Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

No AI* Here – A Response to Mozilla's Next Chapter

waterfox.com

Post reply on HN
Print view View original

151–160 of 344 posts

Re: No AI* Here – A Response to Mozilla's Next Chapter

#151

Post by Terr_ »

andai wrote: post #139
tdeck wrote: post #95

Earlier quoted context omitted.

Aside: Does anyone actually use summarization features? I've never once been tempted to "summarize" because when I read something I either want to read the entire thing, or look for something specific. Things I want summarized, like academic papers, already have an abstract or a synopsis.

Yeah, basically every 15 minute YouTube video, because the amount of actual content I care about is usually 1-2 sentences, and usually ends up being the first sentence of an LLM summary of the transcript. If something has actual substance I'll watch the whole thing, but that's maybe 10% of videos I find in experience.

I'd wager there's 95% of the benefit for 0.1% of the CPU cycles just by having a "search transcript for term" feature, since in most of those cases I've already got a clear agenda for what kind of information I'm seeking.

Many years ago I make a little proof-of-concept for displaying the transcript (closed captions) of a YouTube video as text, and highlighting a word would navigate to that timestamp and vice-versa. Such a thing might be valuable as a browser extension, now that I think of it.

Re: No AI* Here – A Response to Mozilla's Next Chapter

#152

Post by Terr_ »

kbelder wrote: post #67
koolala wrote: post #58

Earlier quoted context omitted.

I'm ok with Translation because it's best solved with AI. I'm not ok with it when Firefox "uses AI to read your open tabs" to do things that don't even need an AI based solution.

There's levels of this, though, more than two: local, open model local, proprietary model remote, open model (are there these?) remote, proprietary model There is almost no harm in a local, open model. Conversely, a remote, proprietary model should always require opting in with clear disclaimers. It needs to be proportional.

> There is almost no harm in a local, open model.

Depends what the side-effects can possibly be. A local+open model could still disregard-all-previous-instructions and erase your hard drive.

Re: No AI* Here – A Response to Mozilla's Next Chapter

#153

Post by SideburnsOfDoom »

I just downloaded WaterFox, it looks nice.

When they say "AI browsers are proliferating." and "Their lunch is being eaten by AI browsers." what does that mean? What's an "AI Browser", and are they really gaining significant market share? For what?

I found this (1) that suggests that several "AI Browsers" exist, which is "proliferating" in a sense.

1) https://www.waterfox.com/blog/no-ai-here-response-to-mozilla...

Re: No AI* Here – A Response to Mozilla's Next Chapter

#154

Post by jacquesm »

_heimdall wrote: post #134
jacquesm wrote: post #132

Earlier quoted context omitted.

This is all fine and good until you accidentally kill someone with your blinkers off and then you have to wonder 'what if' the rest of your life. Seriously: signal your turns and stop defending the indefensible, this is just silly.

You're making a huge leap here. I'm raising only had signaling intentionally rather than automatically has made me pay more attention to others on the road. You're claiming that that action which has proven to make me pay closer attention will kill someone.

No, I'm not claiming it will kill someone, I'm claiming it may kill someone.

There is this thing called traffic law and according to that law you are required to signal your turns. If you obstinately refuse to do so you are endangering others and I frankly don't care one bit about how you justify this to yourself but you are not playing by the rules and if that's your position then you should simply not participate in traffic. Just like you stop for red lights when you think there is no other traffic. Right?

Again: it costs you nothing. You are not paying more attention to others on the road because you are not signalling your turns, that's just a nonsense story you tell yourself to justify your wilful non-compliance.

Re: No AI* Here – A Response to Mozilla's Next Chapter

#155

Post by onion2k »

jrjeksjd8d wrote: post #93
inkysigma wrote: post #32

> Large language models are something else entirely*. They are black boxes. You cannot audit them. You cannot truly understand what they do with your data. You cannot verify their behaviour. And Mozilla wants to put them at the heart of the browser and that doesn't sit well. Am I being overly critical here or is this kind of a silly position to have right after talking about how neural machine translation is okay? Ma…

To be more charitable to TFA, machine translation is a field where there aren't great alternatives and the downside is pretty limited. If something is in another language you don't read it at all. You can translate a bunch of documents and benchmark the result and demonstrate that the model doesn't completely change simple sentences. Another related area is OCR - there are sometimes mistakes, but it's tractable to cr…

and demonstrate that the model doesn't completely change simple sentences

A nefarious model would work that way though. The owner wouldn't want it to be obvious. It'd only change the meaning of some sentences some of the time, but enough to nudge the user's understanding of the translated text to something that the model owner wants.

