Two thoughts:

• It’s snowing in Minnesota. I love it.
• This Kirkland coffee has been my go-to for a while. Every time I look at the drawing on the box, it looks like a bbq potato chip in the middle of whitewater.

Extra thought:

This box is at my desk. I think about that bbq potato chip every day.

2025 did not go as planned. Honestly, it was mostly because I didn’t have a well-defined plan to begin with. However, the narrative of things going wrong seems more compelling so I’ll stick with that for this blog post.

I started the year by finishing a sci-fi novella. “Finish” is generous here: it still needs lots of love but it has a beginning, an end, and doesn’t sag too much in the middle. This novella was going to be the bulk of my 2025 aspirations.

A few things transpired in my life in April. While cleaning out my Google Drive, I found a paper I wrote in 2009 for my MBA. This was one of many papers evaluating my leadership style through the lens of a framework. This one happened to be a “Jung Typology Test”. I believe by that point in my MBA I had written so many responses to flavor-of-the-week business books that I found clever ways to diverge from the assignment and write papers that interested me. This was a prime example.

I wrote a paper on systems thinking and folded my Jung test results into it, mapping the systems I could influence along with their details and interconnections. The crucial part was my conclusion: if I wanted to have more influence for the good of the world, I should embrace writing because it’s what I’m actually good at and my Jung test results highlighted this with neon lights.

Sidebar: I have written poetry privately for decades. Primarily haiku and senryu. Finding a message from my younger self telling me to embrace it and realizing I'd ignored it for sixteen years was a wake-up call.

With this at the top of my mind, a week or two later I found myself in NYC looking out of the window of an office in One World Trade Center from the 40th-something floor. The pervasive feeling I kept having was this: these are not my people. Thinking of going to a happy hour after work with them filled me with anxiety. I stood there as a person who was valued because I could write code and build things. That is only a fraction of who I am. I knew there was more. I was done denying it.

I flew home and typed up and sent my resignation letter on the Uber from the airport to my house. I had another job lined up, so it’s not as dramatic as it may seem initially. It was more about leaving start-ups behind and seeking out something that fit my desired lifestyle and left me with mental capacity to do the things I want with people I like. When I left that Uber a version of myself stayed behind.

I wrote poetry pretty furiously and, more importantly, I started putting some of my more subversive things about work and life on… LinkedIn.

Why LinkedIn? Why not?

To my surprise, people connected with it. Senryu like these seemed to really connect with people:

employee handbook
didn’t specifically say
naptime’s not okay

and

company vision
is posted in the break room
right by the trash cans

I didn’t stop at poetry. I realized LinkedIn rewards the cadence of authority more than actual meaning, so I started writing technical thought-leadership articles that were actually subversive art.

I published a piece on "Commit-Driven Development" that used corporate-speak to advocate for untested code deployment. I wrote a tutorial on "Understanding Recursion" that was just the definition of recursion looped over and over again. I wrote another article about "11 Super-Effective AI Prompts" that masqueraded as a productivity guide but was actually a linear horror story about a man using ChatGPT to navigate his divorce and bankruptcy.

I wrote “Programming Tutorials” that slowly descended into Lovecraftian horror.

People liked them. They shared them. I think a majority did not bother to read them before blasting them out to their own network. It was a test of the system, and the system confirmed what I suspected in that tower in NYC: the code is running, but no one is at the controls.

My base assumption had always been nobody wants to see what an engineer writing poetry looks like. Especially if that engineer is obsessed by systems, thinks in systems, and writes… mostly about systems.

But I wondered if I was wrong.

I brushed off things, expanded some long-dormant senryu into longer free-verse forms and submitted them to journals. I was as surprised as anyone when I got some acceptances.

I put together a chapbook, including it in my application materials for an MFA and was accepted… with a scholarship. Everything I loved about writing haiku, senryu and prose was especially intensified by free verse poetry.

I put together a manuscript. The Controls Are Not Responding is, more than anything, the sort of poetry I want to personally read. I don’t try to dazzle or impress anyone with it. It’s just me putting thoughts and ideas into the right forms that can contain them.

One last divergence on this origin story and I'll wrap it up. I acknowledge many people want to write. Simply having the urge to write things would not be compelling enough for me to have made these moves. I have a certain inheritance from my dad that I mentioned here in Rolling. You can see this in my LinkedIn examples above: it's an extreme love of hypothetical scenarios. This is the engine behind it all. My dad never did anything with it except pass it on to myself and my brothers. I decided to start writing it down for others to see. I didn't want to be looking back in my last moments realizing I'd squandered this one very unique thing that absolutely makes it all work. Poetry enables me to use this in a way writing prose barely did. That's the difference.

So yeah. That novella? I’ll get back to it eventually. But I have other things going on now.

Today I distilled my manuscript back down to a chapbook: commit, merge, push --force.

Everything I’ve learned about structuring meaning across a full-length collection made this version cleaner, sharper, more deliberate.

The path was nonlinear:

1. Finish a chapbook
   
2. Submit it widely
   
3. Start the full manuscript
   
4. Realize halfway through the chapbook no longer holds up
   
5. Withdraw it from consideration
   
6. Focus fully on the manuscript
   
7. Burn out
   
8. Create a chapbook, this time with precision
   
9. Submit both
   
   It’s not that the original chapbook was weak. Most of its poems survived. What changed was the architecture: order, style, and an understanding of how to modulate pressure across a sequence.
   
