.

Ideas do not respect disciplinary (or bibliographic) boundaries. When you traverse paradigms, traditional systems of organization no longer make sense.

In the old paradigm, we used books and cataloging systems to provide crude maps to the ideas in them, but in the process of organizing information, these systems siloed the ideas. Generative AI doesn’t need to do that and that has the potential to transform our relationship with the ideas contained in them.

As an experiment, I’ve been scanning and otherwise digitizing key books in my library. I then analyze them by feeding them into a Retrieval-Augmented Generation (RAG) model called Notebook LM.

My goal is to build an idea extraction machine. I want the tools to help me find connections I may miss. This experiment is ongoing.

I want to build a tool to do this visually because this will give me the ability to analyze ideas in completely different ways. That is one of the many ways in which the Knowledge Navigator will change how we see information.

These systems are not replacements for my bookshelf. I will still read for pleasure. I also hope that once we stop flogging learners with books they don’t want to read, perhaps they will read for pleasure as well. We’ve already ruined learning with grades. We’ve done the same thing to casual reading.

The skills learners will need are not dissimilar from the skills we developed as readers, but AI can make those much more approachable. If we digitally adapt our information systems so that generative AI can analyze them, this will make both learning and research easier and more approachable for both formal and informal learners.

To achieve that, we need to free the ideas contained in them. That is a systemic problem, not a technological one. However, the payoffs in both equity and the progress of human ideas will be immense.

“Finite players play within boundaries; infinite players play with boundaries.” – James Carse

 We like our boxes. They are how we make sense of an overwhelming world with too much information and demands for rapid decisions. Yet, they also form prisons for our imaginations. The digital world has always threatened these boxes. AI has the potential to liberate us from them altogether.

The industrial world was a world of boxes. Most of us worked in boxes, and if we were lucky, those boxes intersected with others to provide a richer experience. If we weren’t, we never left our boxes.

The boxes in question here could be physical, like the infamous cubicle. However, these boxes were usually also manifestations of conceptual boxes.

Those boxes are layered: this is my job, this is my company, this is my career, and these are the bounds of my universe.

If you didn’t clearly fit into a box, you were suddenly unhoused from a career perspective. For some of us, that has been our fate. It has also made me exceptional at any job that I applied my creative energies to because I never let myself be fooled or constrained by any boxes I find myself in.

I’m also very aware of the artificiality of edges that others may seek to impose on me. We construct our existences. We can also reimagine them.

I’m constantly running up against people who are stuck in their boxes and find sanctuary and solicitude in them. The rules provide comfort and companionship comes from those who are stuck in the same boxes. If you give them opportunities to transcend those boxes, they ignore or violently reject them.

Generative AI doesn’t respect boxes. That is why so many people feel threatened by it. AI breaks their rules. It brings rapid fire outside influences into their work. It shatters paradigms.

Franz Johansson argued almost 20 years ago now that it is in these intersections between boxes that true innovation emerges. You do nothing new if all you do is the same thing repeatedly.

The digital world has been threatening boxes for decades. Once information became digitized and was no longer confined to paper, books, filing cabinets, etc., it started leaking across boxes. Insularity became a liability, but many institutions resisted pressures to unbox themselves.

Leaders may expound the virtues of interdisciplinarity, but in the real world, most of the people who would benefit from it resist change. Innovation suffers. Networks atrophy.

The other insight that has driven much of my innovation work was Steven Johnson’s liquid networks. Liquid networks allow ideas to flow but give them channels for application.

For liquid networks to work, people must flow between boxes. Channels must be flexible and adapt to the needs of the project.

I have been trying to break down boxes within the networks of education for over two decades. However, until recently, I have been reluctant to step outside the larger box of education itself. I have a deep and abiding love for the work of education, and I come from a family of educators.

One frustration that drove me to pursue my passion outside of formal educational structures is that education boxes are extremely inflexible. One of the more liberating aspects of my startup journey has been to step outside the education world. This move has allowed me to create completely different networks that serve my passion for teaching and learning.

Not that the startup world doesn’t have its own boxes. Many people in it come from heavily boxed environments. They bring that with them because they are seeking to understand and impose a structure of thinking on what they see is a chaotic system.

For a while, I took these boxes at face value because I was used to an environment of navigating a honeycomb of boxes. But over time, I perceived a collection of boxes rather than a rigid system.

We design boxes to mitigate risk. Risk takes many forms. There’s personal and career risk, which is something I have unfortunately never paid much attention to. There’s corporate risk. And then there’s the risk associated with money.

Money drives the creation of boxes. That is what I found in the more structured elements of the startup world. Those who have the money want quick returns. They want to reduce the risk associated with making those returns. This creates boxes because “no one will give you money unless you [insert truism here].”

I am designing the Knowledge Navigator to break down boxes. It does so by showing connections around any chunks of information or between them. It doesn’t actually break down boxes, but at least it shows you what you’re missing if you never look over the edge of where you are.

As part of my marketing efforts, I designed a map of some products that the Knowledge Navigator could provide in various fields. The map is necessarily incomplete because, like every tool I’ve ever built, people will find uses for it I have not imagined.

However, in building this diagram, I immediately saw connections between the boxes that the use case scenarios represent. The Knowledge Navigator would do this work automatically. It took me quite some time to do it manually.

When I make these intersections, it breaks boxes. The true power of the tool lies in connecting the unconnected. For instance, a company needs someone to do a particular set of tasks. That person is out there.

However, the systems for connecting that person to company needs are horribly inefficient and time-consuming. They stretch through personal, informal creation to the education system to various layers of human resources before they get anywhere near the person who actually needs the skills. Those boxes are all disconnected.

The Knowledge Navigator could break through all of those boxes by connecting information siloed within them. The potential employer gains immediate perspective that allows the company to navigate to the needed talent.

So many of the problems of the world are at root problems of a lack of perspective. We can educate and feed everyone, but our boxes tell us that’s impossible. As the boxes collapse, we feel an increasing sense of desperation and retreat further into what’s left of those boxes.

We are not rabbits. I hope we’ve evolved past that. However, those instinctual flight instincts run deep and are associated with fear. It’s way past time that we put our heads above the horizon and see out of our perceptual boxes. That’s what I hope to change with the Knowledge Navigator project.