The latest news and developments on Firefox and Mozilla, a global non-profit that strives to promote openness, innovation and opportunity on the web.
You can subscribe to this community from any Kbin or Lemmy instance:
Related
While we are not an official Mozilla community, we have adopted the Mozilla Community Participation Guidelines as far as it can be applied to a bin.
Always be civil and respectful
Don't be toxic, hostile, or a troll, especially towards Mozilla employees. This includes gratuitous use of profanity.
Don't be a bigot
No form of bigotry will be tolerated.
Don't post security compromising suggestions
If you do, include an obvious and clear warning.
Don't post conspiracy theories
Especially ones about nefarious intentions or funding. If you're concerned: Ask. Please don’t fuel conspiracy thinking here. Don’t try to spread FUD, especially against reliable privacy-enhancing software. Extraordinary claims require extraordinary evidence. Show credible sources.
Don't accuse others of shilling
Send honest concerns to the moderators and/or admins, and we will investigate.
Do not remove your help posts after they receive replies
Half the point of asking questions in a public sub is so that everyone can benefit from the answers—which is impossible if you go deleting everything behind yourself once you've gotten yours.
I was wondering why there were so many people tagging others in this thread, it is not normally what I see on Lemmy, so I had a look at the instances people were posting from, as well as where this thread was posted.
It was quite impressive.
Oop created this thread using Mastodon, the thread is located on an mbin instance, other mastodon users posted in it, and here I am reading it and posting in it from Lemmy.
Really cool that it works this well!
Adding for anybody that wants to disable it all:
user_pref("browser.ml.enable", false); // general switch for machine learning features in Firefox, though it might only apply to some of them, https://bugzilla.mozilla.org/show_bug.cgi?id=1971973#c11
user_pref("browser.ml.chat.enabled", false); // AI Chatbot (https://docs.openwebui.com/tutorials/integrations/firefox-sidebar/#additional-about-settings)
user_pref("browser.ml.chat.sidebar", false);
user_pref("browser.ml.chat.menu", false); // remove "Ask a chatbot" from tab context menu
user_pref("browser.ml.chat.page", false); // remove option in page context menu (https://github.com/mozilla/policy-templates/issues/1230)
user_pref("extensions.ml.enabled", false); // might only be relevant for app developers
user_pref("browser.ml.linkPreview.enabled", false);
user_pref("browser.tabs.groups.smart.enabled", false); // "Use AI to suggest tabs and a name for tab groups" in settings
user_pref("browser.tabs.groups.smart.userEnabled", false);
user_pref("pdfjs.enableAltTextModelDownload", false); // "This prevents downloading the AI model unless the user opts in (by enabling the toggle to "Create alt text automatically" from "Image alt text settings" when viewing a PDF)"
user_pref("pdfjs.enableGuessAltText", false); // (disabling this might be redundant when AltTextModelDownload is disabled)
[This comment has been deleted by an automated system]
This right here. It's one of those situations where always asking the company to "Just listen to your users, ez, lol!" results in something that, surprisingly, quite a few people dislike.
Because yes, I'd say Most People™️ actively like AI-generated stuff, and of course there are valid use cases (like summarizing text and video content, the latter even helping with accessibility). But this in turn means that if a company listens to user input, they will shove AI into their products even if they would otherwise not be looking at doing it.
of course there are valid use cases (like summarizing text and video content, the latter even helping with accessibility)
only if you don't care about accuracy or completely made up transcription that just didn't happen.
If a majority of users want it, offering the feature makes sense, as long as you can disable it.
Yea, the uncreative shitstains who can make nothing on their own because they do not understand the first thing about hard work or art.
I am not that person but I use AI to figure out mostly computer issues that I have troubles to Google for.
I guess Google got so fucked up that even adding reddit to a prompt doesn't yeld a good result. And if you didn't add "reddit", results are all over the place and you still fail to find an answer. Fuck quora, who dafuck uses quora? Never saw a good answer there like ever.
And then there is chatgp that gives you options and so far it's been quite on point at least for me.
My guess would be it is general Google entshittification rather than superiority of LLMs. People want answers for their questions and Google is failing to provide any good info for past 5 years.
SEO killed the useable web and AI is killing the open web.
I feel bad for people who never got the chance to learn things in the golden age of the internet. It was so easy.
[This comment has been deleted by an automated system]
grow a goddamn spine and listen to your users. People do not like these features and expect better from you
This might shock you but Firefox is trying to appeal to people who are not already firefox users.
An additional shocking revelation of which you seem to be unaware is that a lot of people do find value in these chat bots.
I do mostly kind of agree with you. Yes Gen AI is over-hyped. Yes their is a dangerous investment bubble. Yes I wish Firefox would just focus on being a great browser.
Yeah this is a feature we have because mozilla listens to their users.
This is tough to swallow for most of us - I hate this crap except for a minority of ultra-narrow use cases - but the general population of users loves AI. They want more of it.
This might shock you but Firefox is trying to appeal to people who are not already firefox users.
No they aren't. They are trying to appeal to investors and advertisers.
https://blog.mozilla.org/en/advertising/index-exchange-partnership/
... and what, pray tell, do advertisers and investors desire?
The same thing they always want: Surveillance, behavior tracking, and avenues to advertise.
This might shock you but Firefox is trying to appeal to people who are not already firefox users.
Fuck them if they want the opposite of what firefox users want. Stop fucking chasing whoever isn't already on board with your current offering by fucking over your current users.
Firefox isn't really viable with it's current user base. Mozilla is circling the drain really.
