Disable AI in Firefox
flamedfury.com
flamedfury.com
Or—and this happens—it "summarizes" the same text differently, depending on whether the author's name happens to fit a certain ethnicity.
With how inexpensive trainings are starting to get, it will not be long until we can train our own specialized models to fit our specific needs.
I am pessimistic on that front, since:
1. If LLM's can't detect biases in their own output, why would we expect them to reliably detect it in documents in general?
2. As a general rule, deploying bias/tricks/fallacies/BS is much easier than the job of detecting them and explaining why it's wrong.
Maybe in five years they will be useful enough that it would have been worth including these features
This is nice to know but in future versions of Firefox that single config switch (browser.ml.enable) will both change names and split into multiple sub-switches, most likely appearing in different pages of about:config.
These sub-switches will then not remain consistent.
Bank on it.
Almost no one changes defaults, according to the experts
It's perverse to think about:config is doing favours for anyone except Mozilla and its business partners
Otherwise the defaults would be false, i.e., opt-in
Also pick a lane mozilla. Either use "enable" or "enabled", don't use both.
I agree with the sentiment, but you underestimate the level of engineering, coordination, design work, testing it is to do this.
It is admirable that they even have a half-decent browser, but to compete at the top you need soooo much money and motivation.
Let's not forget the CEO who paid herself a $6.9m salary in 2022, $5.6m in 2023.
I’m guessing Ladybird will prove you wrong in due time
Do you just need to advertise stuff among content creators these days with common sense going out of the window? It'll take them a decade to catch up without any engineering funding at the level that Apple/Google/Mozilla have.
I’m not a content creator and I don’t really care about Ladybird. I use Safari.
I’m just pointing out that browsers have decades of legacy cruft from mis-steps deciding what the web even should be and someone smart can carve out a path to covering 90% of use cases in 10% of the effort and code. And there are the huge organizational costs Google and others pay that a small organization doesn’t have to.
Your argument is the same as looking at a large company (say Microsoft) and saying no one can compete without trillions of dollars and tens of thousands of engineers. Ladybird has the benefit of hindsight, as well as a non-idiotic structure (I assume).
The defeatism among engineers is sad
If you're really smart and you say "I can do it with half or a quarter of the resources with hindsight", sure I might give you the benefit of the doubt, but if you're going to claim you can do it with 0.1% of the resources in a volunteer Discord server effort, no. Not because I wouldn't be happy if that was possible, but because that's not how the world works. Linux is being able to compete with Microsoft because there are now large billion dollar companies like RedHat, Steam and others investing into the development. It takes real money and developer time.
And that's the second point, Mozilla has to make these compromises because they are one of the few companies that actually maintains an independent software project at this scale. And if any other competitor ever wants to get there, they'll need to answer these funding questions too. Even if they're ten times as clever, they'll still need tens or hundreds of millions.
Ask 60% of their (Chrome, Edge, Safari) userbase, and they won't even be able to tell you what their browser is called.
It'll be a usable product, but it will be extremely extremely niche, until the dev burns out or just quit it.
I hope I'm wrong, but a browser is a XXL type project and needs proper funding (means = there should be a reason for it to exist, not altruistic as lets have an alternate because reasons ..)
It's been 20 years since Apple decided they needed a browser of their own, and even then they chose to throw their weight behind KDE's KHTML, not Gecko.
Deleted it in my config. I'm solely relying on DuckDuckGo.
I've been fully on DDG for years but becoming slowly skeptical & looking for alternatives.
1. They're leaning heavily into "responsible AI", much like Mozilla
2. Might be just me but I feel like their algorithm became significantly worse recently. Over the years they've gone from being worse than Google in the early days to steadily improving & overtaking Google on quality (I made heavy use of !g until I started slowly realising it was no longer giving me better results). But now I feel like they've reversed & regressed again.
(though it might be a case of there simply being no good search anymore)
Maybe they'd be in a better position if they focused their resources on building their core product? I know that's a wildly radical concept these days...
https://www.techradar.com/vpn/vpn-services/mozilla-is-experi...
https://ftp.mozilla.org/pub/firefox/releases/
You're on your own running these in the wild; at least use a few other layers of security protection. Later versions of Firefox will not run unsigned addons unless you're running a development build, without running it from the debugger page. (Maybe someone can chime in with a workaround.)
Selecting text and having a action with a custom prompt/tool without needing a browser extension might be nice. It need not even be a llm.
https://nilsheuman.github.io/TIL/2025-04-17-custom-firefox-a...
