Firefox uses on-device downloaded-on-demand ML models for privacy-preserving translation.
Uncategorized
220
Posts
107
Posters
1.1k
Views
-
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?
@firefoxwebdevs As worded, and if we can trust Mozilla, then the acceptable answer should be No for these reasons: ML is not AI, and on-device means nothing is sent out of the device. In exchange you get free translation. Win.
BUT… there’s the trust issue now.
And what we REALLY need is not an AI kill switch but more of a “data transfer/phone-home kill switch”, almost like a firewall, where we know the browser is not taking any data and sending it to a device we don’t control ourselves.
-
@made @firefoxwebdevs There's already lots of work for on-device ML: https://searchfox.org/firefox-main/search?q=toolkit%2Fcomponents%2Fml
Integrating models into a finalized product with the wide spectrum of end-user devices is tricky though, so it has to be done with care.
@gregtatum @firefoxwebdevs great to hear! I can imagine! Thanks for the link ☺️
-
@firefoxwebdevs As worded, and if we can trust Mozilla, then the acceptable answer should be No for these reasons: ML is not AI, and on-device means nothing is sent out of the device. In exchange you get free translation. Win.
BUT… there’s the trust issue now.
And what we REALLY need is not an AI kill switch but more of a “data transfer/phone-home kill switch”, almost like a firewall, where we know the browser is not taking any data and sending it to a device we don’t control ourselves.
@mdavis folks want to disable 'AI' for more reasons than privacy. Privacy is important of course, but folks are also concerned about the training data, and energy used for the training.
-
@firefoxwebdevs not trying to split hairs here but how are the ML models doing translation when they are not LLMs? Maybe they are not as huge as ChatGPT but they are transformers probably with all that entails.
(A Killswitch should of course kill all ML/AI functionality and people could then reactivate certain specific features of they want to, it's really not that hard. Just cause you consider a feature"better" than others does not override consent practices.)
@tante it’s a SLM :)
-
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?
@firefoxwebdevs tbh, the open embracement of AI, the addition of AI into the browser, while full well knowing your user base is well known for being anti big tech and privacy focused, was a mask-off moment.
I've already switched to librewolf, and I didn't have to disable/remove bullshit.
I recommend your ELT 1) get a grip and 2) remember you exist because of your userbase, not to please tech giants. If big tech had their way, they'd eat you alive. people who want AI slop aren't using Firefox.
-
@firefoxwebdevs It would also be compelling if a team at Mozilla were dedicated to building the best browser translation add-on on the market, for all browsers. To promote the power of add-ons and, at the same time, the Mozilla brand.
@fasterandworse there are no such interfaces to intercept input boxes with extensions I guess. And also why should Firefox improve other browsers?
-
@tante it’s a SLM :)
@eckes for that usage pattern the results would probably be even worse with more fabrications. So what are we even doing here?
-
@zzt I posted this poll after a meeting where we discussed the design of the kill switch, and there was uncertainty around translations. I want to make sure the community's voice is represented in these discussions.
@firefoxwebdevs @zzt How about making a poll "Should Firefox include AI/LLM by default?"
-
@firefoxwebdevs That's exactly the motivation behind my suggestion, though - I've attached a mockup in an additional reply to hopefully make it clearer, but the idea here is to not redefine it so much as it is to explicitly pick a definition, and then provide an additional option for the broader definition, so that a user can essentially pick whichever definition they are following without getting into the technical weeds too much.
@joepie91 agreed.
@firefoxwebdevs we're not in those meetings so we don't know what all is actually included within the AI module suite, or even if that has been fully defined internally at this point, so of course there won't be a clean consensus externally from us on what "it" is and if it should be included or excluded, as it's up to our interpretation.
-
@mdavis folks want to disable 'AI' for more reasons than privacy. Privacy is important of course, but folks are also concerned about the training data, and energy used for the training.
@firefoxwebdevs But if the ML/AI training work is processing on the device and not is shared off device, and it is in support of a feature like translating a page (which should be prompted/selectable) then what’s the issue? You can say no and nothing happens. Or you can say yes and the worse that happens is you chew up some local power on your laptop or PC. Or are you saying that even though the translation happens on the device, the RESULT of that training data is sent back out?
-
@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)
-
@firefoxwebdevs But if the ML/AI training work is processing on the device and not is shared off device, and it is in support of a feature like translating a page (which should be prompted/selectable) then what’s the issue? You can say no and nothing happens. Or you can say yes and the worse that happens is you chew up some local power on your laptop or PC. Or are you saying that even though the translation happens on the device, the RESULT of that training data is sent back out?
@mdavis I believe it's a moral stance due to how the models were produced.
-
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?
@firefoxwebdevs Like many others, I have a metric ton of thoughts on this topic. I might even try writing something to consolidate them.
In the meantime: I don't consider the translation models to be part of the major AI hype I loathe so much. Machine translation of language has been happening for a long time and has proven largely useful, and it lacks the stink of desperation which so many of the generative applications of recent times carry.
While I'm already thinking about it: even the name "AI kill switch" feels bad to think about. I know that "AI" is the buzzword that gets upper management giddy and which the untrained public is now used to hearing, but the fact of the matter is that if you can't "sell" a feature without appealing to buzzwords, your feature wasn't worth the time and effort put into it.
