top 83 comments

sorted by: hot top controversial new old
[–] 53 points 1 month ago (1 child)

What happened to just making a damn good browser

  • source
  • hideshow 1 child comment
  • [–] 53 points 1 month ago (1 child)

    Mozilla is in the sweet spot of basically being the only available browser that isnt chrome, that has adblocking.

    and they are doing everything in their power to fuck it up by chasing this bullshit AI trend thats already dying on the vine.

  • source
  • hideshow 1 child comment
  • [–] 15 points 1 month ago

    On one side, its great that forks exist that strip out a lot of the bullshit. On the other hand if main Firefox dies or goes too far into the deep end, there's no way those forks can take its role

  • source
  • parent
  • [–] 82 points 1 month ago (8 children)

    yeah no thanks. luckily librewolf strips this shit out.

  • source
  • hideshow 8 child comments
  • [–] 10 points 1 month ago (7 children)

    or you can just turn it off from settings

  • source
  • parent
  • hideshow 7 child comments
  • [–] 29 points 1 month ago (6 children)

    Or they can stop trying to force it through their users throat and just accept that the biggest part of their users don't want it.

  • source
  • parent
  • hideshow 6 child comments
  • [–] 13 points 1 month ago (2 children)

    I mean, they do give a single toggle that disables every AI feature (new and old) at once... I think that's pretty good

  • source
  • parent
  • hideshow 2 child comments
  • [–] 9 points 1 month ago

    But why isn't it a toggle that turns these features on instead?

    It's bullshit that we're expected to constantly opt-OUT of these features that we supposed want. If we wanted them, we would turn them on.

  • source
  • parent
  • [–] 6 points 1 month ago (2 children)

    accept that the biggest part of their users don’t want it.

    Where did you find this data?

  • source
  • parent
  • hideshow 2 child comments
  • [–] 18 points 1 month ago*

    I don’t trust any feature that uses the word “AI.”

    Ideally, such things should be transparent to the user. As in features that are useful without having to understand the underlying programming mechanisms.

    Take FF’s local translation. That’s an immensely useful feature. No one cares that it uses a language model or about any of the details, because it just works well enough, and is useful enough to stand on its own as an advantage over Chromium.

    …If they have to brand a feature to make it appealing, that’s a bad sign. Otherwise why TF would I want expensive tab sorting or things like that?

  • source
  • [–] 17 points 1 month ago* (2 children)

    It doesn't say anything about how this works. "No data retention" suggests that data is going to a server that's processing it but then not retaining it. That's bad, this type of feature should run entirely on the client.

    Also, there's a proprietary AI model doing the work, even if inference is in the client. It says "open weight" but that's like letting people download a program as a compiled binary, the exact opposite of what FOSS is supposed to mean. It's an open/free model only if the training data and algorithms are also public, not just the trained weights.

  • source
  • hideshow 2 child comments
  • [–] 1 point 1 month ago (1 child)

    Issue is that local AI hog vram, are slow and small models. Or if you don't have a compatible GPU, it uses the CPU which is even slower.

    If AI is the future, I really hope it's local, but with the equipment people have at home, local AIs are a non-starter.

    IMO it's enough to justify not shipping those AI features at all but I guess that's where FF execs disagree.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 1 point 1 month ago

    AI inference is doable on home equipment. AI training is less feasible, but nobody builds their whole Debian distro from source either. It's enough that the source is available and you can rebuild any component if you want ot do so.

  • source
  • parent
  • [–] 45 points 1 month ago (7 children)

    I don't really mind AI features where it makes sense. That being said, as someone else commented, this should have been a plugin.

    I'll add on top of that: this should have been a plugin wholly developed by Mistral. Mozilla should stop spending time and money on side features and focus on the browser and its engine. It still is way slower than any chromium based browser, lacks features that have been standard on chromium for years (eg. HDR video playback) and it's way less efficient on battery than the competition.

    Complaining about Firefox doing this and groveling on "I'm good, I'm on " is spitting upwards: even the most modified forks like Zen still rely on Mozilla's aging engine.

  • source
  • hideshow 7 child comments
  • [–] 12 points 1 month ago* (last edited 1 month ago)

    Exactly. The Blender foundation has stated that it is a tool by humans and for humans, so no AI functionality is planned. However, there are still MCP extensions and the like. There’s nothing stopping anyone from using AI with Blender, but the Blender Foundation funding goes to core development. Not sure why Mozilla isn’t doing the same. If I really want a LLM controlling a browser, playwright-cli works fine anyway.

