I think we’ll unfortunately see a degradation over time of capable programmers, as it’s with most other things: Humans learn poorly from seeing things being done for them, rather than doing it themselves. AI can be a good resource for iteration and a sort of rubber duck, and even for explaining things (with external confirmation). But as it currently stands, people find it much more attractive to just let AI do the work for them. That’s the most prominent work culture after all, based on capitalistic values: Economic growth above all.

I see it in schools as well, unfortunately: Students find it much more appealing to get good grades than to actually learn. Not everyone, of course, but still way too many.

And I do understand many of the students: The bar for junior programmers has now increased, as a lot of employers simply think “why hire a human that is on the same level as AI?”.

I can vouch for that: As a programmer, I can generally see both programming AI and general LLM use if it was copy-pasted. I know a guy who isn’t a programmer who’ve used Claude Code to make an app, and things are working, but it’s sooo clear even to a junior SE that it’s vibe-coded. There’s a clear lack of consistency, architecture, maintainability, and general human readability.
However, if the code/text was reviewed and edited by a capable human programmer, it’s incredibly difficult to single out what is a product of AI and not.

Actual programmers who understand and can write code themselves who use AI simply as a tool, NOT a shortcut, will produce indistinguishable code. Linus is an incredibly smart and capable guy, and for specific uses, I’m confident in him using this tool responsibly.

(The ethics of AI use is a different topic however…)

Interesting! Does anyone know of any C/C++ libraries that implement this? And perhaps some solid performance numbers runnings the different solutions?

This all makes me think of the awesome performance enhancement videos for Super Mario 64 by Kaze Emanuar!

There are so many open source solutions to it at this point, yet nobody has thought of simply aggregating all of the various solutions into one with a simple GUI/CLI basically asking "Do you want to send it directly (P2P E2EE), indirectly (through agnostic server to hide IP), upload it (3rd part account/server needed), or make a share (announce your hash and IP on a discover network)?

How does this compare to wush?

I’m curious, but have many questions. It’s a fork of LibreWolf, but what made a fork necessary, apart from using ESR instead of RR?

The FAQ describes in some ways how it’s different from LibreWolf, in addition to pulling from other projects like the Tor Browser, but I’d love to see some near-exhaustive spreadsheet or table of features and functionalities that differ from the various projects.

What is the motivation for users to switch away from LibreWolf? Who is the team behind Konform, and how can one trust this team to build a “better” (depending on what you value) browser? And furthermore provide a consistent and safe development distinct from other browsers?

I share your interest, and crossposted to !linux@sh.itjust.works and !linux@sopuli.xyz.

Personally I find Navi better to use for the purpose of “not having to remember commands”. It also provides a description of the command, fuzzy search, and essentially pastes the command in the current terminal, not having to think about current directory or user.

Just is essentially a collection of aliases, while Navi is a collection of pre-written commands in full length (with optional description).

This is something I’d love to have for the past decade or so, but as I’m now transitioning to Linux I hope to not have much of this problem anymore.

Nice of you to create this though, and I share your tiredness of settings being changed behind our backs…

This is super cool! I’d love having something like this when I was a kid! I’ll see if I can find some kids I know to take a look at this, and I’m confident they’ll have a blast!

Say it with me now, people!: “GPTs are not mindful!”

LLMs (everything out there called “AI” these days) are predicting words based on statistical analysis of human-made texts. There is nothing “smart” about LLMs, they are only perceived like that because the training data contains a fair share of intelligent works from humans!

So, it doesn’t have “feelings” or “motivations”. It inherits the “traits” from human works that relate to the context given. If I prompt something like “Tell me a short story of one paragraph backwards in leetspeak”, it’s going to generate some sentences with these words being the context. It will sift through its model, prioritizing works it has trained on that has the highest relevance/occurance/probability of these words, mix it up with a bit of entropy, and voilá! You’ll (likely) get a fair amount of gibberish, because this is a very small-to-non-existent part of its model, as very few actually use reverse leetspeak, i.e. it makes very few appearances in the training data. (I haven’t managed to get any model to write this flawlessly yet.)

