Firefox uses on-device downloaded-on-demand ML models for privacy-preserving translation.
-
@mdavis @firefoxwebdevs I do not want to use any product that has been developed using "AI" code generation tools, especially not if it is security critical software like a browser
@mcc @firefoxwebdevs I would mostly agree with this if you added this at the end of your statement: …by an idiot programmer or one who didn’t grow up and learn to code properly during the decades before AI LLMs.
In reality, I don’t think either of us are going to get our way on this one.
-
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?
I would say every feature in everything should be a separate toggle to the best of its ability.
Also, by "open data", I hope you mean "the data's license gives consent to be used in this way", not "the data exists on the web somewhere".
-
@firefoxwebdevs Said translation should be an opt-in extension you can install if you want it. Not a core component at all.
@dalias @firefoxwebdevs Which is also kind of funny when compared to pro-privacy features like containers being put as extensions. -
@Fnordinger https://www.neuralconcept.com/post/ml-vs-llm-key-differences-applications-engineering-impact seems like a good overview
@jaffathecake This article claims that LLMs are always transformers. This is not true, in fact the first LLMs were LSTMs (https://arxiv.org/abs/2405.04517).
-
@firefoxwebdevs Can you clarify the distinction you’re making between LLMs and open data? Was the latter collected with consent?
@knowler @firefoxwebdevs it absolutely was not! he means "open data" as in "we found it lying around, bugger the license" https://mas.to/@twifkak/115849848003348176
-
@firefoxwebdevs what exactly do you refer to as „open data”?
-
@liquor_american @firefoxwebdevs shit, I use it all of the time
@wes @firefoxwebdevs @liquor_american so do I, but also it should be an extension
-
@mcc @firefoxwebdevs I would mostly agree with this if you added this at the end of your statement: …by an idiot programmer or one who didn’t grow up and learn to code properly during the decades before AI LLMs.
In reality, I don’t think either of us are going to get our way on this one.
@mdavis @firefoxwebdevs I don't like the word "idiot". But a programmer who would use LLM codegen is a programmer with bad judgement. A programmer who has bad judgement cannot spot the errors made by LLM codegen. QED.
Anyway I already got what I wanted: Servo, the web browser which will replace Firefox, has *already* banned "AI" code contributions. So it's only a matter of time before Servo is complete enough for day to day use, and I can delete the AI-infected Firefox from my computer.
-
@flxtr @firefoxwebdevs as someone who used these in the early 2000s: no, it's not. It's not as good as DeepL, but it's worlds ahead of machine translation in the 2000s.
@funkylab @flxtr @firefoxwebdevs there's no such thing as good machine translation
-
@jaffathecake @Fnordinger I do! my source is this thread and the thread linked in the OP: https://wandering.shop/@xgranade/115772870672213549 category IV is the most relevant one but you’ll want to read the entire thing
@zzt @jaffathecake This is an interesting read, but I struggle to find a proper demarcation criterion that
(1)separates LLMs and other types of „AI“, while
(2)allowing for a translator to be part of the last group.I think we might have reached a point where not just „AI“, but also „LLM“ is starting to lose its (already underconstrained)meaning.
-
I think the challenge with everything going on here is one of clarity.
@sil, you are asking them about disclosure of models and sourcing. But that is far from the only AI that is in the system.
The tool that does grammar checking and language identification does not leverage an LLM, and while there may be some type of model underneath, the context is very different. Tools that detect spam pages or faulty JavaScript that locks the pages, that's another type of AI hard at work.
Is the browser allowed to support speech to text?
@jmax You're calling out that Firefox may not be able to do this, but I think that mischaracterizes the scope of what's happening here.
The browser has several types of non-deterministic, probabilistic tools in it that provide useful services. Now there's a backlash against one very specific version of those non-deterministic, probabilistic tools. But the backlash is vociferous, often unsolvable, and incredibly broad.
It's hard to engage with non-specific anger.
@gatesvp @firefoxwebdevs @sil @jmax this is the sort of obfuscatory claim I see from AI marketers. "You say you hate slop, so that means you must hate X-ray scanning! Checkmate, AI hater!" It's not convincing.
-
@eckes @fasterandworse To further the charitable mission, pretty obviously.
-
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 the AI kill switch could be better rebranded as the GenAI kill switch (I get that marketing has muddied the water here).
ML models like translations are very much not the same thing as the LLMs being pushed at the moment and should be treated separately.
And I truly appreciate the work out into getting it performant and accurate enough for my needs.
For me, ensuring that the browser isn't leaking what I'm reading is an important privacy control, and I wouldn't trust a 3rd party plugin or an online translator service (or Google translate in Chrome).
-
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?
donate to servo if you can
https://opencollective.com/servo
they have a roadmap that is dedicated to making an actual browser engine, not a collection of browser features on top of one
-
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 The problem is "AI" is a meaningless marketing term, and if you have to pose a question using it, then you aren't being specific enough about what you're asking. More generally, if you find you have to pose an ethical question at all to the public, it's a big signal that you shouldn't be doing what you're doing.
-
@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.
> I want to make sure the community's voice is represented in these discussions.
if that were true, the poll would have had a "remove all LLM functionality" option.
-
@firefoxwebdevs going through all the other replies and your lack of response to any of them..
“why are there flaming bags of poop on my porch, and why do they all have different postmarks”
-
@funkylab @flxtr @firefoxwebdevs there's no such thing as good machine translation
@typhon @flxtr @firefoxwebdevs oh I'm sure not being able to read what the Japanese seismographic agency's website said about the underwater quake is much better than having had a machine translation of that page for my friend in Alaska. </sarcasm>
-
@eckes @fasterandworse To further the charitable mission, pretty obviously.
q1 - design one
q2 - see the post you responded to -
@twifkak Also notice that Mastodon instances are using LibreTranslate.
Has that been debated as well?
![[object Object]](https://media.mas.to/accounts/avatars/109/287/758/133/964/366/original/0f3099c0adfb36a0.png "[object Object]"){kind=avatar}