   

This is what it looks like to be actively trying to get something published.

Late nights. Weekends.
Nearly a dozen versions of the same book.

I love that you can see the file sizes change across submissions as I add and remove things.

It’s the tiny weight shifts of poems moving in and out, like I’m packing and repacking a suitcase for a trip I’m not sure I’ll get to take.

#poetry

Billy Collins Poetry Masterclass
Metallica - Ride the Lightning
Grandaddy - The Sophtware Slump
Some The Adventure Zone podcasts
Run DMC - Raising Hell
A random smattering of Clint Mansell and Ludovico Einaudi
Der Dritte Raum - Raumgleiter
Tool - Lateralus
Ulrich Schnauss - A Long Way To Fall
A whole lot of silence

Over the weekend this post became something longer.

I’m pretty happy with it. It’s one of the handful of things I’ve written that did not start off as haiku/senryu and when I read it, I can tell. It’s strange how skeptical I am of anything that doesn’t use my process.

I keep adding this new poem to my manuscript and then taking it out. I’m just continually re-reading, testing how it feels in context. Sometimes it’s like a stumble and others a slide without really having changed anything.

One of the greatest things is just going to dinner with someone and being fully present.

Joking. Talking.

There is darkness in my poetry, but it’s in my pocket. I pull it out when I need to have a chat with it, but it’s not a hat.

That’s about as much metaphor as I have in me right now.

Closing out an amazing four-day weekend. Ridiculously productive. The feeling, all around and within, is simply gratitude.

There is no fickle muse. No angels of inspiration.

There is, however, rare alignment in the machinery of the universe that allows things to happen.

When this happens, you must seize it.

I have been writing all day. Undoubtedly, it could continue through the night.

But I am going to sleep.

What would life be like if we did not want to matter? This question has been on my mind.

Many people choose small lives, rejecting the pursuit of power or fame. This is the burning ambition we conflate with mattering. But even the quietest life is built on a framework of necessity: to be missed, to be relied upon, to occupy a chair that would otherwise be empty. Mattering, then, is not ambition; it is being necessary to the algorithm of daily life. The true question is: What if we wished to remove ourselves from that algorithm entirely?

Could you love?

Can you consciously choose not to matter and circumvent the irony of making a choice that matters?

This is what I want to write next. I’ve just got to figure it out.

One further thing I want to talk about as it pertains to the Poetry Playground project I mentioned earlier.

Many, well… actually most… of my ideas for interesting transformations came from reading thousands (really) of Reddit haiku. There were a couple of failure modes I found.

• sentences: If each line can stand independently without the poem collapsing into one flat sentence, you’re often in good shape.
• line jam: a very common failure mode is that a Reddit poet just throws three random things together. Often these are gestures towards profundity of some sort, but it’s not uncommon for them to just be bizarre mutations of language.
• weak L3: probably the most common problem with Reddit haiku. Reddit haiku doesn’t really know about “a turn” and often the poem runs out of steam after L2. So L3 just throws its hands up like “LIFE, AMIRITE???” and exits stage left.
• bridging L2: this one is less common. You have limited space in a haiku to get your point across so if all you do in L2 is pass the time until you get to L3, your poem would have been better off without that line at all.
• L3-L1 swap: I can’t believe how much this one appears. If swapping L1 and L3 doesn’t break the poem, that’s a problem.

I applied this learning directly in my newest feature, which is contextual sampling. This works in this way:

• You choose a word. It’s best to choose a bridging word like “therefore”, but really any word can work.
• You choose a number of sentence pairs to return.
• The system then finds instances of your chosen word amongst the Project Gutenberg corpus.
• It extracts the sentence before and after the sentence in which your word appeared, discarding the sentence in which your word appeared entirely. This is sampling the context.
• Optionally, it can transform the sentences through another thing I have called the “Ship of Theseus” transformer (adds flavor).

This feature was driven entirely by my realization of the bridging L2 phenomenon I mentioned earlier. The results are amazing. Lots of awkward fractures, strange plot twists, and impossible shifts.

So I said I was going to take a break. I did. By doing something new.

I decided to see if I could build a stack of “smart” generative word and sentence fragment tools. I say smart because it’s relatively simple to create random words and not hard to use simpler techniques like markov chains to occasionally eke out a few coherent words. What’s more difficult is to algorithmically and programmatically generate things from human-provided seeds without leveraging an LLM.

I started with just a few ideas, like:

---

• what if I mine Project Gutenberg texts for random sentences and just take them out of context?

  

  ---

• what interesting things can be done with calculating the levenshtein distance between two words and finding other words at the same distance?

  

  ---

• what if I took a word like “thought” and used a dictionary, but instead of looking up the word, I found every definition that has the word “thought” in it?

  

  ---

• what could I build that takes me on a word journey from a word like “hope” to “despair” in a definable number of steps?

  

  ---

• what if I combine a whole bunch of concepts? Say, take the output of one generator and send it into another with the overarching goal of sketching an idea for a poem. Not the poem itself, mind you, but just some words.

  

  ---

• what if want statistical and stylistic analysis of my own poetry?

  

  ---

  None of this uses Generative AI as we refer to it today. Just plain old math, machine learning libraries, and the datamuse API.
  
  There’s more to it than this. But this description should suffice.