nadiyar
60 following, 141 followers
Firefox uses on-device downloaded-on-demand ML models for privacy-preserving translation.
They're not LLMs. They're trained on open data.
Should translation be disabled if the AI 'kill switch' is active?
| Yes: | 755 |
| Yes, but let me re-enable just translations: | 1456 |
| No: | 661 |
| 🤷: | 104 |
Let's ask the real question:
Firefox users,
do you want any AI directly built into Firefox, or separated out into extensions?
@firefoxwebdevs
@davidgerard
@tante
| I want AI built into Firefox: | 2 |
| I want AI separated into extensions: | 6 |
| Mozilla should not focus on AI features at all: | 24 |
@duke_of_germany @firefoxwebdevs @davidgerard @tante
We were longtime users if Firefox.
AI is crap.
Nobody wants AI.
All of us are Librewolf users now.
@Compassionatecrab @duke_of_germany @davidgerard @tante fwiw, Librewolf includes the same AI translation engine as Firefox.
@firefoxwebdevs @Compassionatecrab @duke_of_germany @davidgerard @tante
Firefoxwebdevs, would you please stop muddying the waters by conflating machine translation with generative AI? You know they're not the same, you pointed it out in your poll.
@cryptica @firefoxwebdevs @Compassionatecrab @duke_of_germany @davidgerard @tante I'm happy you felt it was clear from the text of the poll, yet look at the results
@jaffathecake @firefoxwebdevs @Compassionatecrab @duke_of_germany @davidgerard @tante
The poll you made specifically to muddy the water?
@cryptica @firefoxwebdevs @Compassionatecrab @duke_of_germany @davidgerard @tante you said I clearly made the case for them being different, and yet respondents disagreed.
Not sure how I could have made the point clearly, yet been misleading.
@jaffathecake @firefoxwebdevs @Compassionatecrab @duke_of_germany @davidgerard @tante
🤷
By making a flawed poll based on falsely tying all machine learning to GenAI/LLMs.
Do you seriously think this poll shows that only 1% want translation features? Of course not, you know better.
@cryptica @firefoxwebdevs @Compassionatecrab @duke_of_germany @davidgerard @tante see this, and the following post in the thread https://mastodon.social/@firefoxwebdevs/115859958902197087
Thank you everyone who responded to this.
For context: I saw mocks of the kill switch where translation was included, but it lacked the ability to enable the kill switch but still enable particular features (such as translation).
The results of this poll helped me successfully push for more granular control in addition to the single AI kill switch. So again, thank you for that.
@jaffathecake @firefoxwebdevs @Compassionatecrab @duke_of_germany @davidgerard @tante
Thank you for pushing for, and getting, more controls for the end user.
I know you are not directly involved in that part of the organization.
Please stop conflating all machine learning with generative AI/LLMs.
@cryptica @firefoxwebdevs @Compassionatecrab @duke_of_germany @davidgerard @tante I agree they're different. I feel I laid out the difference clearly in the poll, and from your messages, it seems like you agree.
And yet, the vast majority considers that to be the type of thing that should be covered by the kill switch.
So, if the kill switch didn't include translation, folks would have accused us of cheating/dishonesty. I'm glad we avoided that.
@jaffathecake @firefoxwebdevs @Compassionatecrab @duke_of_germany @davidgerard @tante
I still think the poll results are skewed based on the construction of the response options tying translation features that people do want to the genAI features that no one wants.
Hence why I'm asked for you (and whoever was responsible for the mockup in the first place) to stop tying them together.
@cryptica @firefoxwebdevs @Compassionatecrab @duke_of_germany @davidgerard @tante I really agree with you in terms of translation, but look at some of the replies to the poll posts. I think if Mozilla tried to say "no, this isn't the type of AI you hate, it's different", the response would be furious.
@jaffathecake @firefoxwebdevs @Compassionatecrab @duke_of_germany @davidgerard @tante
I just reread an entire third of the responses.
I saw two who wanted translations to count as AI for the kill switch.
I did see scores of folks, however, pointing out the things I pointed out, namely that your poll is flawed, you are intentionally muddying the "ai" waters, you are intentionally choosing to not understand that you cannot point to the results of a flawed poll as an indicator of anything in reality, you are cherrypicking comments, and you are inventing a fictional future response based on your biased flawed poll.