It could have been viable with the user base that they had years ago, and it probably could be viable today, but Mozilla keeps throwing away all of their goodwill so that they can keep paying absurd salaries for shitty executives.
@null_dot If all of the not Firefox users are AI-shills, then I don't think we're missing much from excluding them.
And what value? Wasting a load of water just for the chatbot to tell you that "strawberry" has 2 r's in it? There is nothing generative-AI can do that other tech can't do better and costing far less resources.
The "value" of generative-AI outside of highly niche scenarios is an illusion. For example, people think generative-AI in the workplace improves productivity when it actually does the opposite: https://www.forbes.com/sites/carolinecastrillon/2025/10/02/ai-workslop-could-be-the-biggest-threat-to-productivity/
This is a good quote from that article: "A recent MIT Media Lab report found that 95% of organizations see no measurable return on their AI investments. Workslop helps explain why. When employees use AI to create low-effort output that shifts the burden downstream, any productivity gains are lost."
Generative-AI has also actively made search engines worse, especially with their "summaries" containing significant errors about 50% of the time: https://www.bbc.com/news/articles/c0m17d8827ko
And I can go on and on. There is an entire laundry list why generative is just... not it, and whatever "value" people get from these can be gotten elsewhere, usually in higher quality and more ethically than relying on this tech.
I don't think we're missing much from excluding them.
Another shocker for you... mozilla isn't viable with it's current user base.
Idiots love AI chat bots, and there are plenty of idiots around.
@null_dot Even if that were true, that doesn't consider the fact that generative-AI is a money black hole and literally the only company type profiting from this bubble are ones like Nvidia that are producing the graphics cards these models use. Generative-AI is not and will not be profitable; Mozilla is losing money by shilling this tech.
So even if you do attract a high volume of users by shoving in generative-AI (which I DOUBT), I doubt that would offset the money burning from running these models. It's not a smart business move, either for users who despise the tech or users who love it. It's a lose-lose.
Generative-AI is not and will not be profitable
Sorry, this is objectively false.
Gen AI is not presently profitable due to the rate of innovation. Developing a new inference model needs a lot of cash for salaries. Populating that new model needs a lot of compute.
Once you have the model, the cost to query it is minimal. You can literally buy a $500 graphics card and download a model and have it perform useful tasks.
The difficulty is, if it costs you $n billion to develop whatever model today, and it's obsolete in 1 year, then thats a $n billion hit to your profitability this year. If innovation slowed down so maybe a model is still competitive after 10 years, then your costs have reduced by 90%.
There's loads of things to dislike about AI, but the profitability thing is borne of misunderstanding on your part.
The arguments against AI you're looking for are:
I would take issue with your implicit suggestion that humanity becoming slaves to AI would be fine if the machines doing it weren't under the direct central control of their elite owners. The gradual replacement of human thinking with rapid low-cost mediocre pseudo-thought is a problem for more than just the worlds of media and art, no matter who's nominally the owner of the process.
No, we're not going to be slaves to a general AI, but it will grant unlimited power to whoever controls it.
There's a large ecosystem of "open" models available to download and develop with. Sadly IMO this open movement is too small and too slow to assail the likes of openAI, but it's something.
An open model available to everyone doesn't concentrate power.
What I'm thinking of depends more on the training methods and data used. But I guess it's too early to say much about what strong AI would be like, we're not really close to it yet.
I mean some of it might turn out to be "profitable" in the narrow sense of earning financial returns for somebody, but that does not mean it will be a profitable endeavour for humanity in general.
i had a bunch of AI BS disabled from a prior paste into my user.js so i actually got jumpscared with an unlabeled entry in my context menu. i decided to dig into about:config to find anything else under browser.ml. and set to false or blank it out.
Generative AI CAN be privacy sensitive, if you use local models. Most of them are too big to do that though, so AI is almost always the biggest fucking security hole possible.
@Eeveecraft @mozilla @firefox When they started adding all that crap, I tried & just kept using Librewolf. Still firefox based, but I don't see all that new bloat anymore. It just sucks there isn't LibreWolf on android.
Ironfox is similar
What about Fennec on Android?
@Eeveecraft @mozilla @firefox What are you talking about? Mozilla saw that everyone was going to their browser to get away from Google, so naturally they did the intelligent thing and became more like Google. I mean, it makes perfect sense, right?
There is a setting in about:config to completely disable it, but I'd recommend that you switch to a fork like LibreWolf on PC or IronFox on Android to ensure stuff like that will be disabled to begin with even if later updates would turn it on.
@nazokiyoubinbou @Eeveecraft @mozilla @firefox Is IronFox similar to LibreWolf default settings profile? I haven't heard of that one & i hate that #librewolf isn't on android...
@Zander @Eeveecraft @mozilla @firefox Yes, I'd say they're fairly similar. It's a fork of the mobile version of Firefox, so it's kind of a different beast in its own way, but they take similar views and methods on privacy and security.
@nazokiyoubinbou @Eeveecraft @mozilla @firefox Hell-Yeah! I'll look it up on fdroid or aurora store. Thanks! 😉
I find this very useful, I wouldn't normally click in articles just read the title, so at least now I can read a summary.
I'm one of the people who actually likes the (e.g. DDG) AI overview for searches. Mainly because it cites its sources.
It allows me to bypass the SEO'd shit factories that make regular search useless.
Having said that, I'd seen that context menu option too and cringed.
I have the ddg ai for on demand. But I remember the good old days when it would still provide you and answer summary box but without using generative AI