  • source
  • parent
  • [–] 3 points 1 month ago*

    It still is way slower than any chromium based browser, lacks features that have been standard on chromium for years

    I don't think its slower at all. Most of the time same if not faster. Maybe it depends on where you go?

    Features in chromium are often there because they are going their own way and not being standard, is that Mozillas fault?

    HDR video is in Firefox, HDR images is experimental. Personally I don't care, why would I watch videos in a browser that I think should have HDR?

    I will agree that Mistral should have done the lifting, and Firefox need not be working so hard on adding AI. However, if they don't people on the other side of the fence will complain that Chrome gets all the AI, why is Firefox so lazy?

  • source
  • parent
  • [+] -1 points 1 month ago (4 children)
  • [–] 1 point 1 month ago (3 children)

    I am on the latest version of FF on Linux and I don't have HDR working, not on YouTube, nor on standalone players like Jellyfin. Yes my display is HDR and Wayland+gnome is exposing HDR properly to games and Helium (chromium fork).

    Since some govt and utilities sites only work on "chrome" (borked even with user agent switcher set to chrome+win on FF) I had to get this other browser, and the speed difference is noticeable on most sites, even tho theoretically FF should be faster since its stripping a lot of crap with uBlock Origin.

    Tried flatpak'd Brave and also runs snappier than FF.

  • source
  • parent
  • hideshow 3 child comments
  • [–] 14 points 1 month ago (1 child)
  • [–] 14 points 1 month ago (6 children)
  • [–] 7 points 1 month ago (5 children)

    The translation feature has been very useful, though it's not an LLM. Still a neural network though. Saves me from going to google translate when browsing websites in e.g Russian (drive2.ru or whatever turns up in google search every so often when looking up obscure information on some cars and european oils that bitog forums don't see as much).

    It entirely depends on what the AI does and if it's opt-in, opt-out, or forced on you. Opt-in local AI with sensible features is actually a good thing IMO.

  • source
  • parent
  • hideshow 5 child comments
  • [–] 1 point 1 month ago (2 children)

    On principle I only want a pure FOSS program. I wouldn't want to use a binary-only browser even if it was a great browser with good privacy. I do use the built in Firefox translate feature but I've been wanting to get around to seeing where its model comes from. The Mistral stuff is completely proprietary except for the final weights as far as I can tell. I'd rather resist that since I want to be able to build and modify it myself.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 3 points 1 month ago (1 child)

    The Mistral stuff is completely proprietary except for the final weights as far as I can tell

    Same as most open weight models. There are very few truly open source models, particularly if your definition requires distributing the whole training dataset.

    I’d rather resist that since I want to be able to build and modify it myself.

    I hope you've got some damn beefy hardware then. Even small models are a pain to train from scratch. You CAN still fine-tune it or train with LoRA.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 5 points 1 month ago

    Same as most open weight models. There are very few truly open source models, particularly if your definition requires distributing the whole training dataset.

    Yes exactly. Open weight is like a compiled binary. Misleading terminology made that sound desirable.

    I hope you’ve got some damn beefy hardware then. Even small models are a pain to train from scratch. You CAN still fine-tune it or train with LoRA.

    It's ok I can manage without AI.

  • source
  • parent
  • [–] 33 points 1 month ago (2 children)

    As long as this is opt-in (i.e. disabled by default), then sure. Otherwise, I'll just have to find a new browser.

  • source
  • hideshow 2 child comments
  • [–] 9 points 1 month ago (1 child)

    ive been seeing screenshots of the new vibe coded new tab replacement that are totally broken and the redesigned ui has overly rounded components that cover up text components and hard coded "ai" purple gradients that are nearly impossible to turn off with going into about:config...

    So I have low expectations of this going well and seeing that second option in my future.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 11 points 1 month ago (2 children)
  • load more comments (1 reply)
    [–] 17 points 1 month ago (1 child)

    I gotta admit, I don't really mind AI as long as it doesn't exist anywhere in any amount

  • source
  • hideshow 1 child comment
  • [–] 17 points 1 month ago

    They should stop doing that, instead

  • source
  • [–] 17 points 1 month ago (2 children)

    Even if LLMs were environmentally and socially sustainable they would still be a massive IP theft meant to make the rich even richer. Mozilla and Mistral can go eat a bag of dicks.

  • source
  • hideshow 2 child comments
  • [–] 14 points 1 month ago (1 child)

    I don't believe in IP, so that really doesn't bother me much. The problem is when information gets harvested from the public and suddenly becomes someone else's IP.