I found family to do a decent job of not resolving ads and crapware. Other providers are available.
Your comment is a bit self-contradictory in that it doesn't say what it says.
I don't want anything even vaguely related to spicy autocomplete on any of my machines, and I go to great lengths to kill anything that even resembles it with fire.
IMHO just like many companies are obsessed with adding AI features, some users are obsessed with rejecting them. Both seem mostly senseless to me, especially when it's a local-first AI implementation.
The toggle for the popup is in (as you might expect) the settings hamburger menu in the AI Panel. There's even a remove button! Lots of new things have been added to browsers over the years, and these AI features are becoming incredibly popular as more users recognize the utility (and what they actually are).
This just seems like some more anti-AI hysteria.
It's clutter for a feature that I'm not going to use. I'm not upset it's there for those who want it, but it's also nice to be able to get rid of it.
brow-ser.ml.enable
However, my main reason for ditching Chrome years ago was the fact that I think a browser engine monoculture is bad for the web as a whole, especially if that engine is primarily controlled by a single corporate entity.
Manifest v3 and other Google nonsense came later, and are extra reasons to stay away from Chrome, but I still feel strongly that a good alternative needs to use a different engine.
That lack of capability prevents it from being my daily driver, even if the rest were good enough (I’m not saying it isn’t, I’m saying I have no reason to find out).
I am certain I have inadvertently pushed many people away from Firefox for that reason alone, because when they ask for me to add Firefox support for my tools, I have to tell them it’s impossible.
I have tried to talk to Firefox developers about that a few times, at open-source conferences and such, but they think AppleScript is some power-user feature and fail (refuse?) to understand power users drive adoption and create tools that regular users rely on.
I remember whenever a Firefox story was submitted on HN, multiple people commented “I want to use Firefox but it’s missing <whatever>”. Then Mozilla started doing a lot of questionable stuff (all of which they eventually abandoned) outside their core competency and even pulling distasteful marketing stunts, and at some point people started commenting even that. I presume many got tired and gave up on Firefox entirely. I almost have. I now root for them only conceptually, because browser diversity is good.
I also noticed that no matter how politely someone pointed out on HN “Firefox doesn’t fit for me because of <whatever>”, they always got downvoted. If valid polite criticism is buried, no wonder things stay the way they are.
https://data.firefox.com/dashboard/hardware
They're really not going to be able to dedicate resources to something as bijou as AppleScript.
They don’t need to do it themselves, they could just not stifle the efforts of third-parties who do want to and have worked on it. Multiple people started on it over the years and were simply ignored by the devs.
Anyway, the reasons are irrelevant and I’m frankly tired of explaining this to Firefox defenders. Someone asked “what about Firefox are you missing” and I responded with what it’s missing for me. Plugging your ears and coming up with excuses doesn’t move the needle. Accept it or don’t, it makes no difference. In the meantime I’ll continue being honest with my users that I would like to support Firefox but I can’t, and many of them will keep switching browsers.
As a user, I do not want nor need my browser to support AppleScript. AppleScript is something that should not exist. In somewhat typical apple fashion, it's some NIH platform-specific bullshit that nobody really cares about and is only half-assed supported even on it's native platform. The only way to deter Apple from creating these sisyphus-ian pieces of software is to just stop supporting them and force their hand to use something less bespoke. Although, Apple is not the only culprit of this - nor are they even the worst about it.
If I had my way, Mantle would not exist, iMessage would not exist, and some others. We would live in a perfect utopia and then we'd all hold hands and sing Kumbaya.
I’ll say it again:
> Plugging your ears and coming up with excuses doesn’t move the needle. Accept it or don’t, it makes no difference.
> If you’re going to keep ignoring my points and making up new things while moving the goalposts, I don’t see the point in having a discussion.
I'm not ignoring anything - it's called disagreeing with your points. Because I do disagree with them, to an extent.
I agree that Firefox is missing features. I DO NOT agree that this is generally, keyword generally, a bad thing.
Also, Firefox is not even the most behind browser. Uh, that would be Safari, and it's not even close.
If you don't support Firefox but you DO support Safari, then:
1. You must not give a flying fuck about web standards or features, as Safari is missing the most of both OR
2. You squarely target primarily Apple users so you have no choice but to put up with Apple's subpar software, i.e. you're Stockholm'd in
For option 2, the only way to stop that is to do what I said - stop playing Apple's games and don't target their bespoke barely-functional bullshit.