Watch this if you want a bite-sized explanation of how a GPT works: https://youtu.be/LPZh9BOjkQs

They updated the article to provide further context:

Later, in June 2025, the mod’s original uploader deleted it from Steam entirely. In other words, Valve did not remove the mod from Steam (as our original article suggested).

So Valve only restricted access in South Korea, not the rest of the world.

For such mods, there is pretty much always a very firm emphasis on “this is a work of fiction” when starting the game/mod. Example is the Assassin’s Creed games, where they don’t always have historic accuracy, but that’s okay because they have emphasized that this is a liberty they have taken, and they have already made the user aware of it and agreed to the terms.

This mod however (without having tried it myself) does not make any emphasis on this, and hence uncritically portrays it as part of real events.

Does Actiona suit your needs?

Yeah, a bit of an xy problem, I’ll admit. My ultimate goal would be invisible blocking/hiding from certain expected behaviors like the ones anti-adblockers employ. I’m not sure if what I have in mind specifically exists or not, but I find it fun to tweak and experiment with these kind of things. So I don’t necessarily have a problem I’d like to solve, it’s more an ask for directions/experiences.

I have tried Fiddler before, and it almost has the capabilities I’m looking for (it replaces an entire file; nothing granular), but it was a bit of a hassle to get working well last time I tried it, with the CA certificate, decrypting the TLS and stuff.

I haven’t heard of ZAP before, so this is new and intriguing! Thanks!

uBlock Origin I very much like, and I think it’s near-impossible for me to use a browser without it installed anymore. But as far as I know, it’s only used for blocking (dropping network requests) and hiding (CSS-like rules), but no manipulation?

The monkeys I’ve also used a little bit of, but I wasn’t aware that they supported being run ahead of the rest of the page like how I wanted, so this is definitely worth looking into, I appreciate it!
For reference, I found this answer on StackOverflow, which is promising.

I most likely am going a bit to the obscure/convoluted solution, yes.

My goal is to do something to the effect of uBlock Origin, but instead of just blocking/hiding, either replace with local files, or intercept req/res in order to manipulate them favorably, without being detected. I don’t know what uBlock does under the hood though, apart from its resource blocking and CSS-derived hiding.

Example:
Watching a video on youtube, an ad is about to get loaded, but instead of the hiding/blocking strategy uBlock uses, intercept the GET/POST, save the important flags that are uniquely served to your device that would indicate that you have successfully been served the ad, drop the rest, and then answer with what would be a valid response for “I have watched the ad in its entirety”.
So the server basically saying “Here, I give you this page and this script with both vital and ad contents. I now expect you to provide the corresponding hash that these two files will create through a series of functions. If you don’t, I will assume you’re blocking me, and I won’t provide further contents.”, and I’ll simply respond with “Here’s your hash! *wink*”.

Essentially, I wish to experiment with trying to be completely invisible in the blocking, by providing responses as if I have loaded and watched the ad, with all anti-adblock implementations through scripts and dynamic loading “intact” and unaware.

You’re quite right, I should include my goal in the post - editing it in now.

Do you often (if at all) run into conflicts with the .. alias? I can’t think of any case on top of my head, but it feels a bit sketchy. The more than 2 dots however I imagine is pretty safe.

I’d say Linux Mint, ZorinOS, and Manjaro Linux are all viable options for Windows users who want an easy transition. Although I don’t think any distro will ever be considered a “plug-and-play” experience. There are varying degrees of user-friendliness, but if one wants user-friendliness like not having to do root/sudo actions even once, I think one might be better off with MacOS…? Though from what I’ve heard, the main reason Windows users are looking towards Linux and not MacOS is exactly because of the ability to customize more than just the wallpaper (and also the entire boycott US movement).