    So not really theft, more like gatekeeping, but as it stands, its the same result: rich get richer, we can agree on that.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 6 points 1 month ago

    I think that IP may be a good idea on paper, but as it stands it only benefits the powerful, as in "if I pirate content I risk having my life ruined, companies fed their LLMs with everything but the kitchen sink and nothing happened"

  • source
  • parent
  • [–] 21 points 1 month ago (4 children)

    As long as it's easy to turn off, I'm okay with it.

  • source
  • hideshow 4 child comments
  • [–] 43 points 1 month ago (2 children)

    It’s still code bloat. A lot of these new “features” should have been plug-ins.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 19 points 1 month ago (3 children)

    Mozilla can fuck right off with that forced ai horse shit.

    Waterfox is the way.

  • source
  • hideshow 3 child comments
  • [–] [S] 41 points 1 month ago (8 children)

    Mozilla choosing Mistral for Firefox Smart Window feels like a rare AI partnership that actually fits the product. Privacy, zero data retention and multilingual models are a much better match for Firefox than turning the browser into another Big Tech AI funnel.

  • source
  • hideshow 8 child comments
  • [–] 23 points 1 month ago (4 children)

    Why do almost all your comments sound like AI

  • source
  • parent
  • hideshow 4 child comments
  • [–] 17 points 1 month ago (2 children)

    IMO nothing in that comment sounds particularly AI generated.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 16 points 1 month ago*

    Maybe not AI, but that specific comment does read like marketing copy. Then again, there is that list of three in there. That, plus just sounding like marketing, it could be AI. But I also concede that we, people that is, think in patterns of threes, so that's not a strictly an AI behavior, but I can see how someone could reach that conclusion.

  • source
  • parent
  • [–] 3 points 1 month ago (2 children)

    Why do you say privacy and zero data retention? Has there been any proof of this? What we have learned from the last few years of investigations is companies lie and lie and lie through their teeth about anything regarding privacy.

  • source
  • parent
  • hideshow 2 child comments
  • [–] -1 points 1 month ago*

    Mistral makes a lot of local models you can run yourself, assuming that's what this is about. Nothing would leave your computer

    Doesn't mean it's a good idea to integrate, but it could be worse 🤷

  • source
  • parent
  • [–] 7 points 1 month ago* (3 children)

    tbh i would not mind it that much if its local LLM or if its opt-in (opt out as bare minimum)
    i also wonder if they are gonna use this for the built-in translator (if you think about it technically most translator software are AI,DEEPL uses a hybrid between LLMS and NMT IIRC. )

  • source
  • hideshow 3 child comments
  • [–] 2 points 1 month ago (2 children)

    They have the option to use a local llm but default to online inference. Local is not good enough for most people and people would be really unhappy having their browser use 10gb of ram and spin their fans every query.

    Minstrel is working on specialist small models. Maybe one day they are small enough to run in the background no worries.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 2 points 1 month ago (1 child)

    i see, thanks for the information.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 2 points 1 month ago

    Honestly im not sold yet on the AI in the browser meme but I see a lot of normies at my office say the cant live without google chromes AI integration. Maybe I just hate AI summaries.

    My dream is AI navigating websites I hate like facebook marketplace and finding me deals i might like.

  • source
  • parent
  • [–] 3 points 1 month ago (1 child)

    Is it just open weights again?

  • source
  • hideshow 1 child comment
  • [–] 3 points 1 month ago

    dick, hammer, ... etc

  • source
  • [–] 2 points 1 month ago (1 child)

    Oh yeah, great, sounds awesome. Can I get in line now to not use it?

  • source
  • hideshow 1 child comment
  • [–] 2 points 1 month ago* (last edited 1 month ago) (5 children)

    The obsession with multilingual is ruining internet.

    If you look at Reddit now you have posts which are written in several different languages because everyone relies on backend automatic translation or browser automatic translation. But those content become hard read by anyone not using these technologies. Harder to properly archive in a readable manner, and dependent on expensive tech as AI translation.

    This multilingual, everything is in your language, internet is one of the worse things that has happen to the net. I'm tired of having to change my settings all the time to be able to see what the original person wrote, or spoke in the case of automatically translated videos".

    On a paranoid note, this might be a plan so people stop learning english so we won't be able to communicate with each other without their translation technologies.

  • source
  • hideshow 5 child comments
  • [–] 2 points 1 month ago (4 children)

    Totally!

    We all should be white, blond with blue eyes and speak German, am I reich?

