My spam filter on my email server. Autocorrect on Heliboard. I think that’s it. Luckily I picked up English at a young age so I don’t need it for translation.
Also for people having a kneejerk reaction, “AI” and “ML” as terms long predate the current slopwave. The ML I am talking about in the previous paragraph is not genAI, it is the kind of ML to classify things (emails into spam/ham, words into correct/incorrect), that you can feed training data to improve its accuracy rate.
El Pueblo Unido Jamas Sera Vencido
I don’t think the main users of these services are Twitter users as you can’t log in to Twitter using the alternative frontends, nor can you follow accounts to create a timeline afaik. They’re just for viewing tweets eg if someone has sent you a Twitter link and you want to see what it is, or for conducting open-source research on what a person has said or put out historically. At least that’s how I use them.
yovday
Looks nice though, and it’s cool how well Linux works for touchscreens these days.
Ubuntu was my first Linux distro and I liked it back then. I would’ve recommended the old Ubuntu to non-tech-savvy people, or people who want a Linux that “just works” (although back then no Linux “just worked”, there was always at least a bit of debugging for me). However it’s enshittified a lot. A lot of basic software is only available as snaps. It generally feels quite corporate these days.
Seconding Niri, but also I’ve seen swayfx setups that have Hyprland-like aesthetics.
If you don’t care about fancy effects, I’ve been a river user for years and I like it. Nowadays it is just a compositor and the WM is separate, but river-classic (which includes a WM) has a dwm-like tag system, is programmatically configured (ie you speak to the river IPC in any language of your choosing—by default it’s a shell script but the WM just executes ~/.config/river/init on startup, so that file can be any executable), and is simple and lightweight. The new 0.4.x release separates the compositor and WM, so you can pick or write your own WM that suits your own workflow/use-case, if that’s your thing—it might not be if you don’t want customisability and you just want something that makes the decisions for you to a good enough standard.
Some people do live off the record. It’s somewhat inconvenient but far from impossible. A lot of people I know have their address off the record such that the government won’t have it on record, let alone a private company.
Yes, I said “no they don’t” because translation has the ethical implications of global inequality. Which, btw, climate change greatly exacerbates. I never said that LLMs don’t have environmental impacts if used for translation; I said they have additional ethical implications when used for machine translation. Which is true. I have my doubts that you are particularly concerned with tech ethics and global inequality given how flippant you have been about the real issue of exclusion of global majority populations just from things as simple as language choices.
A small fraction? I’m talking about the majority of people on Earth who are disadvantaged by English centrism.
And where did I say “it’s ok to burn the world”? I never specified any position on technology, but to be clear, I am a luddite, I don’t think modern tech is ecologically justifiable, and is built upon a foundation of slavery and ecological destruction. It is a separate question of a policy of a small software forge (which already runs on ecologically destructive hardware; the vast majority of emissions for modern tech come from the manufacture of the devices we’re using to talk now, from your own client device to the devices being used as servers to relay our messages) to allow or disallow AI translations. I do not think that our energy use is justifiable. Now LLMs use a comparable amount of energy to, say, AAA video games. I think this is a bad thing and video games should be more computationally efficient. Does that mean I support banning people from playing AAA video games? No, I don’t think that’s the way to address this problem. Questions of policy are separate questions to questions of technology; a technology can be bad without the solution being to ban it, because the effects of a ban are different to just getting rid of the banned technology.
And LLM translations have all the same political and ethical as coding
No, they don’t. As I said, machine translations make contributions more accessible to ESOL speakers (ie global majority populations) and can help reduce Eurocentrism. The reason why I said I think an exception should be made is specifically for ethical and political reasons.
If you want to skip to the changes.
I think this is a really solid LLM policy for a platform like SourceHut. Since they rely on a policy that’s deliberately open to interpretation to allow for reasonable discretion, and rely upon the honour system for enforcement/adherence, I don’t think we should necessarily be telling larger projects to adopt the same policy because this becomes untenable for big projects or platforms. But it’s still good to see policies in this spirit being adopted. The only change I’d like to see is explicitly allowing AI translation for non-code, like comments and emails. It reduces the barrier of entry for ESOL contributors. If there’s concern about the accuracy of translations, they can always include the original alongside the machine translation.
It’s FOSS has a tendency to write in that kind of cadence. In general the articles they put out don’t seem to be ones that are tedious or boring to write, so I don’t see why they’d AI-generate their articles. I also remember their old articles before the AI hype train being written in the same kind of cadence. It is grating but I think it’s an annoying human, not an annoying LLM.
I don’t know about Nazi specifically, but people often use Nazi to mean fascist in general. And he is definitely far-right. I would prefer that foss remain hostile to fascists.
On at all times at the router level. Because it’d be annoying to selectively disable it. My VPN isn’t really an anonymity feature anyway; I have a dedicated browser profile for things like banking where I intend the fingerprint to be associated with my government identity.
Converting a GPL codebase to another language creates a derivative work which, by the terms of the GPL, must be GPL-licensed.
Non-native but spent a lot of my life living in English-speaking contexts. I know how to pronounce it, but my tongue seems to like tripping up over “hospital”.
0 days, who cares. Call yourself from a place if you want to.
Julia Evans has a bunch of these really handy cheatsheets. I have several saved that I reference semi-regularly.
I don’t particularly care in the abstract but this particular piece of software is also technically awful vibecoded slop. And DHH isn’t a random nazi, but a nazi with a platform provided by the popularity of his slopware.