Also, let me just say: there is absolutely no shame in number 2. You have to make your money. Just a couple years ago, I was maintaining an application that still targeted IE 6. Yes, really. Sometimes you have to do what you have to do to reach the users where they are.
> You asked for one reason
No, I did not. I didn’t ask for any reason. I don’t need one, the result is the same.
> I gave it to you
No, you did not, you gave speculation. Which changes nothing.
> you didn't like it
I couldn’t care less about your speculation. I specifically said I don’t care. The only thing I don’t like is how you’re just making up reasons based on what you want and like, as if your tastes and needs are all that matters, instead of what’s the truth for Firefox development.
> I'm not ignoring anything
You moved the goal post without addressing my initial reply to your response, which poked a direct hole in your speculation:
> Then they should just say so and close the open issues, instead of letting them linger for literal decades and have people waste time on them then ignore them. That’s just bad stewardship.
You continue:
> it's called disagreeing with your points.
It makes no sense to disagree with a personal opinion! My reasons are my own. It’s like if you said you disagreed with me for saying I don’t like chocolate. Your opinion is irrelevant to someone else’s taste or needs.
And then the rest is just more speculation. Again, Firefox is the only major browser not supporting AppleScript. And just so you know, Chromium browsers typically support it better than Safari.
- Change to first browser tab whose URL or title matches <whatever>.
- Close every browser tab matching <whatever>.
- Grab all your tabs and backup their URLs to a file.
- Join all tabs from all windows into a single window.
- Execute JavaScript on a page and get results back.
- Grab the URL of the current tab and open it in a different browser in a Private window.
- And many more things.
You don't even need an mcp server. Claude Code can just run osascript. https://news.ycombinator.com/item?id=44492369
I don't see why a browser should have to support AppleScript specifically. The Chrome DevTools Protocol and WebDriver BiDi are the standard protocols for interacting with browsers programmatically. Firefox supports WebDriver BiDi. Just use any tool that supports it, or talk to it directly. Maybe AppleScript can do that, I wouldn't know.
(not to diminish the css, it's great for theming and correcting the many usability mistakes in browser defaults, and wish all the browsers used that)
I'm not sure what you mean with the context menus, but Web Extensions can add things to context menus.
If you've tried chrome recently, you'll know that it's jam packed full with even more stuff you don't want. And the article lays out how to easily disable all AI in firefox (which you cant do at all in Chrome)
I disagree with Mozilla here, too - but you can't cast Chrome as a magic spell. Chrome sucks ass. Google sucks ass. It's trivial to suck less ass than Google.
It doesn’t to tabs, and links that the site forces to open in a new tab often don’t work. It also doesn’t do JS well by design.
I use Firefox focus for throw away links I come across, but for everything else I need a full browser
If you’re running in a Custom Tab on Android, you need to switch to the full Focus if multiple tabs are involved.
Step 2: ??
Step 3: Profit
Sometimes HN commenters are so funny.
What makes you think it'd happen if full-time employees at Firefox cant do it? We can poop on the leadership over at Mozilla, but there are FTEs getting paid to work on Firefox.
You cant just replace with few people running passion project on weekends, and even get the remote success Firefox has.
Browser is extremely complex. HN is underestimating how much work goes into making a browser.
The full-time engineers are given work to do by incompetent Mozilla management. It's the management that are driving Mozilla into the ground and setting baffling goals. Remove the management and have work based on features that users want, then you can see Firefox develop in the right direction.
As an example of how to organise this, you could have a bounty system for feature requests. Users define a feature they want to see and in negotiation with developers set acceptance criteria for when it's delivered. Users can then assign money as an incentive to complete the feature request. In this way, users can ensure they support developers to deliver the features they want to see.
> Browser is extremely complex. HN is underestimating how much work goes into making a browser.
Nobody is underestimating this. Firefox is already a mature product that can serve a wide range of user needs. What it lacks is effective leadership. I could live with slower development if the development it had was based around features that users most wanted. I don't need Firefox to support every web feature under the sun, the features it already supports is good enough for the vast majority of websites. Letting the users call the shots about it's future direction will help to guard against irrelevancy.
Could you actually provide any examples of large-scale software projects working this way, and not the usual way with coordination being done by special people or w/o any (e.g., by the devs)
A good idea is nothing without good execution. And execution requires patience, coordination, a good amount of luck .. and so much ideas.
It is like what they say - Ideas are infinite. Nobody cares about them.