Mozilla's pivot to AI first browsing raises fundamental questions about what a browser should be.
  • Quibblekrust@thelemmy.clubEnglish
    126·
    8 hours ago

    There’s so much repeated paranoia in this article. He makes the same weak points over and over.

    But how do you keep track of what a black box actually does when it’s turned on?

    And later,

    And yes, yes - disabling features is all well and good, but at the end of the day, if these AI features are black boxes, how are we to keep track of what they actually do?

    Why would you have to care? You turned them off. The browser is open source. You can see how it invokes the LLMs. If you turn off the features that invoke LLMs, it will not invoke LLMs. I don’t get it. Where’s the disconnect here? The browser is not a black box. The LLM it talks to in the cloud is a black box. If it doesn’t talk to any LLM… 🤷‍♂️

    The core browsing experience should be one that fully puts the user in control, not one where you’re constantly monitoring an inscrutable system that claims to be helping you.

    Jesus… The bias in this article is extreme and repeated often. “claims to be helping”… he even said earlier that LLMs have a measurable utility. Why are they suddenly merely “claiming” to be helpful?

    Why do you have to constantly monitor something you turned off? Really? Constantly?

    Even if you can disable individual AI features, the cognitive load of monitoring an opaque system that’s supposedly working on your behalf would be overwhelming.

    “Overwhelming cognitive load”. Riiight. I turned off telemetry in Firefox as soon as I installed it. I don’t constantly monitor that setting. There is zero cognitive load. I’ll do the same to the AI features if I don’t want them. Also, again with the “supposedly working for you” or “claiming to be helpful” language. Such bias.

    They promise AI will be optional, but that promise acknowledges they’re building AI so deeply into Firefox that an opt-out mechanism becomes necessary in the first place.

    That’s such terrible logic. If something has an opt-out, it has to be “so deeply” built into it? Are the current new-tab features deeply built into Firefox? Like Pocket and such? They’re opt out. Are address bar completions “so deeply” integrated? They’re opt out, too. Is the crash reporter “so deeply” integrated into Firefox? That’s opt out!

    Hell, you could argue crash reporting is deeply integrated because maybe there are many try-catch blocks all over the code which use it, but if you’re the kind of person who turned it off, does it require an “overwhelming cognitive load” to keep it off? Nonsense.

    This article is a bombastic mass of paranoia and bad logic.

    If Firefox releases AI things you can’t just turn off, that can be easily invoked by accident—gestures, keyboard shortcuts, or whatever—that might send page content to an LLM, then I will stop using it. Until then, I’m happy with Firefox. It will always be more up to date than the derivatives.

    • pory@lemmy.worldEnglish
      8·
      4 hours ago

      Firefox is a black box because you can’t opt out of stuff before you click the ‘update browser!’ button. They’ve added default-on data harvesting, telemetry, ads, and now chatbots to Firefox that you have to track down and disable every time it happens. All self-updating software is a “black box” like this, but Mozilla has lost my trust that their updates will have more good than bullshit. So now I use Waterfox and don’t need to worry that there’s some new scheme to monetize me every time I get a browser update.

    • Undertaker@feddit.org
      10·
      7 hours ago

      Your post is really straining.
      You don’t get several points regarding monitoring. An evolving browser hast to be monitored for these anti features because new ones could be added at any time and thus has to be disabled immediately.

      ‘Claim’ to help is correct. While AI can have benefits, is is advertised as always helpful and has to be integrated into a browser. And that’s two different points with the latter one being wrong. The opposite will be the case. People will lose their ability to think, analyze and decide by themselfes.

      they’re building AI so deeply into Firefox that an opt-out mechanism becomes necessary in the first place

      That’s such terrible logic. If something has an opt-out, it has to be “so deeply” built into it?

      If you don’t get it, I can’t help you.
      Article says: A causes B.
      You argue: B causes A?
      Your following examples are pointless unfitting as well.

      Maybe you should take a deep breath and consider a factual view instead of pseudo arguments.