Tell your bosses that there is zero trust and goodwill left for anything Mozilla, much less for hot garbage takes meant to boiling-frog/nazi-bar us into accepting the inevitable creep of techbro number-go-up enshittification that is blatantly being forced on us even as we speak.
That is why no one trusts your "AI kill switch," Mozilla. You already proved you can't be trusted.
@firefoxwebdevs Where's the option for "I do not want this bullshit toy anywhere near my browser"? Is someone forcing you at gunpoint to be pro-slop? Why are all the executives so into this crap? Can't we just let them have their cocaine daydreams without subject the rest of us to it?
@StarkRG @firefoxwebdevs @davidgerard Firefox only exists because Google subsidize them so they can point to FF as "proof" that Chrome isn't a monopoly. With the new regime in power, that's a dead issue. So Google want FF to push AI adoption now because they've figured out how to monetize it and they don't want precious eyeballs evading their slopware. If Google cut off their "search" payment to FF, Mozilla goes bust and the C-suite lose their jobs. QED.
@cstross @firefoxwebdevs @davidgerard
Sure, I just wish they'd stop lying about it and pretending like it's anything anyone actually wants.
@cstross @StarkRG @firefoxwebdevs @davidgerard Yep. Their paying customer is Google, not the people using Firefox.
@bjn @StarkRG @firefoxwebdevs @davidgerard Wasn't it Mark Zuckerberg of whom it was said, in the early days of Facebook, "if you're not paying for the product you ARE the product"?
@cstross @bjn @StarkRG @firefoxwebdevs @davidgerard it's older than that. The phrase dates from "free" broadcast TV at least.
@cstross @StarkRG @firefoxwebdevs @davidgerard My first encounter with the idea was reading Chomsky in the 1980s, in relation to media. The idea is likely way older.
@cstross Ufrisken ligg i at vevaugo som Firefox, Chrome, Brave og Vivaldi omfamner virknadsflod, og dette gjer at dei er bunden av ein godsleg overflod for å ilivehalda seg.
virknadsflod er til gagn for makten 👑
Om me røyndleg har lyst på eit vevauge som ikkje opptrer i maktens namn, så er me naudd for å nøya oss med mindre.
Vevauget Lagrange som er bunden av Gemini framgangsmåten er ein lækjing i dette spørsmålet.
@cstross @StarkRG @firefoxwebdevs @davidgerard ...has google figured out how to monetize it? I still only see nonconsensual unasked for ai results in my searches. I'm definitely not paying for any of that
@autonomousapps @StarkRG @firefoxwebdevs @davidgerard They have studies showing that if they show the AI results first, a significant proportion of google users stay on google (to explore the AI results) rather than following outbound links to the public web. Thereby giving google more opportunities to shove google's own ads under the users noses.
It really *is* that moronic.
@cstross @StarkRG @firefoxwebdevs @davidgerard oh geeze. I'm clearly in the minority. Thanks for explaining!
@firefoxwebdevs My closest answer would be "no", but I think the question is kind of mis-phrased here, and that's probably going to lead to a confusing and potentially misleading outcome.
The problem that people have is not with "AI" as a generalized category, but with the current generation of thieving, climate-destroying, grifting systems that are marketed as AI to an overwhelming degree - notably LLMs and "generative AI", but really anything with those inconsiderate properties.
If your kill switch is presented as an "AI kill switch", then depending on the person they're either going to understand that as "exploitative tech", or as "machine learning", and so make different assumptions as to whether local translation is included in that.
So I think you'll have to be a lot more explicit about what you mean; either by describing clearly what the kill-switch includes, or what it excludes, right in the place where the option is offered. Otherwise it's damned if you do, damned if you don't; depending on whether you include translations, either one or another group is going to be upset with the unexpected behaviour.
So, ethically, if the translation feature is built on ethically collected data, and it has no outsized climate impact, then I would not consider it something that needs to be included in a "get rid of all of it" kill switch. But to convey this clearly to users, both that and why it isn't included should be explained right there with the button, with potentially a second-step option to disable it anyway if someone still feels uncomfortable with it.
That way you've transparently communicated to users and shown that you have nothing up your sleeve by immediately and proactively offering them an option to disable that, too, if they have already shown interest in removing "AI" features.
