It is important to let people be wrong, especially when the stakes are not high (Hint: in most work, most of the time the stakes are not high). Being wrong is one of the best ways to learn to be right

Also, it’s best not to try to convince someone they are wrong, but to highlight the risks associated with their position, and try to get them to work with you on finding a way to mitigate the risks. Half the time they will realize they’re wrong, and half the time you will find you weren’t as right as you thought.

https://wangcong.org/2026-06-30-why-i-stopped-arguing-with-people.html

Back to pen and paper: Norway bans AI from primary schools.

For decades, Norway has been a pioneer in digital education. But now the Scandinavian country is making a U-turn.

The Norwegian government is reacting to declining test results among students and is issuing a far-reaching ban on the use of generative artificial intelligence (AI) in teaching.

https://mediafaro.org/article/20260620-back-to-pen-and-paper-norway-bans-ai-from-primary-schools?mf_channel=mastodon&action=forward

#Norway #AI #PrimarySchools #Skills #Education #Tech #Teaching

The funniest thing about the Softbank goose memes is that their chucklefuck metaphors for things learned by GCSE economics students in their first 15 minutes of class are patently absurd in their own right.

"Eggs do not lay eggs" yes they do, that is how reproduction and compound investment - the thing you are presumably trying to explain - literally works.

I dunno, guys. I think the world might just be run by morons with more money than sense. I don't think there's much more to it than that, I really don't.

@troublewithwords

This quote has been circulating recently, from a 1985 paper that was part of a conversation at the time about what a computer program *is*. Is it the source code? Or the final executable binary? Or the sum of its versions? Or the people who wrote it?

I've been programming for over 40 years, and I can tell you with great confidence that Naur was right. When I'm building software, first and foremost what I'm building is me.

I do think that make (the tool) is by far and away the best thing that came out of the whole UNIX era/ethos. I'm always fascinated by how versatile it is.

My text editor has integrated terminal windows. Nowadays, opening a new terminal window means immediately opening my editor. Why not skip the extra step and just have it *be* the shell?

Why can't a text editor also be a shell? Why have we decided that shells always take the form of a blinking line parser at the bottom of the screen?

I'm quite proud of the fact that at $day_job I've spent more time deleting than adding code over the past year or so. I'm an anti-software absolutist: code is bad and dangerous by default and only good and useful on rare occasions.

RE: https://neuromatch.social/@jonny/116764269580776228

Lines of code is a terrible proxy for software productivity, but there is something it's actually quite a good proxy for: complexity. The bad kind, in particular.

Building a Soviet Nail Factory: how KPIs killed efficiency. A story that I often tell. It is now old enough to not upset my former colleagues. https://vincent.bernat.ch/en/blog/2026-kpi-goodhart

A quantization/dithering and outline shader for Veloren. Quite happy with it.

The UK government's plan to teach 10 million British children how to use VPNs may be one of the most ambitious IT education projects ever launched. Experts have praised the scheme, saying that a deft combination of incentives and peer education make it more likely to succeed than other, comparable initiatives.

"With the rise of autocratic governments worldwide, VPN-literacy is more essential than ever.” said one expert, “This bold project definitely comes at the right time.”

#UKSocialMediaBan

https://shawnsmucker.substack.com/p/please-use-ai

Title: "Please Use AI"

Excerpt:
Be sure to use AI when making
your next, I don’t know, meal plan,
for example. Definitely do not call
your friend who loves to cook and ask her
for her favorite recipes or tips or ways
to save time making meals,
because you will end
up talking for longer than you had hoped,
hearing, perhaps, about her father’s cancer
diagnosis or how lonely she’s been or even
what she’s planted in her spring
garden and then lost with the early frost.

Recommend you read the whole thing.

I've started asking people the thought experiment "are there more doors or wheels on planet earth?". I don't really care about their answer, but the speed at which they ask ChatGPT allows me to assess how much I care about their thoughts and opinions on other matters.

Almost 25 years ago, I wrote a blog post with the title ‘jumping ship slowly’ about leaving Windows (XP was awful, it was mind boggling to me that Vista managed to make people nostalgic for XP). My advice remains the same:

Don’t try switching OS first. The OS is the most easily replaceable bit in the stack. Switch applications first. Most ‘Linux’ apps are cross platform. They’ll run on Windows, and the few that don’t will run in WSL2. You can switch out apps one at a time, and take the time to get comfortable with the alternatives.

Once you’re comfortable not using any Windows-only apps, changing the OS but using all of the same applications is very easy to do. Changing OS and application stack at the same time is an enormous obstacle.

I believe this is also why a lot of corporate and government Linux migrations fail: they try to change everything at the same time and that’s too steep a learning curve.

It's really difficult to put hard numbers on it but I think this has made me substantially more productive as a developer, especially when working with unfamiliar code. I no longer have to pull the whole codebase into my mental cache to start working with it: I can ask questions of the code with the project search at the speed of thought. It feels *amazing*.

I don't care how fancy you think your AI-powered workflow is, there's nothing like wondering "which components consume this type?" and getting the answer back before you've even finished formulating the question.

Another thing I added recently is post-search filtering. Once I search for a term, I can then keep typing additional space-separated filters to refine the search.

This has had a surprising effect on the way I write code. I now mostly don't think about directories when I'm moving around the codebase. Instead, it's much more like I have a tagged filesystem: I can specify little fragments of the path in whatever order I like based on whatever pops into my head first and it zeroes in on the result almost immediately.

Since I made project-wide search fast in my text editor I've realised I've mostly stopped using the regular find/open dialogues.

The project-wide search nudges files that are close to the current file to the top of the results anyway, so I'm seeing the same files regardless: but I also get to see everything in nearby directories too, just in case I forget where the file I'm looking for is.

Originally I built project-search as a heavy-weight tool with the expectation that I'd use it only a few times per programming session but now that it can sift through millions of lines faster than I can blink I've ended up using it to query pretty much anything about the code I'm working on.

The mode in which your phone can receive calls and notifications is called “Disturb.”