-
@joepie91 agreed.
@firefoxwebdevs we're not in those meetings so we don't know what all is actually included within the AI module suite, or even if that has been fully defined internally at this point, so of course there won't be a clean consensus externally from us on what "it" is and if it should be included or excluded, as it's up to our interpretation.
@chillicampari @joepie91 fwiw I asked about translation because we're figuring out what to do specifically about translation.
-
@firefoxwebdevs Like many others, I have a metric ton of thoughts on this topic. I might even try writing something to consolidate them.
In the meantime: I don't consider the translation models to be part of the major AI hype I loathe so much. Machine translation of language has been happening for a long time and has proven largely useful, and it lacks the stink of desperation which so many of the generative applications of recent times carry.
While I'm already thinking about it: even the name "AI kill switch" feels bad to think about. I know that "AI" is the buzzword that gets upper management giddy and which the untrained public is now used to hearing, but the fact of the matter is that if you can't "sell" a feature without appealing to buzzwords, your feature wasn't worth the time and effort put into it.
@bersl2 I agree it's a meaningless buzzword, but a lot of tech folks are saying they want "no AI" - they're using the buzzword. So the poll is about finding out what folks mean by "no AI".
-
@eckes for that usage pattern the results would probably be even worse with more fabrications. So what are we even doing here?
@tante hu? I guess a slm is much better suited as the old ispell dictionary, I don’t see an issue with offering that (as an option)
-
@mdavis I believe it's a moral stance due to how the models were produced.
@firefoxwebdevs Hookay… then this is less about a local feature or data sharing and more about an overall “Made with AI” concern where nothing related to AI *at*all*ever* taints the user’s browser, in or out. In that case, if the user turns on the AI kill switch, it should totally kill anything having to do with AI for those who take that position.
That’s an issue with these polls — too much undisclosed nuance to be able to answer properly.
-
@angelfeast @twifkak No, I don't think so. It says this (with a takedown compliance process posted afterward)...
License
These data are released under this licensing scheme: PD
We do not own any of the text from which these data has been extracted.
We license the actual packaging of these parallel data under the Creative Commons CC0 license ("no rights reserved").@tasket @angelfeast https://paracrawl.eu/moredata says "This is a release of text from Internet Archive.... The project also used CommonCrawl which is already public." Those crawls quite famously/infamously include copyrighted content. I don't see anything to suggest they filtered those datasets for public domain annotations. (Not that such an annotation would be enforceable, but it would at least be an indication of intent.)
-
@firefoxwebdevs Hookay… then this is less about a local feature or data sharing and more about an overall “Made with AI” concern where nothing related to AI *at*all*ever* taints the user’s browser, in or out. In that case, if the user turns on the AI kill switch, it should totally kill anything having to do with AI for those who take that position.
That’s an issue with these polls — too much undisclosed nuance to be able to answer properly.
@firefoxwebdevs But wait… what if the developers used AI to help develop the code in the browser itself? Does that mean AI kill switch purists should then rather not even use the product at all?
-
@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.
Gli ultimi otto messaggi ricevuti dalla Federazione
-
Hmmm.... I'll just put him in Andorra, since that's right smack in the middle
-
Se esistesse una comunità di #T1, ben vengano i confronti.
Foto puramente dimostrativa, clickbait e nu*ity.
-
@thedarktangent This is interesting.
Every time a defcon.social user tries to interact with feddit.it, this error appears:2026-02-26T09:23:02.075679Z WARN lemmy_server::root_span_builder: CouldntFindObject: Failed to parse object https://defcon.social/@signorina37/116135473370481795 with content {"error":"Request not signed"}: data did not match any variant of untagged enum SearchableKinds
by @signorina37
or
2026-02-26T07:16:22.978630Z WARN lemmy_server::root_span_builder: Unknown: Failed to parse object https://defcon.social/users/grabbi_it/statuses/116133671937824462 with content {"error":"Request not signed"}: data did not match any variant of untagged enum PageOrNote
by @grabbi_it
We are currently receiving errors from the following instances, but they don't have much in common: they are running different versions or even forks.
https://bsd.network
https://cathode.church
https://defcon.social
https://kolektiva.social
https://mastodon.art
https://social.lol
https://universeodon.com
https://wehavecookies.social@informapirata@defcon.social @support @TCMBC @informapirata@mastodon.uno
-
@eniko I voted for “other” because although I have never personally roasted a whole chicken, we do have “girarrosto” shops here that basically sell just that
-
@NIGHTEN the right one (second)
-
@GossiTheDog @switchingsoftware Ridiculous.
“Sandy, we have to let you go because you aren’t using Copilot enough.”
“But… I already use it for literally everything I do here.”
“Well, then I guess you should have made up stuff to get it to do for you. Security is waiting to take you to your desk. You have fifteen minutes to collect your things.”
“But—“
“If you’ll excuse me, Sandy, I need to ask Copilot to tell me how to word a call to my mother for her birthday before I fire the next person.”
-
Have you looked at your sidekiq logs to see if there are any errors?
Thanks for the feedback; Lemmy doesn't have sidekiq, but I'll try to see what the logs say anyway.
@informapirata@defcon.social @support @TCMBC @informapirata@mastodon.uno