@firefoxwebdevs Here's a concrete example of what I mean, that should be pretty consistent with the Firefox UI design:
@joepie91 I think a lot of people in the replies would consider this sneaky. It's a tricky UX problem. But yes, granular control needs to be part of the solution, along with a kill switch.
@firefoxwebdevs I can only speak for myself of course, but I'm someone who is strongly opposed to sneaky approaches, like hiding things in submenus or requiring people to go back later to disable new things, for example. And I'm also strongly opposed to basically everything in the current generation of "AI" (LLMs, GenAI, etc.) - but personally I wouldn't consider this sneaky, as it's immediately visible that there's a second choice to make, at the exact moment you disable "AI".
Of course if that stops being the case and the second option gets hidden behind an "Advanced..." button or foldout for example, it would be sneaky. But in the way it's shown in my mockup, I would consider it fine as it's both proactively presented and immediately actionable.
(I do still think that exploitative "AI" things should be opt-in rather than opt-out, but it doesn't seem like that's within the scope of options that will be considered by Mozilla, so I'm reasoning within the assumption of an opt-out mechanism here)
@joepie91 they will be opt-in, but different people have different opinions about what that means. For us, it means models won't be downloaded or data sent to models without the user's request.
However, some folks have said the only meaningful opt-in would be a separate binary for the browser-with-AI, or even having to compiling it manually.
@firefoxwebdevs "Without the user's request" is quite ambiguous, though. I'm reminded here of Google, which put the AI tab before the Web/All tab, displacing it so that people would unintentionally hit the AI button and "request" it. It's a small and plausibly-deniable change that nevertheless violates the user's boundaries, and difficult to call out and stop even internally within a company or team. I've seen many companies and software do the same thing.
A genuine opt-in would, in my opinion, look something like a single "hey do you want such-and-such features? these are the implications" question, presented in a non-misleading way, and if that is not answered affirmatively then the various UI elements for "AI" features should not even appear in the UI unless the user goes and changes this setting. It's much harder for that to get modified in questionable ways down the line, and reduces the 'opportunities for misclick' to a single one instead of "every time someone wants to click a button". It also means users aren't constantly pestered with whatever that week's new "AI" thing is if they've shown no interest.
Such a dialog could still specify something like "if you choose Yes, Firefox will still only download models once you try to use a feature", to make it clear to users that it's not an all-or-nothing, and they can still pick-and-choose after selecting 'Yes'.
@joepie91 @firefoxwebdevs Mozilla's tortured definition of opt-in seems to predict that Mozilla will invent features to nag you into enabling AI, as they have already done with Link Previews: https://www.quippd.com/writing/2026/01/06/architecting-consent-for-ai-deceptive-patterns-in-firefox-link-previews.html
Had to convince my mom to scroll down from Google's AI answers to find more reliable results -_-
Me at school: ok so I suck at this subj I gotta revise a lot at home and in my free periods At home: -_- im too lazy can I just chill
I saw "LLM's are unregulated automated yes-men" and holy shit yeah ☠️
How many naps did you have today?
| One: | 35 |
| Two: | 16 |
| Three: | 5 |
| I'm napping right now: | 22 |
Closed
@vlrny @Paperposts Where is the answer for zero?
@vlrny
Just because I'm a spoonie doesn't mean I have to take naps every day. You really should have given a 0 option.
"[Sam Altman] reckons progress is so fast that he could soon be able to make an AI to replace him as a boss."
🤷♀️
Doesn't sound like the biggest challenge ever.
🤔
Who's up for replacing him with a perl script?
Sven Slootweg, low-spoons mode ("still kinky and horny anyway") » 🌐
@joepie91@fedi.slightly.tech
Especially when working on a project, I sometimes don't have the spoons to write alt text for screenshots; there can be quite a lot to describe, and it often requires a complex description.
What would you prefer that I do in those situations, and do you depend on alt text to understand images?
| I do NOT depend on alt text; post without alt text: | 9 |
| I do NOT depend on alt text; do not post it at all: | 8 |
| I do depend on alt text; post without alt text: | 0 |
| I do depend on alt text; do not post it at all: | 0 |
@joepie91@fedi.slightly.tech My instance allows you to use AI to generate alt text, which is kinda cool so when I'm lazy / the image has lots of details I can let it do the hard work and I can fine-tune it afterward. Maybe you can do something like that? Or maybe you can tag your posts to ask people to write some alt text for you. Iirc # alt4you is one of them
In Norway, they’re building a new stave church, using only Viking-age tools and techniques. 1,000 people are involved, and it is expected to be complete by 2030.
https://www.visitnorway.com/typically-norwegian/how-to-build-a-stave-church/
Would you like to see UNIX V4 on MissPiggy? #retrocomputing #vintagecomputing #unix #softwarepreservation
@icm I'd like to see Xenix on MissPiggy.
