It depends on what they mean by AI. I can think of oodles of great uses:
An AI-powered adblock that removes all trackers, cookie confirmation popups, those annoying “please subscribe” popups, etc. would be badass. It would be virtually invisible but it would make the internet usable again.
A content filter that magically extracts the recipe you’re looking for out of the stupid blog post they write for SEO
Or to expand on that, an AI that goes through the page of search engine results and removes the ones that are SEO spam instead of actually useful content
An AI that can review at a page or email and determine if it’s a scam would save a TON of people by pointing out suspicious features.
Basically anything that requires you to copy data from one context to another is a good use of AI. You could probably have a nice resume-filling feature, for example.
But yeah, Mozilla will probably just go for a “chat with your browser” feature. Total waste of space.
All of those could be terrible to be honest, because AI is a data tracking vacuum. An AI adblocker or content filter sounds cool at first, but it would mean it reads and analyzes your data, just like the shit you do with chatbots too. Reading your mails? That’s basically what Google does for years with gmail, that’s why they have such a good spam filter. I agree that a chatbot would be kinda useless though, even if privacy friendly, which in of itself would be great but I just don’t see the use. This could simply be outsourced to a website.
If they’re local they’d be basically useless due to a lack of computing power and potential lack of indexing for a search engine chatbot, so I doubt it. It would also have to be so polished that it wouldn’t require further user knowledge / input, and that’s just not a thing with any local LLM I’ve come across. Mozilla can gladly prove me wrong though. I certainly wouldn’t mind if they generally can make the whole process of local LLMs easier and more viable.
The requirements to run good local LLMs have really been shrinking this past year… I have a lot of faith that there is a generally useful yet tiny AI tool within the grasp of Mozilla.
I can understand your thinking, but it could be as simple as giving the user the option to outsource the computation to a secure something or other, if their machine can’t handle it.
And yeah, the requirements are still quite high, but they are being reduced somewhat steadily, so I wouldn’t be surprised if average hardware could manage it in the long term.
Edit: For the record, Mozilla is one of the only companies I would trust if they said “the secure something or other is actually secure.” And they’d likely show actual proof and provide and explanation as to how.
Yes, but what would a local model do for you in this case? Chatbots in browsers are typically used as an alternative / more contextualized search engine. For that you need proper access to an index of search results. Most people will also not have enough computing power to make use of any complex chatbot / larger context sizes.
Pennomi wrote a whole list of potential ideas. And honestly, while I agree that local LLMs on typical hardware are underpowered for most tasks, it’s possible they would have the option for those that can run it.
People are getting all upset over this announcement without even knowing what their plan actually is, like the word “AI” is making them foam at the mouth or something. I’m just saying we should reserve judgements for when we have an idea of what’s happening.
Yeesh people here are salty.
Honestly, if they make it optional and/or give the option to run it locally, I could see this being a good thing.
Lord knows the competition is going full bore on AI, and if FF wants to stay relevant with the mass market they’ll need to keep up.
It depends on what they mean by AI. I can think of oodles of great uses:
But yeah, Mozilla will probably just go for a “chat with your browser” feature. Total waste of space.
All of those could be terrible to be honest, because AI is a data tracking vacuum. An AI adblocker or content filter sounds cool at first, but it would mean it reads and analyzes your data, just like the shit you do with chatbots too. Reading your mails? That’s basically what Google does for years with gmail, that’s why they have such a good spam filter. I agree that a chatbot would be kinda useless though, even if privacy friendly, which in of itself would be great but I just don’t see the use. This could simply be outsourced to a website.
The only reason this would be an issue is if it’s sending that data off to a third party. If its fully local, who cares what data it sees?
If they’re local they’d be basically useless due to a lack of computing power and potential lack of indexing for a search engine chatbot, so I doubt it. It would also have to be so polished that it wouldn’t require further user knowledge / input, and that’s just not a thing with any local LLM I’ve come across. Mozilla can gladly prove me wrong though. I certainly wouldn’t mind if they generally can make the whole process of local LLMs easier and more viable.
The requirements to run good local LLMs have really been shrinking this past year… I have a lot of faith that there is a generally useful yet tiny AI tool within the grasp of Mozilla.
I can understand your thinking, but it could be as simple as giving the user the option to outsource the computation to a secure something or other, if their machine can’t handle it.
And yeah, the requirements are still quite high, but they are being reduced somewhat steadily, so I wouldn’t be surprised if average hardware could manage it in the long term.
Edit: For the record, Mozilla is one of the only companies I would trust if they said “the secure something or other is actually secure.” And they’d likely show actual proof and provide and explanation as to how.
There’s no way it would be running locally.
In their defense, Mozilla did create the easiest way to run and integrate an LLM locally, so if anyone could do it, I imagine it would be them.
https://github.com/Mozilla-Ocho/llamafile
Yes, but what would a local model do for you in this case? Chatbots in browsers are typically used as an alternative / more contextualized search engine. For that you need proper access to an index of search results. Most people will also not have enough computing power to make use of any complex chatbot / larger context sizes.
Pennomi wrote a whole list of potential ideas. And honestly, while I agree that local LLMs on typical hardware are underpowered for most tasks, it’s possible they would have the option for those that can run it.
People are getting all upset over this announcement without even knowing what their plan actually is, like the word “AI” is making them foam at the mouth or something. I’m just saying we should reserve judgements for when we have an idea of what’s happening.
And I replied to that comment, without any mouth foaming.
Yes, and then you asked for ideas, which were in that comment that you replied to.
I honestly have no idea what you’re referring to now. I never asked for any ideas.
deleted by creator