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

331–340 of 344 posts

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

#331

Post by account42 »

fsflover wrote: post #272
johnnyanmac wrote: post #235

Earlier quoted context omitted.

Mozilla hasn't had the benefit of the doubt for quite a while here. This isn't just one small kerfuffle coming out of nowhere. They say trust takes a lifetime to build and seconds to break ". We're years into it at this point.

> Mozilla hasn't had the benefit of the doubt for quite a while here In contrast to Google Chrome? This is just FUD. Ublock Origin is still working and will be working. Full customization is still there and isn't going away. All of that is unlike in Chrom(ium).

If you need to compare yourself to the literal devil then you're probably far from a saint yourself.

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

#332

Post by ErroneousBosh »

_heimdall wrote: post #122
jacquesm wrote: post #94

Earlier quoted context omitted.

I would strongly suggest you use your turnsignals, always , without exception. You are relying on perfect awareness of your surroundings which isn't going to be the case over a longer stretch of time and you are obliged to signal changes in direction irrespective of whether or not you believe there are others around you. I'm saying this as a frequent cyclist who more than once has been cut off by cars that were not i…

Its a matter of approach and I wouldn't say what I've found to work for me would work for anyone else. In my experience, I'm best served by trying to reinforce awareness rather than relying on it. If I got into the habit of always using blinkers regardless of my surroundings I would end up paying less attention while driving. I rode motorcycles for years and got very much into the habit of assuming that no one on the…

> Signaling my turns is zero cost with regards to that action. At least for me, signaling as a matter of habit comes at the cost of focus.

What do you mean by "comes at the cost of focus", there? Do you mean you are more distracted by having to use your indicators?

Maybe you're just not a very good driver, if you're so distracted by the basic controls of the vehicle.

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

#333

Post by immibis »

someothherguyy wrote: post #287
PunchyHamster wrote: post #103

Earlier quoted context omitted.

Bad defaults are bad defaults, and "you can turn them off" is not a good excuse for bad defaults continuing to be bad defaults

It is the default in every major browser at this point. https://chromeenterprise.google/policies/#GenAiDefaultSettin...

Bad defaults are bad defaults, and "our competitor does the same thing" is not a good excuse for bad defaults continuing to be bad defaults

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

#334

Post by FeepingCreature »

dkdcio wrote: post #320
int_19h wrote: post #314

Earlier quoted context omitted.

We know how neural nets work. We don't know how a specific combination of weights in the net is capable of coherently asking questions asked in a natural language, though. If we did, we could replicate what it does without training it.

> We know how neural nets work. We don't know how a specific combination of weights in the net is capable of coherently asking questions asked in a natural language, though. these are the same thing. the neural network is trained to predict the most likely next word (rather token, etc.) — that’s how it works. that’s it. you train a neural network on data, it learns the function you trained it to, it “acts” like the d…

It's like you're describing a compression program as "it takes a big file and returns a smaller file by exploiting regularities in the data." Like, you have accurately described what it does, but you have in no way answered the question of how it does that.

If you then explain the function of a CPU and how ELF binaries work (which is the equivalent of trying to answer the question by explaining how neural networks work), you then have still not answered the actually important question! Which is "what are the algorithms that LLMs have learnt that allow them to (apparently) converse and somewhat reason like humans?"

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

#335

Post by FeepingCreature »

schoen wrote: post #299
FeepingCreature wrote: post #245

Earlier quoted context omitted.

> There's an older tradition of rule-based machine translation. In these methods, someone really does understand exactly what the program does, in a detailed way I would softly disagree with this. Technically, we also understand exactly what a LLM does, we can analyze every instruction that is executed. Nothing is hidden from us. We don't always know what the outcome will be; but, we also don't always know what the o…

The grammars in rule-based MT are normally fully conceptually understood by the people who wrote them. That's a good start for human understanding. You could say they don't understand why a human language evolved some feature but they fully understand the details of that feature in human conceptual terms. I agree in principle the statistical parts of statistical MT are not secret and that computer code in high-level…

Sure, I agree with that, but that's a property of hand-writing more than rule-based systems. For instance, you could probably translate a 6B LLM into an extremely big rule system, but doing so would not help you understand how the LLM worked.

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

#336

Post by schoen »

pona-a wrote: post #325
schoen wrote: post #126

Earlier quoted context omitted.

There's an older tradition of rule-based machine translation. In these methods, someone really does understand exactly what the program does, in a detailed way; it's designed like other programs, according to someone's explicit understanding. There's still active research in this field; I have a friend who's very deep into it. The trouble is that statistical MT (the things that became neural net MT) started achieving…

Do you know what is the SOTA rule-based MT? I used to be deep into symbolics but couldn't find much in the way of contemporary rule based NLP.

