On the 12th of February, I made a blog post where I laid out my disappointment in barely blogging in 2025, how it was such a blur and 2026 was going to be #NewYearNewMe and I was going to get back to blogging more.
However, I wasn’t going to get New Years Resolution syndrome and make an empty promise, so I added some accountability to the post:
If I fail to hit the end-of-February deadline, I have to post a clearly GenAI Hustleslop LinkedIn Status about my failure to hit the first accountability deadline of 2026
Not only that, I took the most hallowed bond possible… I posted it as a LinkedIn status with that very promise.
So did I make it? Yes and no. I managed to force something out literally at the last minute, cutting it super fine and pushing it with seconds to go:
Sat, 28 Feb 2026 23:59:26 GMT
I refreshed the website… and it didn’t show.
+++
author = "Peter Souter"
categories = ["Tech"]
date = 2026-02-28T23:55:00Z
description = ""
draft = true
slug = "supposed-to-be-fun-why-youre-going-to-learn-about-flags-today"
tags = []
title = "Supposed to be Fun: Why You're Going To Learn About Flags Today"
keywords = []
thumbnailImage = ""
coverImage = ""
+++
I’d forgotten to switch it from draft…
Sigh
Accepting the L
Technically… TECHNICALLY! I did blog three times. Just one of them was a draft…
But let’s be real, that would be a letter-of-the-law thing. And honestly, I wasn’t proud of the post. It was unfocused and rambling and clearly done in a rush. I’m not happy with that and it’d leave a sour taste about blogging for the future.
So let’s nuke that (feel free to go read the post from the commit if you want I promise you’re not missing much) and replace it with something better.
As for the LinkedIn punishment… well, a bet’s a bet…
But doing the thing and failing actually helped me reset a bit. Throughout this whole thing, the stakes were incredibly low - it was just for me, nothing work or family related, and my main punishment for failure is a little bit of cringe.
So I dusted myself off Saturday morning, had a screen break for most of the weekend and touched some grass, including baking a cake with my kiddos. Then I used the wee hours of a lazy Sunday to do some self-reflection.
How did this happen?
First Blog: 2026 Blogging Plans
The first blog went smoothly: small scope contained scope, mostly pulling from existing Todoist notes and git history. Just me and my notes. One down!
Second Blog: Cfgmgmtcamp 2026 Trip Report
The Cfgmgmtcamp 2026 trip report started well - I had a solid base from my usual conference prep: photos, notes, schedules and scratchpad writeups. Plus it was a fun event to think back on, especially as it’d been so long since the last one!
But I got bit by the scope creep bug… I wanted to elaborate on the talks I’d attended, then the one’s I’d missed, then about the general vibe of the con, patterns from last year…
I’d already been tinkering with a bunch of GenAI stuff for this kinda stuff, so I thought that’d help with burning down the backlog. I was re-checking my own notes with NotebookLLM’s reports on Youtube transcription, Claude Code for drafting, Gemini 3 with it’s increased 1m context for fact-checking, ChatGPT for restructuring. Each tool to speed thing up actually bogged me down.
On top of that, half way through, I started having a nagging gut feeling. I’ve been an AI skeptic until pretty recently, as the tools tipped into acutally useful, and things like Claude Code genuinely felt useful and helped me with coding and creating sandboxes for customers. But with writing, especially about an event I’d attended in person and had so much nostalgia and good feelings about being able to go back to after Covid, a new job, new house and a kid?
What’s the tipping point when there’s so much LLM in your writing process that you’re losing the actual human side of writing?
It did lead me down some interesting rabbit-holes on how people are addressing this, such as the /ai Disclosure Manifesto, a framework for transparently labelling AI involvement in your websites prose.
But I realised I was getting a bit navel gaze-y and the clock was still ticking, so I split the difference and added an AI disclosure note to the post as a starting point, but all of that gilding the lilly took it’s toll.
I didn’t get it merged until the 24th, leaving only 4 days for the last blog.
Third Blog: Inspiration Overload
To add to this, I’d fallen for the rookie mistake: Trying to start new WIP when the existing wasn’t finished.
And maybe it wasn’t just one…
Honestly I was spining my wheels at this point, with 3 earlish drafts across wildy different topics:
- A tribute to [Terrance Gore] who had tragically passed in the middle of my
- A pondering on Staying Human In The Age of AI
- … And a post about cool and unusual flags from across the world.
There is some context to that last one (I’m a huge vexillology nerd at heart) but each one had some thing that would give me a bit of a mental block when I tried to think about the structure and the flow.
The AI discomfort really flared up around the Terrance Gore tribute. It felt… I dunno… icky really. Anyone with access to a ChatBot can recurgitate boxscores, stats and timelines, but a tribute to someone needs to come from a human place. One’s memories, feelings, voice.
I ended up flip-flopping between my own from-the-heard feelings about the loss of someone so young and such a unique and electric character to run the bases, then it’d emotionally-jackknife into a litiny of stats and rote history where the AI-ness practially dripped through the screen like a miasama.
So, like anyone sensible, I ended up picking secret option 4 and trying to cleverly do a replay of my first blog post where I just talked about my approach for the 3 Blogs In Feburary thing (meaning only 1 of my 3 blog posts actually had a specific point vs being an elevated planning doc), but I don’t like losing and I thought, do the damn thing and then use that as motivation to actually go back and put your heart into one of the others later when the deadlines done.
