Highly recommend to read their report, which bypasses the opaque corporate language.
Seems they want to replace Google search revenue by AI money. I would rather they go with this rather than hide with weird explanations.
They want to bring new users, but they don’t realize they risk losing existing users by doing so.
Not an enviable position to be in IMO.
New AI-native interfaces will replace traditional browsers and email clients. This creates a market opening for products that use chat, fragments, voice to offer new ways of interacting. However, Mozilla’s experiments are early and limited — and we don’t have an overall theory of action.
Oh man I really hate this. Why does AI need to be arbitrarily shoved into absolutely everything? They say later they want to focus more on subscription services and the like, I really hope they lean more on that than the AI stuff.
In fact having a full Mozilla ecosystem of services would be really sick, like Proton but also integrated in the browser really well.
I wonder if their “subscription” with this massive AI focus would have their VPN, Relay, and AI in one (since they’re clearly going head on into this), like Kagi has their AI assistant as part of their subscription. I agree it would be great to have a whole Moz suite in the browser, but I also don’t want an AI centric browser that just decides to do things for me. I can check my own email, do my own searches, and do my own shopping just fine. Moz needs to diversify away from the Google income, but this has so much risk attached.
Its always a roller coaster will Mozilla. I want off this ride.
No denying that LLMs are really good at language understanding. Browsers are primarily used as communication tools, and there’s no way, fortunately or unfortunately, for a browser to ignore LLMs, given their potential utility.
Discounting the (former crypto/web3?) grifters riding the “AI” hype train, devs/orgs genuinely trying their hand & putting in an effort do make “AI” products that feel clunky and unthoughtful, but the ChatGPT moment was Nov 30 2022, a mere 3y ago. This paradigm is all new for everyone, and only through shipping products & features will anyone have a shot at figuring all of this out, imo.
The pushback against LLMs is all too familiar to the one that smartphone apps got back in the early 2010s (ex1 / mirror), and some times for totally valid reasons (ex2 / mirror). Fast-forward a decade and most don’t even think twice before ditching the web/laptop/desktop version for the corresponding smartphone app (ex: banking, social media, news, streaming etc).
I hate this, and except for banking/messengers, I don’t want to switch to mobile. In fact, the more they insist, the less I comply.
The web is wonderful, and easier than apps (no need to support dozens of platforms/devices). In fact, WhatsApp just dropped its Desktop app for web.whatshapp.com embedding.
If you want AI browser, you should be able to download Comet or Atlas. Just like you can have a bloated OS with Windows and open with Linux. Choice is key, and Mozilla by following the crowd like a sheep, erase their identity and edge.
Not saying some AI features can’t be integrated. The new local translation feature is very cool. But AI agents no thanks.
Maybe I am missing it, but I don’t think that Brave gets as much flack for all the AI they do in their browser compared to what Firefox has done. Whether folks like what Mozilla is doing is one thing, but I feel like criticism should also be spread fairly to other companies in this space.
From my perspective, I don’t like the increased attack surface. That being said, having some kind of AI feature does exist in other browsers, and depending on how it’s done it doesn’t have to be inherently compromising (as long as you know how you’re using it and that it was well-implemented). The impression I have of this with the AI window would put this in the category of being solidly compartmentalized, or at least that’s what I hope happens. In that case, I’m fine with leaving it alone and moving on.
I think a big problem is that many people remember when Mozilla used to be cool.
I think having some AI features in the browser is fine, but they need to be tastefully integrated, like any other feature, instead of shoved down everyone’s throat. Also, since these tools need to process user data, that processing can’t happen anywhere but the user’s device. One good example would be web translation in Firefox: it’s AI, nicely integrated into the browser, that runs on-device. I want more of that and less cloud-based LLM chat sidebars.
Hopefully their on their way since the scare about search engine deals. I’m glad that they brought back tab groups, are making profile management easier to use, and I think are working on workspaces and PWA support. For me, they’re doing a lot of things right from a feature perspective. I still don’t see the sustainable business model, but if AI helps for those who want to use it, I think I can be ok with that.
Definitely true. There is two main reasons IMO
- They always targeted techie/nerdy users, and always had bloat, but this is counterbalanced by being Chromium and having integrated anti tracker
- They are integrating AI in a better way
- More private: they only use their own Chatbor service. The later uses both Brave-hosted models and Claude models with a no-train agreement. (Firefox just propose you 4 chatbots with an embedding. Most of them require an account)
- More secure : it has been months that they research security vulnerabilities in other AI browser with agentic capabilities. (Mozilla on the other hand has treated Firefox security as secondary)