My friend is working on Grammatical Framework, which has a Resource Grammar library of pre-written natural language grammars, at least for portions of them. The GF research community continues to add new ones over time, based on implementing portions of written reference grammars, or sometimes by native speakers based on their own native speaker intuitions. I'm not sure if there are larger grammar libraries elsewhere.

There could be companies that made much better rule-based MT but kept the details as trade secrets. For example, I think Google Translate was rule-based for "a long time" (I don't remember until what year, although it was pretty apparent to users and researchers when it switched, and indeed I think some Google researchers even spoke publicly about it). They had made a lot of investment (very far beyond something like a GF resource grammar) but I don't think they ever published any of that underlying work even when they discontinued that version of the product.

So basically there may be this gap where academic stuff is advancing slowly and yet now represents the majority of examples in the field because companies are so unlikely to have ongoing rule-based projects as part of projects. The available state of the art you can actually interact with may have gone backwards in recent years as a result!

nimi sina li pona tawa mi.

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

#337

Post by dkdcio »

FeepingCreature wrote: post #334
dkdcio wrote: post #320

Earlier quoted context omitted.

> We know how neural nets work. We don't know how a specific combination of weights in the net is capable of coherently asking questions asked in a natural language, though. these are the same thing. the neural network is trained to predict the most likely next word (rather token, etc.) — that’s how it works. that’s it. you train a neural network on data, it learns the function you trained it to, it “acts” like the d…

It's like you're describing a compression program as "it takes a big file and returns a smaller file by exploiting regularities in the data." Like, you have accurately described what it does, but you have in no way answered the question of how it does that. If you then explain the function of a CPU and how ELF binaries work (which is the equivalent of trying to answer the question by explaining how neural networks wo…

…except we know what every neuron in a neural network is doing. I ask again, what criteria do we need to meet for you to claim we know how LLMs work?

we know the equations, we know the numbers going through a network, we know the universal approximation theorem —- what’re you looking for exactly?

I’ve answered the “what have they learnt” bit; a function that predicts the next token based on data. what more do you need?

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

#338

Post by FeepingCreature »

dkdcio wrote: post #337
FeepingCreature wrote: post #334

Earlier quoted context omitted.

It's like you're describing a compression program as "it takes a big file and returns a smaller file by exploiting regularities in the data." Like, you have accurately described what it does, but you have in no way answered the question of how it does that. If you then explain the function of a CPU and how ELF binaries work (which is the equivalent of trying to answer the question by explaining how neural networks wo…

…except we know what every neuron in a neural network is doing. I ask again, what criteria do we need to meet for you to claim we know how LLMs work? we know the equations, we know the numbers going through a network, we know the universal approximation theorem —- what’re you looking for exactly? I’ve answered the “what have they learnt” bit; a function that predicts the next token based on data. what more do you nee…

Yes, in the analogy it's equivalent to saying you know "what" every instruction in the compression program is doing. push decrements rsp, xor rax, rax zeroes out the register. You know every step. But you don't know the algorithm that those instructions are implementing, and that's the same situation we're in with LLMs. We can describe their actions numerically, but we cannot describe them behaviorally, and they're doing things that we don't know how to otherwise do with numerical methods. They've clearly learnt algorithms but we cannot yet formalize what they are. The universal approximation theorem actually works against your argument here, because it's too powerful- they could be implementing anything.

edit: We know the data that their function outputs, it's a "blurry jpeg of the internet" because that's what they're trained on. But we do not know what the function is, and being able to blurrily compress the internet into a tb or whatever is utterly beyond any other compression algorithm known to man.

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

#339

Post by nijave »

dredmorbius wrote: post #286
nijave wrote: post #14

Earlier quoted context omitted.

Super charged search on page would also be nice Agents (like a research agent) could also be interesting

Mozilla implementing a search feature which renders Google and/or its advertising capabilities irrelevant is highly unlikely so long as Mozilla is a financial vassal of Google.

Search on page (ctrl-f), not search on internet

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

#340

Post by dredmorbius »

nijave wrote: post #339
dredmorbius wrote: post #286

Earlier quoted context omitted.

Mozilla implementing a search feature which renders Google and/or its advertising capabilities irrelevant is highly unlikely so long as Mozilla is a financial vassal of Google.

Search on page (ctrl-f), not search on internet

Ah, I'd missed that, thanks.
Post reply on HN