Firefox uses on-device downloaded-on-demand ML models for privacy-preserving translation.
-
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 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.)
-
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 "AI" isn't a real thing. When we use the word "AI", we (and you) mean something completely different from "Artificial Intelligence", basically referring to "things that we wouldn't have used machine learning for before 2018, because before 2018 we recognized it does not work for those purposes".
However, translation should still be a removable extension, for a variety of reasons, one being that the Simple Translate plugin is actually better than your builtin translation support.
-
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 Can you clarify the distinction you’re making between LLMs and open data? Was the latter collected with consent?
-
@firefoxwebdevs it would be nice if the "AI kill switch" had:
a list of each of the models used, what for, and whether they're trained on open data, each having a "disable this" switch
a thing right at the top of the list which says "I don't care, kill all this AI stuff"but that would require putting a list of all the different things that Firefox is now using AI for and whether each is using fair models or not, which I suspect a lot of management won't want to document clearly to users
@sil @firefoxwebdevs I suspect they can't, even if they wanted to.
-
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 translation feature was unnecessary to begin with. I suspect y'all know this.
-
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 shouldn't be an AI killswitch. There should be an AI enableswitch.It's like programmers forgot the default bit is 0, not 1.
-
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 personally think just a translation model doesn't need to be killed by the AI switch, but perhaps a toggle, right next to it to also disable that? That way people who don't want that either can just disable that. Adding context regarding what the translation model actually is may be a good idea as well.I'm honestly impressed that Firefox is now asking direct feedback for stuff like this.
-
@zzt @firefoxwebdevs please don't call it the "design" of the kill switch when you have to ask *us* what it should kill—as some kind of transparency/openness posturing.
@zzt @firefoxwebdevs You'd never have to say "consent", "opt in", "opt out", or "kill switch" again if you put design energy into overcoming whatever (WHATEVER) barriers are preventing all of these things being add-ons.
-
@zzt @firefoxwebdevs You'd never have to say "consent", "opt in", "opt out", or "kill switch" again if you put design energy into overcoming whatever (WHATEVER) barriers are preventing all of these things being add-ons.
@fasterandworse @firefoxwebdevs it’s bold of you to tell Mozilla to throw their entire executive board and most of their PMs into the office dumpster (the one the kitchen uses) and reform as a co-op but I agree
-
@fasterandworse @firefoxwebdevs it’s bold of you to tell Mozilla to throw their entire executive board and most of their PMs into the office dumpster (the one the kitchen uses) and reform as a co-op but I agree
@zzt @firefoxwebdevs now would be good, before that new CMO is breathing heavily over the shoulder
-
@firefoxwebdevs jonah, I hate to break it to you and the LLM shaped like a product manager that’s setting the agenda for your meetings, but the only time I hear about Firefox translations in any context is when Mozilla PMs try to hold it up as an example of an ethical, low-resource, useful AI feature so they can convince to be a fan of the worthless LLM shit they’re actually there to push
the reason why I don’t hear about translations otherwise is simple: it’s shit
@zzt @firefoxwebdevs I've used it numerous times this week and it looks good to me.
-
@zzt @firefoxwebdevs I've used it numerous times this week and it looks good to me.
@tasket @firefoxwebdevs holy shit Josh you’ve done it you’ve found the user!
quick ask them if the LLM kill switch should also turn off manifest v2 they might go for 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?
If a kill switch doesn’t kill can it still be called a kill switch?
The best solution is to divide those functionalities. Translation can be a completely separate function.
-
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 your translations are bad and you should feel bad. both about them and about this poll.
-
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 You sold out to the clankers.
-
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 How are the models not LLMs, if they are trained on large datasets and generate text?
-
@firefoxwebdevs How are the models not LLMs, if they are trained on large datasets and generate text?
@Fnordinger https://www.neuralconcept.com/post/ml-vs-llm-key-differences-applications-engineering-impact seems like a good overview
-
@tasket @firefoxwebdevs holy shit Josh you’ve done it you’ve found the user!
quick ask them if the LLM kill switch should also turn off manifest v2 they might go for it
@zzt That would be funny.
But look at the Firefox forks... some had to bring back translation after (mistakenly) disabling it. I don't think any of the local ML API should be suppressed. The discussion should be about shoving LLMs into places where they don't belong.
-
@sil @firefoxwebdevs I suspect they can't, even if they wanted to.
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.
-
@firefoxwebdevs The translation feature was unnecessary to begin with. I suspect y'all know this.
@liquor_american @firefoxwebdevs shit, I use it all of the time
