What Elon Musk could learn from my poetry collection

The power and purpose behind a good, if imperfect, bureaucracy

I’ve been reading poetry pretty regularly for almost ten years. This might not seem like much time, and it isn’t compared to many. But enough mornings of reading poetry have passed for me to go through quite a few dozen sets of small sticky flags I use to mark poems I like, or would like to return to. Some people have other methods, like dog-earing these pages; others consider this practice nothing short of a capital offense against the book. I’m not quite that rigid with my rules, and dog-ear when I have to, but I always make sure later on to add the appropriate flag.

Color is the main differentiator here. Purple flags denote poems about God, cosmic mystery, or spirituality. Red ones are love poems; green ones, tree or nature poems (oh, so many tree poems); yellow ones for pretty much any poem in which the sun is a featured character, blue ones for sky, water, or ocean imagery; black or other darker shades for poems about death or sorrow. I also try to have handy a set of unique top-tier flags to note poems not just that I liked, but that really blew me away or stayed with me. These are usually wider than the typical flag, and/or have a special design: for awhile, I used a set that were lovely varying shades of tartan to help these ones stand out amidst the parade of flags that litter my most-favored works. For me, a system like this is necessary if I’m ever to stand a chance of retaining or returning to these poems later in a time of need, or if there’s a theme I’m aiming to reflect on for this newsletter.

If you’ve ever wondered where I find all the poems I include in this newsletter… now you know.

Systemization is even more important when dealing with something as fluid and mercurial as a poem, much less a large quantity of them. Even more so than other forms of the written word, poems are not clear-cut math problems; they can be unintelligible in what precisely it is they do for us even as they fulfill an essential purpose. In a poem I read for the first time just a couple of days ago, fittingly titled “To Poetry”, Dean Young describes this phenomenon in a direct address to his own work of art: “No, you don't add up/although you make some crazy square/roots and repeating decimals.” My flag system was helping me cut through some of poetry’s abstractions by attaching colors and concepts to them so I could more easily recall them later.

But before too long, the system I was implementing wasn’t pulling its weight. It turned out that I really liked poetry and, as a result, had accumulated many hundreds of flags across dozens of books. Over years, I swapped in new flags as old ones went out of print, or even began to fall out of their original pages. Although attaching these sticky flags to well-worn pages was deeply pleasing from a tactile perspective, it was quickly becoming clear that this was not a viable long-term solution.

Not only was one flag turning out to be an insufficient method to encompass everything I wanted to remember about each poem, but the sheer number of them was hobbling my attempts to effectively and reliably recall them. My brain’s limitations meant that I couldn’t, and still can’t, immediately conceive of all the creative information I’ve taken in from these poems, nor instantly recall the text of all the poems I’ve ever read about (for example) “the manifestation of God in the natural environment.” Even the flags weren’t really accomplishing this, so I needed a newer, better system to do it for me. What it turned out I needed more than anything else was a bureaucracy: a complex, well-organized and robust system to be able to meaningfully differentiate poems from each other, and to attach to them more detailed data than just “yellow” or “this one was truly great.”

In my case, the answer turned out to be a computer application called Obsidian. In Obsidian, you can create notes — say, a new one for each poem you collect — and attach to them metadata like the author’s name, the poetry collection it came from, or the year it was written. I then could copy in the text of all the poems I’ve flagged, enter them into this application, and have them all at my digital fingertips.

This mad database of poetry is bigger and better than the flag system for a few key reasons. For one thing, it’s searchable: I can search for the word “Winter” and digitally leaf through all the poems I’ve saved that include the word in the text or title. For another, I can tag poems with keywords to make the data richer: instead of just a blue flag indicating that a poem is maybe(?) about the ocean, I can tag “ocean” or “water” in the metadata, and then also include other words that I think are embodied in the poem that have nothing to do with the original flag, and which are far more complex than any flag of any color or design could be (“melancholy”; “American exceptionalism”; “family”). Finally, I can create entirely new notes for themes that you can then backlink, and then easily find all the poems that relate to that theme, or which authors have written on that theme.

