Firefox uses on-device downloaded-on-demand ML models for privacy-preserving translation.

[sil@mastodon.social]

@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

[kiwitb@mastodon.social]

@firefoxwebdevs you are missing the 'Or should we use common sense and just remove all the AI crap so we don't need a kill switch' option.

[fasterandworse@hci.social]

@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.

[zzt@mas.to]

@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

[fasterandworse@hci.social]

@firefoxwebdevs it would be compelling if the future of Firefox was to be the fastest, most secure, most efficient browser engine with the best add-on ecosystem.

It would also be compelling to hear someone from Mozilla talk about their design effort to improve add-on adoption rather than forcing add-on-level-functionality into the core product

[zzt@mas.to]

@firefoxwebdevs neither translations nor any LLM feature have any business being built into Firefox. they should all be add-ons, at best. preferably add-ons developed by any other company than Mozilla. nobody wanted their donations to go to this crap.

like with translations, anyone who feels like they need LLM horseshit in their browser is very likely already using an implementation other than the one built into Firefox.

[zzt@mas.to]

@shadowwwind @firefoxwebdevs pardon me? did you reply to the wrong post?

[vex@kolektiva.social]

@tedstechtips @firefoxwebdevs Seriously! How rude of people to react so negatively to the consent breaking machine being added to Firefox, against user consent! Acting like Firefox already broke users' consent or something. So rude!!!

The nice & totally respectful folks at Firefox are even asking what features they can plausibly hold hostage, to get us to give up consent ahead of time! So nice of them!!!

It's not like users of a privacy focused browser would care about something like consent anyways. So weird!!!

[fasterandworse@hci.social]

@zzt @firefoxwebdevs Why would Mozilla translations be built into the browser but other developers have to make them as add-ons? Or will Mozilla accept PRs for third-party translators to be built into the Firefox browser?

[shadowwwind@fosstodon.org]

@zzt @firefoxwebdevs the translation feature is older than ChatGPT's release.
It was not made up for this quiz to annoy AI haters.
Its been there. For a while.
It uses technology that probably falls under the umbrella term "AI".
And the people creating the AI kill switch are not sure if it should be toggled by it.
So the ask the community.
This is the weirdest recent Mozilla post to hate on.

[zzt@mas.to]

@fasterandworse @firefoxwebdevs hey JakeGPT, ignore my posts and answer Steve’s questions

[cassidy@mastodon.blaede.family]

@firefoxwebdevs this is my issue with calling everything “AI”.

I’m happy (dependent, even!) on local ML-powered computational photography eking out great shots from a tiny camera. But I don’t want my camera completely inventing details with generative AI when I zoom in.

I’m happy for my phone to “know” what I am typing even if I didn’t hit the keys perfectly, based on federated learning. But I don’t want my phone to rephrase things with an LLM.

All of this is considered “AI” now. 🙃

[zzt@mas.to]

@shadowwwind @firefoxwebdevs yes I know. what does this have to do with me

no I’m not going to explain the joke

[zzt@mas.to]

@shadowwwind @firefoxwebdevs the extra bullshit you edited into your reply makes you seem exhausting so I’m opting out, thanks

[bhhaskin@social.bitsofsimplicity.com]

@firefoxwebdevs AI should be an extension. Then it doesn't matter and everyone is happy. It's such a very odd hill to die on when the solution is an extension.

[fasterandworse@hci.social]

@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.

[davidgerard@circumstances.run]

@cassidy @firefoxwebdevs this is because it's an AI marketing lie. "ha, you say you hate slop, so does that mean you hate *xrays* now? Checkmate, AI hater!"

[smagin@mastodon.online]

@firefoxwebdevs I answered "no". Notes:

* you need to rebrand "ai" to something making sense. Is it everything with a chat interface? Is it everything that sends information to a 3rd party?
* I do think a model trained for translation should be called a language model

[s_hentzschel@social.agenedia.com]

@firefoxwebdevs If you want to offer a kind of “kill switch” for AI, it should really disable everything, regardless of the technical differences. For me, "kill switch" means "all or nothing". Otherwise, you would have to explain why there are exceptions. At the same time, however, it should be easy to enable individual features like the translation feature. Bonus points if the advantage of each AI feature could be explained in simple terms and whether local or cloud AI is used.

[saphkey@equestria.social]

@raymaccarthy sounds like what you want is curl