The new timeline implementation is now available for testing on Google Play 👀👀👀
🧵👇
@ConnyDuck ok but why google play. why r new features exclusive to the corpo silo
kind of a bad look imo
I have been working on this almost every evening since months and duckfooded it for weeks. It is now ready for some more public testing. Please let me know any and all issues you encounter.
Some of the improvements in the new timeline include:
Improved html rendering: Html rendering is now way better than before. Things that did not display correctly, e.g. links in strikethrough tags, nested lists, ordered lists now display correctly.
More than 4 media previews: Tusky now indicates when a post has more than 4 attachments - before you had to click through into the media viewer to get that info
Timeline position remembered reliably: When "oldest first" readin mode is active, Tusky will always remember the home timeline position, even when completely closing the app.
Full screen scrollable in tablet mode: On tablets, the whole screen area including the margins is now scrollable. Before only the posts themselves were scrollable.
Huge shoutouts to @Tak and @gryzor for reviewing my code and @bladecoder for his html parsing library that is the foundation for the new timeline implementation.
I'll now start adding support for quote posts, which *should* be a breeze because the new code is way more maintainable than the old one.
And then it is time for a Tusky 2⁵ release 🥳
@ConnyDuck 🤩 that sounds great.
Did you change anything about the way the timeline is being formated or is the new implementation matching the current look & feel?
@mxk it should mostly be the same, except how overlong posts are collapsed
@ConnyDuck since i saw this thread i went as fast as possible to install Tusky Test and checking it 👀
I'm afraid the latest Tusky test build keeps spontaneously closing for me - either immediately after I open it or as soon as I scroll. Android 14 on a tablet, portrait orientation.
Have force-closed and cleared cache, but no difference.
When I scroll down, it always shuts down at exactly the same point. Can post a screenshot of my feed just before the point where it shuts down, if that's any use.
@ratcatcher sounds like you have a post with some unexpected properties that causes the new implementation to crash. Can you check with a regular Tusky build or another app which one it could be? Should be 1-3 posts after the point where it crashes
It's odd because I have three accounts set up in Tusky. All three follow the same accounts so the feeds are pretty much identical.
Viewing via my mastodon.ie account, Tusky crashes consistently. But scrolling the same range of posts via my c.im account, it doesn't.
I have screenshots from both (which I've sent to you).
@ConnyDuck I'm using the oldest-first order with full timestamps and see no more timeline skips or false ordering in tusky nightly.
Excellent work. Thank you very much.
A dream: scrolling bottom-to-top to constantly read top down.
If so, I'm available between 5:30 - 21:30 UTC 7 days of the week
Have you seen a programming language with explicit environment passing?
@lxsameer
Compulsory or optional?
@kupac both, if you know any let me know.
@lxsameer
In R, you can call a function (which is a closure), and redefine its environment. https://www.rdocumentation.org/packages/base/versions/3.6.2/topics/do.call
The guy is an engineer, and to support his claim, he pasted two links to Instagram videos as a proof. 🤯
@nadiyar 😂 Nah, I know him. He is not. He just likes to “read” scientific papers on Instagram 😂.
And yes, I'm being sarcastic.
ugh I want to discover new music and yet I am also too lazy to discover new music and I like the comfort of songs I already know but I'm also bored of them
why is inaction sooo easy
inertia
@cybar.dev@bsky.brid.gy a @writeblankspace at rest will remain at rest until a force is applied to it frfr
the existence of GNU+Linux implies the existence of GNU-Linux
Why isn't there an Android app to easily change the preferred network type?
Ie, what if i wanna be stingray secure in the city but if im in the woods without 5G.... You know???
Also why does Android call LTE "5G"??
Have to use the
*#*#4636#*#*
Dial combo to change it now :(
But yeah it still flags lte as 5g sometimes lol
@nadiyar nope!
That doesn't not disable ie lte :D
The "preferred network type" in the dial menu forces ie; lte or NR only!
