Firefox uses on-device downloaded-on-demand ML models for privacy-preserving translation.
-
@firefoxwebdevs doing a great job at regaining users' trust there, I see
In other news, you've done such a great job at regaining my trust that I've switched browsers to anything but Firefox. Well done, Mozilla.
@mxjaygrant what was it about this post that made you switch?
-
> A web browser should load web pages, allow you to interact with them ...
I would point out that translating a web page written in a non-native language allows me to interact with said page. Your argument can go both ways.
@Cappyjax Good point.
It is indeed not that simple to define what should be or what should not be a core feature. Even if for translation I am more in the "it should be add-on" team. -
@firefoxwebdevs Also as a side note: The org I'm working on has banned genAI tools for projects above a certain level of confidentiality. Guess what? Firefox is banned as well and probably stays banned regardless of any kill switch.
@sebastian which feature resulted in the ban? Given that you can access eg chatgpt in any browser, shouldn't your company ban all browsers?
-
@mage_of_dragons @cassidy @firefoxwebdevs Right, LLMs are unquestionably an AI technology, as are ML, neural nets, expert systems, and so on.
But your response misses the point. The complaint was:
Firefox users: We hate these new AI (implicitly: generative AI, LLM slop) features, please let us turn them off! (Ideally, stop wasting developer effort on them!)
Mozilla leadership: Oh, you mean you hate the AI (willfully misinterpreted to mean existing ML systems) translations?The compliant is not “It's incorrect to call LLMs AI”, the complaint is “You know perfectly well what we mean when we use "AI" in this context, stop disingenuously pretending you don't know what we're talking about”.
@RAOF @mage_of_dragons @cassidy @firefoxwebdevs the question was started clearly, yet 75% of respondents feel translation should be disabled by the switch. It doesn't seem like willful misinterpretation when the evidence is right there.
-
Signed repository metadata isn't the norm in the Red Hat family. It exists in CentOS because of community efforts (that admittedly I was involved in), and basically nowhere else.
I would like that to change, but saying that Red Hat is secretly undermining the world because of this is somewhere between laughable and insane.
Someday, we'll get there. Conspiracy theories are not required to fix it, though.
Oh, and Fedora updates are extremely transparent. All the information about them is present in the updates management system, Bodhi: https://bodhi.fedoraproject.org/
-
@firefoxwebdevs The frame of this question is risible.
I am begging you to just make a web browser.
Make it the best browser for the open web. Make it a browser that empowers individuals. Make it a browser that defends users against threats.
Do not make a search engine. Do not make a translation engine. Do not make a webpage summariser. Do not make a front-end for an LLM. Do not make a client-side LLM.
Just. Make. A. Web. Browser.
Please.
-
Signed repository metadata isn't the norm in the Red Hat family. It exists in CentOS because of community efforts (that admittedly I was involved in), and basically nowhere else.
I would like that to change, but saying that Red Hat is secretly undermining the world because of this is somewhere between laughable and insane.
Someday, we'll get there. Conspiracy theories are not required to fix it, though.
@neal The dependency of fetching metalinks from fedora.org for each update is of primary concern. Decent package management should be able to verify everything even if the central org is never contacted. Everyone should be able to see what everyone has access to – including signatures or hash lists. Making that a secret between the vendor and each client for each update is unacceptable.
but saying that Red Hat is secretly undermining the world because of this is somewhere between laughable and insane.
Excuse me, but major vendors are complying with spyware orders in a growing number of jurisdictions. Samsung, a top Android vendor, is now openly doing this in some countries. Gosh, what are Microsoft and Apple doing in China and similar states? The answer is in the news. The EU and UK have been on the edge of making this mandatory for years. So, save us the 2009 pearl-clutching about "CT".
And it's not "subscription control", the TLS certificate is used to authenticate you to the Red Hat CDN and get you access to the download location.
You contradict yourself.
As for Red Hat's motivations, Red Hat has declared they will not honor the GPL any longer when distributing patched code (i.e. their modifications) to customers. This is despite the GPL being focused on redistribution of modified code. They've gone to war with it and that is a simple fact. The takeover and dissolution of stable CentOS should have been a lesson learned; they weren't doing it out of any kind of "community spirit".
