Hello, World

2026-02-12T00:00:00Z

Welcome to the blog. I've been meaning to set this up for a while, and it's finally here. This first post covers what's under the hood, because the infrastructure is more interesting than a generic introduction.

How it's built

The site runs on 11ty (Eleventy), a static site generator. I write posts in Markdown, push to a GitHub repo, and Netlify builds and deploys automatically. No CMS, no database, no JavaScript frameworks. Just static HTML and CSS served from a CDN.

The blog supports tags for filtering by topic, reading time estimates, and an RSS feed if that's your thing.

11ty was the right choice here because it does one thing well: turn Markdown files into HTML pages. I didn't want to maintain a build system more complicated than the content it produces. The entire config is about 60 lines of JavaScript.

Quality gates

One thing I wanted from the start was a way to enforce content standards automatically. Not just "does it build," but "is this actually worth publishing." So before any post reaches the live site, it runs through a validation pipeline.

A Node.js script checks every post against a set of targets: word count, internal and external link counts, summary length, heading structure, tag hygiene, anchor text quality, paragraph length, and a few other things. Each check gets a pass, warn, or fail. Any failure blocks the deploy.

This runs in two places. GitHub Actions runs the validator on every push, so I can see failures in the commit checks before I even think about merging. Then Netlify runs the same validator as part of the build command. If a post has failures, the build exits before 11ty even starts. Nothing gets deployed.

There's an override ([skip-validate] in the commit message) for cases where I've reviewed the output and decided a warning is acceptable. But the default is: if it fails, it doesn't ship.

The workflow

Posts start on a development branch. I push, the validator runs, I see what needs fixing. Once everything passes (or the warnings are intentional), I merge to main and Netlify deploys. The whole loop is usually a few minutes.

I also have link checking set up through lychee, which scans every page on the built site for dead links. That runs on every push and on a weekly schedule so links that rot over time get caught too.

What to expect

I plan to write about technology, tools, projects, and whatever else I find interesting enough to put into words. Networking, self-hosting, security, Linux, hardware, AI, reviews, project updates. No schedule. Posts go up when they're ready.

The site will keep evolving alongside the content. There's a list of improvements I'm working through: syntax highlighting for code blocks, per-post sharing images, pagination, and eventually a Substack integration for email subscribers.

If you're here, thanks for reading.

The Bubble and the Science

2026-03-02T00:00:00Z

Look, I'm just a guy. But I think that artificial general intelligence (AGI) isn't really a discrete threshold anyway. There's a gradient between the auto-completing of sentences, and actual reasoning. While mechanistically, LLMs ARE "stochastic parrots," on a purely representational level they could end up being something more profound.

We're used to models that are forced to take deterministic paths through the vector-space, but an LLM that uses the superposition of those representations would end up being able to link data in ways that produce novel reasoning. It still wouldn't necessarily be anything even close to consciousness, but we'd likely find it incredibly impressive (as a massive understatement).

The rough math

The following has caveats, but it's the rough math I just did.

The human brain: 100 trillion connections across 86 billion neurons, or around 1,163 connections per neuron.

Current LLMs (already, classically, on normal hardware): according to research from Anthropic, the most conservative estimates about this you could possibly make would be about 5-10x the feature compression, or 1.6 quadrillion interaction terms per forward pass. That's already 16x the brain's synapse count just from attention geometry, and it'd be the absolute floor.

To put that differently: every time you send a prompt to one of these models, the number of representational interactions happening in a single pass is over an order of magnitude larger than every synapse in your head firing at once. And that's the conservative number. The real number depends on how much information is being packed into each dimension, and the research suggests it's a lot more than we assumed.

The caveats

I said there were caveats, and there sure are. The brain runs its 100T connections continuously and in parallel, has persistence, and constant feedback loops, and probably a bunch of other stuff that will always make you an infinitely more wondrous machine (that you should be proud of). A forward pass happens once and is done. Your brain never stops. That's a fundamental difference that raw numbers can't capture.

But to be honest, as much as I think there's an AI bubble, and that a lot of you are talking absolute nonsense, there IS a chance that things improve despite the fact that the entirety of human knowledge on the internet and elsewhere has virtually been consumed. The superposition research suggests that models are compressing far more information into their parameters than we thought, which means there may be room to get more out of existing architectures before we hit the wall everyone keeps predicting.

Feel free to have dreams or nightmares about it as you please.

TL;DR: The LLMs are a bubble, but there's still some science left to do.

It's been a while since my first post. I'm still here, still writing. More to come.

Jay Neawedde