The improvement of this new poetic bureaucracy over my rainbow of page flags is pretty clear. And sometimes it’s even a meditative experience, entering all these poems into my little database (for one thing, it’s fun to have an excuse to revisit the poems).

But other times, I must admit, it’s a tremendous pain in the ass. This process has been important and I’m glad I have it, but it’s added quite a few extra layers of work to my poetry-reading process that facially have not much to do with the reading of poetry. I’m really far behind, by the way: I have probably only entered about a quarter of all the poems I’ve flagged over the years, and am reading new poems faster than I seem to be able to enter old ones into my library.

And so we can see that this process is marked by many of the negative descriptors we have for “bureaucracy” in real life: it’s slow, it’s cumbersome, and at some level, it feels coldly detached from the core mission or purpose it set out to achieve. In this case, there’s no denying that it diverts me from the actual reading of poetry. Its highly practical, mechanistic nature is at obvious ascetic odds with what most readers — and, as I understand it, most poets — tend to say is the whole point of poetry: to feel, to be present with the words, to flow emotionally with the text and let it hit you however it hits you, even if you can’t name, categorize, or “tag” it.

At the same time, the results can’t be argued with. My Obsidian library is now an institution unto itself, and an unbelievably useful one. If you’ve ever wondered how I find poems that seem to fit so nicely the theme of what I’m writing about, Obsidian is the answer. It holds more information — not just hundreds of poems, but long passages from books I like and want to refer back to, short biographies of authors and poets I admire, and even the full text of my back catalog of Substack posts, since I can barely remember what I’ve written myself most of the time.

This is the kind of wrestling we have done with bureaucratic systems since we saddled ourselves with them upon becoming a civilized society. The historian and philosopher Yuval Noah Harari spends an excellent chapter on this paradox in his 2024 book, Nexus: A Brief History of Information Networks from the Stone Age to AI. Bureaucracy can be perilously time-consuming and inefficient, yes; but a society as robust as ours (or a library of previously-enjoyed poems as robust as mine) would be lost without it, and could have accomplished few of our great achievements of the past century:

While [bureaucracy] sometimes sacrifices truth and distorts our understanding of the world, it often does so for the sake of order, without which it would be hard to maintain any large-scale human network. While bureaucracies are never perfect, is there a better way to manage big networks?

Anyone who fantasizes about abolishing all bureaucracies in favor of a more holistic approach to the world should reflect on the fact that hospitals too are bureaucratic institutions. They are divided into different departments, with hierarchies, protocols, and lots of forms to fill out. They suffer from many bureaucratic illnesses, but they still manage to cure us of many of our biological illnesses. The same goes for almost all the other services that make our life better, from our schools to our sewage system.

Is the “bureaucratic institution” I use for cataloguing poetry fast? Is it efficient? No, it is not. Nor does it always succeed; after all, the bureaucratic tool (like all bureaucratic tools) is hampered by a human’s imperfect usage of it. But the sheer amount of information contained in the hundreds of poems I’ve done my best to preserve can’t be memorized, categorized, and cross-checked by one brain alone (certainly not mine), and so I need flags and buckets and tags and apps to put them in a clear order so that my brain can find them again. I could throw the situation out entirely — delete Obsidian, rip out all the flags, and say “to hell with this slow-moving, complex bureaucracy” — but I would lose something valuable and hobble my ability to do one of my favorite things, which is to read, understand, interpret, and share poetry.

The road blocks, inefficiencies, and “red tape” of bureaucracies can drive us understandably crazy. I’ve been tempted to toss my computer out the window after seeing how little progress I’ve made, and I begin to wonder if the whole endeavor is even worth it (I do think it is). Thankfully, even if I did, not much is at stake in the scheme of things. If my Obsidian library disappeared tomorrow, I’d be properly pissed but I’d still manage to scrape some poems together for the Substack.

But what happens when seemingly everything is at stake; when instead of poems, we’re talking about dedicated public servants, human people with jobs and missions like feeding the hungry, caring for the sick, and educating our children? What happens when the bureaucracy being disassembled is the federal government of the United States?