Also… and it’s not an excuse because it was the last two days so I could’ve done it earlier, but I was also on full dad-duty a sick kiddo who had to get picked up from nursary on the Friday, and needed a lot of daddy hugs and company on the Saturday (I probably watched maybe 8+ hours of Gabby’s Playhouse
But, I got her to bed, managed to wrangle the other non-sick older kiddo as well, and opened my blog’s repo in the terminal at around 11:15pm. 45 minutes was gonna be pretty tight…
I quickly generated a new post, rambled out a few hundred words about how this had been much more challenging than I’d expected but I’d just about manage it, and I’d come back to the other drafts later, this was all a Valuable Lesson about Planning and Scope creep.
The deadline was getting close, but it seemed I’d just about made it, but I’d forgetten I’d not thought about what I was going to use for my thumbnail or cover image, let alone actually got the pictures in the right folder, resized them and all that jazz.
I realised I’d probably not have enough time but in a moment of inspiration I was wearing my Zheleznogorsk flag shirt and I thought about that original vexiology nerd-out post and went with that as a wrap up line.
Yes, that is a bear tearing an atom in half
By the way, on this topic there was method in my madness originally, my point was going to me “When you make something you’re supposed to be doing for self-fufillment a chore, you’re gonna end up both burned out and unsatisfied”
Or even simpler “Isn’t this supposed to be fun?” hence vexiology and having a bit of whimsy going…
As I was quickly running the “git commit and merge straight to prod branch yolo mode to speed things up”,I had a flashback to a study that I’d intended to use as my central thesis of my “Staying Human in the age of AI” blog…
Despite the performance benefits, participants who collaborated with gen AI on one task and then transitioned to a different, unaided task consistently reported a decline in intrinsic motivation and an increase in boredom. Across our studies, intrinsic motivation dropped by an average of 11% and boredom increased by an average of 20%. In contrast, those who worked without AI maintained a relatively steady psychological state.
Hmm, so…you’ve got an ever increasing backlog of ideas, you’re solving them with AI and improving the core work, but you’re not scratching that loop of satisfaction when you deliver.
So you think about your backlog some more, maybe you’ll work on that other thing instead, or maybe a new idea comes in.
But… actually it’s just a oroborus of more work, less satisfaction and more watching AI agents do the
In retrospect, this was a delicious irony as this was basically one of the root causes of my 2025 blur-of-a-year and lack of blogging: the AI-assisted scope bloat loop I was in meant I was doing a lot of wheel spinning and initial spikes but never reaching the tangible, satisfying deliverable bits.
But, I didn’t have to feel so bad, as I was not alone…
The AI Vampire
A bunch of people have written about this core concept:
- Token Anxiety, or A Slot Machine by Any Other Name - jkap on how AI token usage triggers slot-machine-like compulsive dynamics
- Butlers or Architects - Murat on the tension between AI as a helpful assistant vs. AI as something that reshapes how you think
- The AI Vampire - Steve Yegge’s framing of AI as something that “sucks the mass and energy out of your projects”
Im Yegge’s piece, he literally says “You’re not getting more done. You’re getting more started.”
Why does that sound so familiar…
So, What Now?
Like a lot of people I’m trying to figure out some balance. I want to get things fun and engaging again and stop this vampire token cycle!
So, for me I think that means figuring out a new writing flow: proper planning using old school writing methods way away from distractions and LLM, at least to start with. I’m thinking about how I would write blogs in ye olde days, outlining first in plaintext, planning out structure and trying to be as compact and consise as possible. After that, maybe some basic prompting to help figure out structure and nailing down the central story I’m trying to tell. Ultimately my ideal scenario is the voice is back to being mine, with the AI back in more of a proof-reader-y role.
One term that really came up as I was thinking about this was “cognitive debt”, the idea that offloading thinking to AI doesn’t make the thinking disappear, it just defers it, and the interest compounds.
I found it via Benjamin Breen’s “What is Happening To Writing” substack post.
They talk about how they, as a writer-by-trade, were thinking about writing in the age of AI really stuck with me, especially one particular phrase:
“Writing is thinking in public”
Ultimately that’s the fun and joy of it: unpicking an idea from your brain, converting it into prose, and getting it into the brains of others. If you let AI do the unpicking, you lose the part that makes it worthwhile!
Also, I gotta figure out a way of scratching the itch that comes with the smaller ideas that I keep in my notes that I want to talk about but probably don’t work as a full-on blog post, but maybe would be cool as a lil paragraph or space to grow as I build out and feel things out. I really like the concept of a digital garden, which I found on Daniel Corin’s blog whilst trying to fix Bluesky embedding in Hugo. A garden would be a nice way to clear out the various sparks and ideas without committing to a full post!
And beyond that, I (like most internet denizens!) need to disconnect and touch grass! My reading, movie watching, powerlifting and walking have all slipped noticeably. That was actually one of the fun parts of the brief glumness I felt when I missed the deadline because of the draft flag issue, it made me take a screenbake and have some fun with the kiddos. We even cheered up my littlest with the sniffles with a sprinkle cake.
So yeah, apologies you had to read some GenAI hussle slop, and then all of this rambling afterwards, hopefully it’s not put you off ever reading this blog again!
But, all clouds have a silver lining and although I may have failed, it helped me reflect on what I’m blogging for and gave me a bunch more inspiration to do cool stuff in the future, for my own satisfaction and with my voice, not Claude et al.
So, hopefully see you in the next (scoped, fun, concise) blog!