Have a sane day...
-
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 alternative perspective:
Remove all AI-LLM, AI-ML related functionality.
Then have target end-user (web developer) choose, informed by their values & preferences what functional components they’d like to “plug-in” to web-browser for ML content processing for web page-
- Language translation - enable on device locally download-on-demand ML or use your own
- Dictionaries
- …
Once these are real-world validated & functional, they can be shared via open source commons with others. -
@sebastian which feature resulted in the ban? Given that you can access eg chatgpt in any browser, shouldn't your company ban all browsers?
@jaffathecake ChatGPT (and many other web based things) are firewalled.
Also you are looking at a compliance issue from a technical viewpoint. As the implications of genAI generated content wrt. copyright and things like patent applications are still somewhat unclear in many jurisdictions, the simplest solution is to stay well clear of any tool that claims to do anything "AI".
If the contract with the customer says "no AI because it exposes us to legal risks", then the work has to be done in a clean environment where there is nothing that could be considered AI. -
@ShadSterling @chillicampari @firefoxwebdevs @joepie91 translation is already opt-in. You're prompted about it, and the model is only downloaded if you say you want it.
@ShadSterling @chillicampari @firefoxwebdevs @joepie91
You're *constantly* prompted about it on every single site you visit. Calling that opt-in stretches the definition of consent.
-
@Rycochet @firefoxwebdevs @zzt I did not follow all what happened around Firefox and the community. Did Mozilla made a public consultation regarding AI integration in Firefox ?
Do we have some reliable datas about the opinion of the Firefox's users ?I would be interested to know if the critical views (that I mostly share) expressed here are largely shared or not.
@fmasy @Rycochet @firefoxwebdevs @zzt You can look at the discussions on Mozilla Connect if you want commentary from community members.
Mozilla does occasionally run surveys, but results are never public.
-
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 I would rather like for auxiliary features to be added via the extensions API.
-
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 There's nothing intrinsically wrong with AI. If you can do translation on device in a privacy-preserving way, there's no reason to disable it.
-
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 I'm not a fan of this all or nothing approach to functionality. What happened to the paradigm of things being scope and/or time-limited, eg just for this page, for x minutes?
-
@firefoxwebdevs I'm not a fan of this all or nothing approach to functionality. What happened to the paradigm of things being scope and/or time-limited, eg just for this page, for x minutes?
@scribe it isn't all or nothing (the currently winning option is granular control). This is the first time I've heard a request for AI to be disabled in a time-limited way. Tell me more about your use-case.
-
@fmasy @Rycochet @firefoxwebdevs @zzt You can look at the discussions on Mozilla Connect if you want commentary from community members.
Mozilla does occasionally run surveys, but results are never public.
-
@scribe it isn't all or nothing (the currently winning option is granular control). This is the first time I've heard a request for AI to be disabled in a time-limited way. Tell me more about your use-case.
@firefoxwebdevs I think it's less about use cases and more about general trust, as privacy often boils down to. If you're talking about an AI "kill switch", you're talking about trust in what's been defined as "AI", and trust in the browser developer as a whole.
Once definitions are murky, there's an area open for ongoing redefinition. One way to adopt a "private by default" approach is to follow what cookies do, for instance, and allow users to allow limits to the extents of permissions.
-
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 there's a huge difference between calling out to an external llm service and using the translation-specific on-device models.
(If, hypothetically, llm's like chatgpt were not a thing, would people have such visceral reactions against the translatelocally models?)
-
@firefoxwebdevs @yoasif @fmasy @Rycochet @zzt a self-selecting survey with push-poll questions that deliberately leave out the "no LLMs in Firefox" option is unlikely to be statistically valid
(also we know this is just noise and Mozilla will do whatever was planned in the meeting anyway)
-
@firefoxwebdevs @yoasif @fmasy @Rycochet @zzt a self-selecting survey with push-poll questions that deliberately leave out the "no LLMs in Firefox" option is unlikely to be statistically valid
(also we know this is just noise and Mozilla will do whatever was planned in the meeting anyway)
@davidgerard @yoasif @fmasy @Rycochet @zzt I realise your position is immutable, but I've already used the results of this survey to push for a change to the design of the kill switch. I'm grateful to everyone who responded.