  • source
  • parent
  • hideshow 4 child comments
  • [–] 1 point 1 month ago* (last edited 1 month ago) (3 children)

    Nicht Deutsch, sondern Englisch (oder nicht einmal Englisch, sondern nur die Sprache, in der die Website geschrieben ist, ohne viele Sprachen zu mischen). Zudem spielt die ethnische Herkunft der Menschen für die gesprochene Sprache kaum eine Rolle. Die meisten Englischsprechenden stammen nicht aus dem Land, in dem das Englische ursprünglich entstanden ist. Würde jeder von uns seine eigene Sprache verwenden (meine ist übrigens gar nicht Englisch), müssten Foren ständig Echtzeitübersetzungen einsetzen. Das würde all die zuvor genannten Probleme verursachen und uns für die Kommunikation von solchen Übersetzungstools abhängig machen. Ich weiß, es gibt Leute, die nicht wollen, dass Menschen über einen kleinen, leicht kontrollierbaren Kreis hinaus miteinander kommunizieren können. Aber ich gehöre nicht dazu. Ich betrachte Kommunikation als etwas Positives; deshalb halte ich eine Lingua franca für sinnvoll. Es könnte auch Esperanto sein, aber für mich ist Englisch völlig in Ordnung. Als Gegenbeispiel habe ich diesen Kommentar automatisch ins Deutsche übersetzt, um leichter verständlich zu machen, worum es mir geht.

  • source
  • parent
  • hideshow 3 child comments
  • [–] 2 points 1 month ago (2 children)

    As you say: "use the same language if the website".

    Therefore, people going there regularly aren't using the translator. If you aren't a regular, but the information you are looking for might be there, you use a translator.

    There is no "obsession for multilingual internet" here, nor should it ruins the Internet".

  • source
  • parent
  • hideshow 2 child comments
  • [–] 2 points 1 month ago (1 child)

    What I meant for "multilingual internet" is what companies like Reddit or Google are doing. Forcing automatic translates to the user language automatically.

    For Reddit this means that, as people write there in whatever language instead of the language of the subreddit, when you use the site outside of the official frontend you see a mix of languages. This means that you search for things in one language and get results written in another language, and you might have been looking for something in a specific language, but as automatic translated websites take over they take away your ability to choose in what language do you want to search things. On youtube for instance I cannot longer know in what language the video is going to be spoken. I might look for videos in one language and get mislead by the automatically translated title, and click into the video and get a shitty AI voiceover instead of the original voice of the video.

    For instance I used to search things in Spanish to get "local" websites. Now I search in spanish and I see a bunch of Reddit post about the USA, because reddit forces the translated results into googles feed, and google accepts it. Searching something in a specific language is becoming harder and harder because more and more websites are going multilingual, taking in and serving out any and all languages and doing a bunch of automatic translations in the middle.

    Maybe your perception is different. But for me this forced multilingual internet (and what I mean to this is that the frontends automatically translate the content in the business layer) makes for a poor experience with far less freedom and power over the content I choose to see, compared to what I had before.

    And also the appreciation, that this might lead to more people not seeing the need of learning a second international language. Which, for me, is one of the most important skills in life, as it gives you the ability to communicate with so many people without being dependant on intermediates.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 3 points 1 month ago

    Oh I totally agree we should see the site in its original language, and that we should have the choice to translate it or not.

    Now, what companies accept in their index and what the site owner is doing with automatic translation or not are unrelated.

  • source
  • parent
  • [–] 1 point 1 month ago

    Anything not to use Chinese model huh. Aren't Mistral's models woefully outmatched by literally everyone else in the AI field, by every metrics?

  • source
  • [–] 1 point 1 month ago (2 children)

    Sometimes when I can't find an answer online I'll ask the DDG AI assistant. It's disabled on my PC so I have to ask on a mobile device.

    It still gives me answers that don't make sense. It's still an LLM, Siri with a tad more oomf.

  • source
  • hideshow 2 child comments
  • [–] 4 points 1 month ago* (1 child)

    Honestly I have soured on DDG.

    Not specifically over search assist (even though it’s some useless OpenAI model), but its search is just… bad. I never seem to find what I want, yet it just pops up on some other engine.

    Its big useful feature is the diffusion-generated image filtering, but even that is spotty at best. I wish they would lean on the spam/SEO suppression angle more.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 0 points 1 month ago

    no thanks you can keep the slop machine.

  • source
  • [+] -6 points 1 month ago

    @eicker How do I make the AI Chatbot sidebar fullscreen? The chatbot itself couldn't answer this, suggesting press f11.

  • source