It's been around 6 years that I've started my work and research on Serene's #language and #compiler. One of the big mental challenges that I had, is that I keep rewriting the compiler occasionally due to a new turning point in my understanding. It bugged me a lot and every so often I feel lost. Yesterday, I had a chance to talk to a few devs of Clasp lang. I felt much better when I learned they had the same experience and rewrote the compiler many times. So it's all natural to feel lost.
One reason you should care about problems that don’t affect you personally is because, actually, they do.
I hereby declare that the term "smart" in combination with a device such as a TV or phone is short for "surveillance machines are really tedious". Now you know.
@liw I stand corrected, I thought it's short for
snooping mechanical antagonist.
rotten technology.
(sma/rt)
Twenty five years ago this month I was diagnosed with cancer. First of all I just want to say, look at me, still here. I'm pretty happy about that.
But twenty five years ago, for a year my and my family's life revolved around surgery, chemo and radiation. And then the best part of a decade of medications with with their 'tolerated' side effects. And still the sneaky little blighter tried to stage a comeback a few years in, but we evicted it.
Twenty five years is a while, so my medical experiences are not current but there are a few things I learnt in that year and the ones after and I'd like to say them out loud -
Cancer did not make me a better person. It did not make me re-evaluate my life and see the world in a new perspective. It might for some people, but it just made me tired and grumpy. And intolerant of people's bullshit.
It was not a blessing in disguise. Having people tell me there had to be a silver lining or that everything happens for a reason was not helpful.
Everyone has their own approach. Mine was evidenced based western medicine. Random strangers (no friends, I'm glad to say), telling me I was poisoning myself was not helpful. I was poisoning *it*, the side effects on me were necessary collateral damage
The most important thing my friends did for me was to be normal. Doing stuff we usually did - grabbing a coffee, going out for a meal, seeing a movie. I was thinking about cancer every waking minute, respite from that was what I needed
The other day someone reminded me I told them this - One of the women in my support group talked about how hard it was not to cry in front of her kids. The facilitator asked, what message are you sending to your kids if, when something this bad happens, it's not okay to cry? This is one of the most important things I've ever learnt. Hiding your feeling doesn't let people in and it's not a viable strategy for longer than half a second. 1/2
I find it mildly annoying how in media (books, movies, etc) when the girl is indifferent towards the notion of romancing guys and voices how she doesn't need guys to be fulfilled, she ends up falling in love with a guy anyway. I mean that's fine and I suppose some people have gone through that but, is there like a lesson being taught here? "If you don't believe in falling in love with a guy it's just cause you haven't found the right one yet"? What if the girl really actually doesn't care about guys? Like, imagine a Princess Jasmine-esque intro and then it turns out the girl actually really just isn't into guys and discovers that she's either asexual or a lesbian. It'd make a cool story. Or if she's neither asexual nor lesbian, just like... not give her a romantic interest. Just let that fact that she's never had a crush just hang in the air and exist. Let her beeeeee. Stop getting me hyped over a female character being confused about why people are confused that she doesn't have a crush on anyone.
Actually, perhaps such fiction actually does exist and they're called coming out stories. But like, what about a side story. There are lots of stories where the main plot is adventure or mystery and some random romance happens without the story revolving around it. If we can have romance side stories we should also have anti-romance side stories right. Girl goes.on adventure. Girl gets lost. Girl hoes on quest. Girl meets boy. Girl and boy become friends. Girl and boy use the power of friendship. Girl saves the world woth the help of boy. Girl and boy celebrate. Girl and boy keep in touch as close friends. Yaaay.
I actually really love found family kind of stories. Trials of Apollo was such a good read. Lester (16 y/o mortal, 4000 y/o god) meets Meg (12 y/o) and they both grew up in horrible families. They come across each other and develop a sibling-like bond. Lester reminisces on his many exes and crushes. He offers to be someone's boyfriend and gets rejected. He encounters several exes amd fights to the death with several. And in the end he doesn't get in a romantic relationship. The story isn't about his romance, but with his found family with Meg and co. I loved that.
p.s. except MS Windows, it keeps getting more unusable every month
Oh my god well done Quanta magazine, amazing article.
https://www.quantamagazine.org/why-is-ice-slippery-a-new-hypothesis-slides-into-the-chat-20251208/
🙌