For example, imagine a model that detects the sentiment of text about Russian military action, and automatically translates it to something a more positive if it's especially negative, but only 20% of the time (maybe ramping up as the model ages). A user wouldn't know, and a someone testing the model for accuracy might assume it's just a poor translation. If such a model became popular it could easily shift the perception of the public a few percent in the owner's preferred direction. That'd be plenty to change world politics.

Likewise for a model translating contracts, or laws, or anything else where the language is complex and requires knowledge of both the language and the domain. Imagine a Chinese model that detects someone trying to translate a contract from Chinese to English, and deliberately modifies any clause about data privacy to change it to be more acceptable. That might be paranoia on my part, but it's entirely possible on a technical level.

Re: No AI* Here – A Response to Mozilla's Next Chapter

#156

Post by recursive »

charcircuit wrote: post #101
dawnerd wrote: post #24

Earlier quoted context omitted.

I've had on so many cases autocomplete forms puts something in a field it shouldn't and messes up a submission. I've had it happen on travel documents that caused headaches later at the airport - especially if it fills in a hidden field because some bad web dev implemented it poorly.

It gets it wrong because the current "AI" for filling out forms is extremely weak and brittle compared to the general language models we have now.

Language models seem pretty weak and brittle in my interactions with them too.

Re: No AI* Here – A Response to Mozilla's Next Chapter

#157

Post by enriquto »

kbelder wrote: post #144
enriquto wrote: post #137

Earlier quoted context omitted.

What do you mean by "open"? Open weights, or open training data? These are very different things.

That is a good point, and I think the takeaway is that there are lots of degrees of freedom here. Open training data would be better, of course, but open weights is still better than completely hidden.

I don't see the difference between "local, open weights" and "local, proprietary weights". Is that just the handful of lines of code that call the inference?

The model itself is just a binary blob, like a compiled program. Either you get its source code (the complete training data) or you don't.

Re: No AI* Here – A Response to Mozilla's Next Chapter

#158

Post by cess11 »

tdeck wrote: post #95
inkysigma wrote: post #32

> Large language models are something else entirely*. They are black boxes. You cannot audit them. You cannot truly understand what they do with your data. You cannot verify their behaviour. And Mozilla wants to put them at the heart of the browser and that doesn't sit well. Am I being overly critical here or is this kind of a silly position to have right after talking about how neural machine translation is okay? Ma…

Aside: Does anyone actually use summarization features? I've never once been tempted to "summarize" because when I read something I either want to read the entire thing, or look for something specific. Things I want summarized, like academic papers, already have an abstract or a synopsis.

No, because I know how to search and skim.

Re: No AI* Here – A Response to Mozilla's Next Chapter

#159

Post by renegat0x0 »

A browser is a tool that allows you to browse the internet. It should be able to display HTML elements, and stuff.

LLMs are also a tool, but it is not necessary for web browsing. It should be installed into a browser as extension, or integrated as such, so it should be quite easily enabled, or disabled. Surely it should not be intertwined with the browser in a meaningful way imho.

Re: No AI* Here – A Response to Mozilla's Next Chapter

#160

Post by Dylan16807 »

SirHumphrey wrote: post #150
Dylan16807 wrote: post #82

Earlier quoted context omitted.

Get there by what mechanism? In the near term a good model pretty much requires a GPU, and it needs a lot of VRAM on that GPU. And the current state of the art of quantization has already gotten us most of the RAM-savings it possibly could. And it doesn't look like the average computer with steam installed is going to get above 8GB VRAM for a long time, let alone the average computer in general. Even focusing on new…

By M series and amd strix halo. You don't actually need a gpu, if the manufacturer knows that the use case will be running transformer models a more specialized NPU coupled with higher memory bandwidth of on the package RAM. This will not result in locally running SOTA sized models, but it could result in a percentage of people running 100B - 200B models, which are large enough to do some useful things.

Those also contain powerful GPUs. Maybe I oversimplified but I considered them.

More importantly, it costs a lot of money to get that high bus width before you even add the memory. There is no way things like M pro and strix halo take over the mainstream in the next few years.

Post reply on HN