As discouraging as it is to return to LFS101 material regarding the bootloader process, I cannot think of a better time to review it. I was beginning to get lost in the language and tools of the computer and starting to forget something very simple: it is a machine with simple rules and moving parts, just like a wind turbine or a car. If you understand the machine, you don’t necessarily need to know all the nomenclature. The names of the tools and the parts become less important, and their purpose more apparent. The course of action becomes more natural.
I also panicked as I imagined my first interview—being asked about my credentials from LFS207 and still not being able to fully explain how a Linux system works. If I don’t turn back now, I might get so far ahead that this point—this clarity, this opportunity for impression—becomes a relic: obtainable, but mostly lost and forgotten. Then what? I’m just mechanically typing commands without thinking, and no longer providing any real value as a human compared to a computer.
Last night I had a flight of ideas. One of them is a concept for a new form of education—perhaps two. The first I will refer to as Deliberate Environmental Frictional Learning, or DEFL, as Shevek coined it. I will put the honor where it is due. Shevek merely named it; the thought itself is my own. I feel that in the 21st century it will become increasingly important to distinguish between the two.
Deliberate Environmental Frictional Learning is the antithesis of modern education. Modern education is pedantic (and for good reason), but not interactive by nature. It typically takes place in a classroom with 25–40 students listening to one person describe a thing, taking notes, while never touching or seeing that thing—and often not doing so for years. The person describing the subject (the teacher) is a human being with bias, first and foremost, and cannot interact individually with each student as they learn. This results in endless lost potential.
DEFL, on its own, resolves much of this. DEFL exploits the root of learning: experience. Modern education has become industrialized and commodified; DEFL breaks this structure down entirely. DEFL has no classrooms and no teachers. DEFL does not even have a lesson plan—other than the one that unfolds from the student or students within an immersive experience designed for their education.
On its own, DEFL would resemble what most people would call “field practicals.” Apart from the independent study students are naturally motivated to pursue, there is no instructor, guide, or formal lesson plan. Limited insight is provided at the outset, so there is no expectation of pass or fail as there would be in a traditional field practical.
In the 21st century, we undoubtedly possess the most educated population to have ever lived, across all disciplines. Our collective intelligence—especially regarding how to convey information through experience—is already there. What I propose is that like minds collaborate on a “sandbox” for a given subject (forestry, biology, computer science) and develop a schema based on how their own awareness and understanding of that field first emerged and evolved.
A sandbox is simply this: a prepackaged, fully immersive, real-world experience. The sandbox itself is the lesson. And who provides the information? In the 21st century, it is always at our fingertips. Depending on the subject matter, the sandbox would include curated access to scholarly journals and reference materials, designed not to answer every question directly, but to provoke more questions and deeper introspection.
This is the difficult part to grasp: when does learning happen? It is always happening. It is happening right now as you read and reflect on DEFL. The defining feature—the zinger—of DEFL is that the learning is by design, and it is experienced. The friction of the sandbox experience is intentionally calibrated to inspire inquiry. Those who benefit most will ask the right questions, and those questions will guide the lesson. You get what you put into it. The sandbox exists to help maintain momentum and inspiration. Real learning requires discipline—and discipline is only inspired.
DEFL sounds like it would require massive collaboration, but here’s the thing: that problem is already solved. As we speak, AI systems are actively mining knowledge from professionals across every discipline. We already have supercomputers with access to the answers. We do not need instructors to collaborate anymore—they already have, whether they realize it or not.
What we can employ instead is an AI Sandbox Method. An AI disciplined in a given field could generate a sandbox for its student or students. It could assess learning style, aptitude, and proficiency, and factor those into the design of the sandbox as the technology matures.
It’s just a thought. Think it’s impossible? If you’re reading this, ask yourself: what am I doing right now, and documenting? DEFL is simply the act of transferring this process to a different subject.
But I digress. I am still intimidated by the bootloader process. I will persevere.
One thing that stands out to me about the labs Shevek makes is that you are required to read all of the files before doing anything at all—before even executing a script. A script may deliberately not work because something else must be done first. This is by design. I SCP the sandbox to my VM, extract it, and read the primary README.md. It instructs me to start by running a script—but I know better. I first run tree -R to inspect the directory structure and read every README.md before proceeding. No surprises this time. This is a discipline I intend to continue, along with reading scripts before executing them.
The second mission asks me to examine the contents of run/grub.cfg and run/cmdline. It looks like an alien language—not a good sign. I take screenshots and consult Shevek.
“What am I looking at here?” I ask.
/run/grub.cfgFriction slows the hand, Machines speak when named parts fade— Understanding hums.