---

I’m not going to regurgitate the well-reported details of Elon Musk’s annhilation of the federal bureaucracy at length; mainly because most of my readers are probably aware of it already, but also because it makes me so upset that it’s actually painful to bear witness to it again.

Suffice to say that — via the newly-created Department of Government Efficiency (DOGE) — the richest, pettiest, most terminally online tech mogul on the planet has taken the administrative state by storm, undertaking mass firings of federal workers, imploding entire agencies, and granting himself direct access to the personal data of basically every American who interacts with the federal government. He is doing this all with the blessing of the president, and with the stated goal of streamlining the federal budget by zeroing in on “waste, fraud, and abuse.”

Children at play

Even if we take the man at his word — that he’s genuinely invested in the project of making government more efficient, and not just trying to line his own pockets with government contracts — he’s doing an impressively poor job of it. He’s targeting agencies that have comparatively tiny budgets, make essentially no impact in the national debt, and provide massive on-the-dollar benefits for the people who interact with them. The Consumer Financial Protection Bureau (CFPB), for example — which Musk’s team essentially shuttered a couple of weeks ago — provides a huge return on investment. It costs about half a billion each year to run, but has returned many billions to consumers over the years by taking action against shady business practices.

Musk’s agency purges represent the apex example of the tired, hollow, and vapid “move fast and break things” ethos of Silicon Valley: that ruthless, quick, well-modeled efficiency is King, and never mind how much forest you clear-cut along the way. But aside from heartlessly treating federal employees like lines of code, Musk doesn’t seem to understand at any level that the entire purpose and promise of bureaucracy, the exact way in which it is useful, is exactly the opposite: to move slowly in order to keep systems stable and intact.

The difficulty of moving and changing bureaucracies can be frustrating, but the stability and order they bring are essential to accomplish their goals and serve their purposes, which are too big for any person — even any president — to do by themselves, systemless. We organize all this money, the services they pay for, and the information needed to execute on these services into clear buckets known as institutions — like USAID, or the CFPB, or the Department of Defense — because we’re a big country with big needs.

Are agencies like the CFPB perfect? Is everything that’s included within it as useful as it could be? Of course not. Institutions exist for big reasons, but institutions are run by humans who make mistakes, or don’t always run programs as efficiently as possible. But without the institution, the original purpose in creating it — holding credit card companies accountable when they cheat consumers out of billions of dollars of junk fees — is not accomplishable by one individual person, and it doesn’t happen for free. Bureaucracy, for all its inefficiencies, makes it possible.

---

Musk is gutting these agencies in a way that suggests he neither knows nor cares why they might exist. If his interest were were in delivering government services well, he would make some attempt to improve these bureaucracies, or trim them precisely and surgically, rather than “deleting” (his words) them whole cloth and never looking back except to gloat about how many people you’re putting out of work. Because, again, there are obviously ways these agencies could do a better job, and reforms we could make to save taxpayers money while continuing to deliver essential government services.

But Elon, by blindly zeroing out entire agencies and the people who work at them because they aren’t a perfect algorithm, is doing the government equivalent of deleting entire sections of my Obsidian library in order to save hard drive space (I’m quite sure that, insufferable tech bro that he is, he could appreciate the analogy). Worse yet, he’s not even bothering to make any backups, burning entire books of poetry so that we won’t be burdened with their wasteful contents.

Deleting my little database (which takes up next to no hard drive space, by the way) or getting rid of all my poetry would “solve” my information problem in the same way that disemboweling yourself saves you the trouble of deciding what to have for dinner tonight. But unlike Elon, I’m actually invested in trying to fix a problem rather than “blowing it up.” Good poetry, like good policy outcomes, requires hard, plodding, sometimes-far-too-slow work. It requires patience, and it requires a frustrating and imperfect system to help us do impossible things together.

So I’ll keep plugging away at my bureaucratic task, even if I never fully catch up. I may not be as efficient as DOGE, but at least I have an actual purpose.