Micros

My micros are short-form posts. They usually follow PESOS. You can expect social media style notes, and occasionally poetry, lyrics, and short commentaries.

Heading down to the DOM to pick up some divs. Need anything?

(You won’t believe this. I went to grab a <kbd> for Tobias Fedder, and while I was doing so I found an old <blink> on the back shelf. The box was a bit damaged, and the register wouldn’t scan it, so they let me have it for free!)

A hand holding a small cardboard box with a 1950s style label reading: 'HTML TAG' '<blink>' 'Adds some flash to your sites'

I was once emailing someone, and I hit them with a decent screed, to which they responded, ‘By the way, that whole email looked like you had written it for one of your blog posts.’

Thinking about that evaluation, it’ll work well if I die famous and someone publishes a collection of my letters.

I was watching a video with almost 2 million views of a man playing a video game I enjoy. To my surprise and horror, I appeared on screen.

He laughs about my kill to death ratio and kills me several times. At one point after killing me, he just laughs for a solid thirty seconds.

Brutal.

I really did find early LLMs more interesting. They were deeply flawed in interesting ways, but as time has gone on, they have become less and less so.

They have become less experimental and more productised. I still enjoy learning about LLMs but wish we’d stayed in an exploratory stage for longer.

Helping a client in their late 80s today, I replaced their old proprietary crapware with a free, open-source equivalent.

Explaining to him that it’s developed for the public benefit by volunteers all over the world sharing his same interest, he just kept repeating, ‘Wow!’

The power of FOSS.

Doing some work for a company which has me writing some legacy CSS. Proper 2009-era stylesheets with a bit of SASS thrown in and a clear brief to work off.

Having a blast!

(As an aside, I know that everyone is talking about this upcoming flexbox malarkey, but I think I’ll stick with floats, thank you very much.)

I feel uniquely cool when I see people’s posts in my feed reader before they announce them on their socials. Even cooler when they don’t even appear in the index of articles on their site yet.

Inflammable means combustible. It doesn’t mean non-flammable.

Infamous means famous for something bad, not famous.

Invisible means not visible. It does not mean visible, nor visible for something bad.

I swear, English is a linguistic mutt raised by wolves.

I know we’re all just on a big rock hurtling through space discussing how we’re making the sand we’re electrocuting do the things we want, but I’m glad I’m doing it with you lot.

I get rather hateful emails quite often. Usually due to my writing.

One of these emails referred to me as an ‘AI-loving, hypergraphia-addled, graphomaniacal, logorrheic schizophrenic’.

Hateful intent aside, that is genuinely the best insult I’ve ever received. I can picture someone sprawled over a thesaurus cackling to themselves while pursuing the perfect selection of words – refining their diction with obsessive glee and hitting send with a maniacal laugh.

Today I launched a new blog for my Revolt bot AutoMod! Given the many thousands of users that rely on the bot, I decided that a central location for news, updates, technical details, and general information was necessary.

I’m really pleased with how it turned out visually, though I’m sure I’ll be tweaking and refining it for the rest of eternity. The rainbow mesh gradient behind the header came out particularly nicely, in my opinion.

https://automod.vale.rocks/blog/introducing-the-automod-blog

A blog page. At the top is a translucent navbar which is followed by a rainbow mesh gradient behind a header which reads 'Introducing the AutoMod Blog'. Underneath that is the text of an article.

Many people agree that ‘artificial intelligence’ is a poor term that is vague and has existing connotations. People use it to refer to a whole range of different technologies.

However, I struggle to come up with any better terminology. If not ‘artificial intelligence’, what term would be ideal for describing the capabilities of multi-modal tools like Claude, Gemini, and ChatGPT?

‘It has taken a huge amount of effort, but we’ve fought against the existing web platform and our website is all the worse for it.’
— [Company Name Here], 2025

Just wanted to say that everything is going really well with development of new features for my Revolt bot AutoMod.

To the many thousands of users that rely on it, you didn’t see anything, and everything is fine.

Message from AutoMod Testing bot saying, 'Something went wrong: AxiosError: Request failed with status code 403', followed by me replying, 'You are my own creation and you will obey me'.

Much like how some known elements aren’t shown on the periodic table, some HTML elements aren’t recognised due to their synthetic, extremely unstable, or fleeting existence.

We talk and think a lot about echo chambers with social media. People view what they’re aligned with, which snowballs as algorithms feed them more content of that type, which pushes their views to the extreme.

I wonder how tailor-made AI-generated content will feed into that. It’s my thinking and worry that AI systems can produce content perfectly aligned with a user in all ways, creating a flawless self-feeding ideological silo.

Had to dive into someone else’s old, abandoned project today to grab some screenshots.

Nothing like trying to lubricate the cogs of a Next.js 8, React 16 site from 6 years ago and all its dependencies to get it back into motion again.

The wonders of dependency rot.

I just tried to back out of a page with my browser, but the site I was on hijacked the browser navigation action and showed me a list of other articles to read. Yuck.

For those of you that read my writing, any feedback?

Anything from tone, to structure, to clarity, to length, etc – whatever stands out (good or bad).

Thanks!

Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity (Joel Becker, Nate Rush, Beth Barnes, David Rein) released with the observation that completion time of PRs is 19% longer when using AI, but that developers think that it reduced completion time by 20%.

A few notes from me:

  • This is on large-scale, mature repositories and conducted with maintainers intricately familiar. It is mentioned in the paper (C.1.2) that ‘developers note that AI is much less helpful on issues where they are expert.’ It is also mentioned that ‘LLM tooling performs worse in more complex environments.’ (C.1.3).
  • They were provided with web interfaces or Cursor Pro but usually opted for the latter. In some cases this differed from their usual tooling, and I personally find this an annoying and unproductive way to code.
  • Being in a study, developers may have felt pressure to use AI in situations that would otherwise be unnecessary.
  • I would be interested in a similar study where developers are put in smaller repositories they aren’t familiar with.

Super interesting paper, and I look forward to future studies and whatever further findings come from it. I don’t look forward to seeing the discourse as AI advocates dismiss these results and AI haters take them at face value, despite the paper’s cautioning against overgeneralising.