FINALLY! An article that doesn't call LLM's "AI."

[Maxmars]

I was perusing the internet's finest techno-news press-release journalism, and I ran across an article of interest to me... you see I hate all things "advertisement."  It's a deficit of my charatcer... so anything that works against me being subject to someone else's desire for my money is worth looking into...

How LLMs could revolutionize ad blocking

(First of all... how very brave to buck the "trend" and NOT say "AI.")

I was amazed at the lack of pro-"AI" journalism, and the presence of honest-to-God reporting....
it was refreshing.... here was another opportunity to mischaracterize the "new" AI...
but the author insists on being specific... how rare...

In the short article, the author points out that LLMs provide powerful new ways to combat the
trickery and abuse that consumers face. 
 

“…LLMs allow us to move beyond simple pattern matching and actually understand the meaning of web content. This opens up a completely new, semantic approach to filtering. And second, LLMs give us rapid prototyping. Ideas that used to take months of engineering can now be tested in a matter of hours. While there are still practical challenges to solve, this new approach allows us to rethink what’s possible in the world of content filtering.”

I notice how he doesn't proclaim this a wonder of "AI"....

The "make-believe AI" honeymoon is over?   We can only hope.

[UltraBudgie]

Well that is interesting; I enjoyed looking at the author's fiction books. Unfortunately he doesn't maintain the AI-vs-LLM distinction in his other articles.

But think about this: the premise is to feed the text of everything your browser presents you to read through an LLM, which will semantically analyze for message and meaning. Ostensibly to check for "nasties" such as advertising and helpfully stop them for you. But such a process generates a nice little psychologically-tokened data-stream of all the information that goes into shaping your world view! Do you think there's a chance that could be gobbled up and monetized into the big-data machine? What are the chance this is yet another "free" service that you end up paying very dearly for?

[Kwaka]

Leaves a few implications of where AI is going in dealing with the digital and real world challenges. Are the ad provides also using AI on their side to add up the data collection?

As AI progresses, will it help or hinder another lockstep covid misinformation campaign? Is a big, expensive job compiling and filtering all the training data. So far all the big players are still on the corporate agenda.

[Kwaka]

Google's Digital Fingerprinting
 

 
Google has announced a significant shift in its online tracking policies, set to take effect on February 16, 2025. The tech giant will now allow advertisers to use digital fingerprinting, a method of identifying and tracking users across the internet

The digital landscape is turning into a jungle out there.

[quintessentone]

I was perusing the internet's finest techno-news press-release journalism, and I ran across an article of interest to me... you see I hate all things "advertisement."  It's a deficit of my charatcer... so anything that works against me being subject to someone else's desire for my money is worth looking into...

How LLMs could revolutionize ad blocking

(First of all... how very brave to buck the "trend" and NOT say "AI.")

I was amazed at the lack of pro-"AI" journalism, and the presence of honest-to-God reporting....
it was refreshing.... here was another opportunity to mischaracterize the "new" AI...
but the author insists on being specific... how rare...

In the short article, the author points out that LLMs provide powerful new ways to combat the
trickery and abuse that consumers face. 
 

I notice how he doesn't proclaim this a wonder of "AI"....

The "make-believe AI" honeymoon is over?   We can only hope.

The media I watch seems to be including both, as in they are saying LLM...AI. I don't know if it will become an educational thing to include the LLM beforehand. You may be right that it is a grandfathering type of terminology to do a 'code switch' like they did with Twitter...X.

[imitator]

I was perusing the internet's finest techno-news press-release journalism, and I ran across an article of interest to me... you see I hate all things "advertisement."  It's a deficit of my charatcer... so anything that works against me being subject to someone else's desire for my money is worth looking into...

How LLMs could revolutionize ad blocking

(First of all... how very brave to buck the "trend" and NOT say "AI.")

I was amazed at the lack of pro-"AI" journalism, and the presence of honest-to-God reporting....
it was refreshing.... here was another opportunity to mischaracterize the "new" AI...
but the author insists on being specific... how rare...

In the short article, the author points out that LLMs provide powerful new ways to combat the
trickery and abuse that consumers face. 
 

I notice how he doesn't proclaim this a wonder of "AI"....

The "make-believe AI" honeymoon is over?   We can only hope.

Sounds like you accidentally found the unicorn of tech journalism, lol.

But unfortunately...  most people will always call this stuff AI.
They don’t know what an LLM is and they don’t care... but, maybe one day. 

[DontTreadOnMe]

I was perusing the internet's finest techno-news press-release journalism, and I ran across an article of interest to me... you see I hate all things "advertisement."  It's a deficit of my charatcer... so anything that works against me being subject to someone else's desire for my money is worth looking into...

What's an LLM?
It's just so hard to keep up with all the acronyms, but the source should explain.
I know it could look it up.  But..... why should I have to do that?
Lovely land mines?
Lunar landing mayhem?
Lucky iost marble?
 

[Maxmars]

The point is that when - chasing down "AI" bigfoot - they engineered something called a "Large Language Model" - an LLM.... a fantastic development in "language synthesis" - the ability to 'provide output in 'human' language.

Marketers took it to the world as "Artificial Intelligence" - which it is most certainly NOT - 

but every talking head and every press-release 'insists' by default that this is a "machine thinking."

SO every time I hear about amazing AI... and I look... it's just an LLM with massive algorithms making it 'sound good.'... but not actually intelligent, or even cognizant, in anyway... fully NOT "intelligence."

This article I shared actually talks about LLMs and how they can use them now...

Instead of propping up an investment scam... and simultaneously proclaiming what would be a pinnacle achievement of any species... creating a new independent intelligence.

[chr0naut]

What's an LLM?
It's just so hard to keep up with all the acronyms, but the source should explain.
I know it could look it up.  But..... why should I have to do that?
Lovely land mines?
Lunar landing mayhem?
Lucky iost marble?
 

LLM stands for "Large Language Model", but I would posit that it should more accurately be "Ludicrous Lying Marketing" or "Loads and Loads of Money".

And, honestly, the algorithm won't suddenly get different if you run it on bigger and bigger hardware. All the scrambling around for bigger and bigger datacenters might, possibly, maybe, have payout in the future when they run better algorithms and have better training filters, but they are a scam to get more and more funding at the moment.

It'd be a lot cheaper to ask things of one of the many subject matter experts that hang around at universities and in R&D departments around the world. Imagine if someone thought of a better way to link-up and use 'Natural Intelligences', like some sort of specific peak-academic social network that the rest of us could tap into for answers?