@plexus Amen. Well put ✌🏻

@plexus

Extruded language food product

@plexus LLMs basically generate language pollution.

@plexus any shape you like, as long as it is featureless and smooth — devoid of any nuance, cleverness, or perspicacity.

@plexus LLM create weird crap.

@plexus So true, so on point. Bravo for putting it in such a succinct and easy to understand fashion.

@plexus LLMs so we can all write emails that sound exactly like our least favorite middle manager.

@plexus beautifully said!

@plexus I think I have an issue--unless someone can convince me otherwise--that these "Large Language Models" aren't really "models" at all. "Model" implies some kind of simplified facsimile, a stripped down or scaled version of something. But the machine learning used to create them basically guarantees that their inner workings can never be interrogated, much less used to understand anything about Language itself. They are just more advanced chat bots.

@plexus LLMs produce - by definition - derivative material. Maybe suitable for instruction manuals and the like, but nothing inspirational.

@plexus I’ve been trying to explain this as well as you have. Language is how we create new ideas, dream new solutions - and share them.

@plexus @RevXenoFact You should check out what painters wrote about photography when it first appeared, or what newspaper journalists said about websites.

@plexus May I suggest you read “Figuring” by Maria Popova.

@plexus I am genuinely conflicted about it. I am generally optimistic about technology making things better for humanity in the medium-to-long term (agricultural technology is why so many of us are alive today), and I reject the facile "luddite" dismissals. And yet much of what you write could have been written by, say, a highly trained cavalry officer decrying the advent of mechanized warfare circa 1918. These new tools will take over because it's what humanity does when better tools appear.

@plexus Reminds me of reading LTI, Lingua Tertii Imperii, by Victor Klemperer, on how the NSDAP changed German language and devalued it for their goals (including constant use of invented buzzwords)

https://en.wikipedia.org/wiki/LTI_%E2%80%93_Lingua_Tertii_Imperii

Not sure how well the book works in English translation

@plexus *boilerplate

@plexus
I hate the AI goo. Not because it's AI, but because it's goo.

I also know that this is what calligraphers said about the typewriter.
And it's true, a beautiful calligraphy looks much better than even the most meticulous type set in Helvetica. It's also much less accessible.

Here's hoping the new tools help us reach further in the end, and that the layers of art stick around for additional enjoyment!

@plexus The widespread availability of the tool merely highlights the DEMAND for such crap in the first place.

@plexus same goes for image generators. Their output is always so over the top. I call it visual sugar flash. Nobody stops even to think about why we create those models and for who we want do drive this forward. There is no ethics and no values in these models. But yet they are aiming at those parts of our creative output that ARE driven by values. It feels utterly wrong.

@plexus AI beat chess in the '90s. No one cared, chess spectators are all still most interested in the human players, even 30 years later. These people weren't interested in well played chess. They were interested in the human choices of the players.

Similarly, people don't value language. They value human expression.

@plexus Thoughts to counter your points:

1. Plenty of posts composed by real people leave the same sad taste of unoriginality. I find it sad to come to Reddit every day and read the same useless threads that repeat in style and meaning. Limited repeated vocabulary of online users hurts sometimes.

2. Try local LLMs which vary in style. They are capable of much more varied content than one might imagine initially. With correct prompt, it's able to replicate your post, for example, word-for-word.

@plexus Hang in there, you are heading in the right direction.

Art is supposed to challenge. And good writing is an art.

The idea that written language that challenges you can be found through LLM's and search machines is the ruse.

Aggregators and large platforms have died intellectually a long time ago.

Ofcourse you should take back control. But that word doesn't begin to cut it. It is about you being you. Create your own connections.

Fediverse for the win.

@plexus
what's LLMs?

@plexus I remember reading a good post (I think by @mekkaokereke ) about how language is used by groups to protect against people trying to infiltrate them without a shared background.

It wouldn't surprise me if the same thing happened to defend against LLMs - you won't accept, say, a job application if there aren't at least a couple of uses of 0-day slang or memes. Language evolves fast, it's one of the amazing things about it.

@plexus Language is a tool. It’s a versatile tool, but tool. Not every word needs to be skillfully crafted and sweated out for weeks before it’s committed to the page.

I’m in no position to tell you how to feel about LLMs. Your position is certainly valid and I share it _in some contexts_. LLM is unlikely to write your favorite novel. Not this year, anyway. But it’s probably OK for all the bulshit elsewhere. It can write a decent marketing copy. Results are not stellar but you probably wouldn’t be able to tell the difference from something written by a person. And like it or not there’s a lot of it out there and demand doesn’t seem to shrink any time soon.

Since most of us are geeks around here let me suggest an analogy. Back in the day we used to craft software in machine code. That was to hard so we came up with assembly which was pretty close to machine code but easier to work with. And what we could do with assembly! Read source code of PacMan, or Zork, or Impulse Tracker. Those are works of art.

But then heathens came up with compilers. How those messed up our machine code! All those automatic optimizations… unrolled loops, and constant elimination, and reordering, and speculative execution, and undefined behavior exploits… The blasphemy! The garbage those compilers spit out is unreadable. Will a compiler ever come up with a fast inverse square root on its own?!

Well, it might since it’s us who’s in charge of teaching that optimization to the compilers.

Machine code/language is a tool. It’s primarily use to achieve some goal that is mostly unrelated to the tool itself. It is OK to automate production if the output quality doesn’t suffer for the purpose. Like compilers, LLMs are fine for some use cases.

We still have people who write ingenious assembly and beautiful poetry and we always will.

Let me give you another analogy. In games there’s a notion of skill floor. Effectively, it’s the basis where a player can play the game. We have tutorials in games to ensure everyone has the basic skills in order to not get stuck. There’s also skill ceiling. It’s how many unessential skill you can get before becoming the best player you can be. Games want this interval to be as wide as possible. This would mean that many people can start playing but also they won’t get bored fast as there’s new skills to learn and we tend to be excited about new things.

LLMs are effectively widening this gap with language. Ceiling remains high. But floor gets lower. People who couldn’t write marketing copy now can guide an LLM to produce something acceptable. People who are just starting to learn English and are unsure of their skill now can write an OK blog post on the topic they care about. More people can use language effectively with LLMs. At the same time the value of good prose/poetry remains the same.

I see why you might think the language is devalued, though. Providers of LLMs promise limitless ability to do anything with their product. In that light results are certainly disappointing. But if we look at LLMs and see what they actually are we might get some use of them.

@plexus another thing I was pointed out by a friend is that the generated language has only one emotion - pleasing. If you want to generate content - say neutral - it's very hard to do as it has been trained to be pleasing.

@plexus
We have an unasked for 'feature' in Atlassian where every time you highlight something, even a single word, you get a popup across the screen asking if you would like this reformatted to sound 'business like' or 'casual' or summat...
Goes well with the two dozen AI macros you have to scroll past asking to do translations into pretty much any language...

At our last account meeting I asked them to turn the effing feature off and don't come back till they offered Pirate Speech...

@plexus 'sugary pink language goo' 👏 👌

@plexus AI seems to make it increasingly difficult to distinguish the 'in' of, for instance, in-telligence and in-formation from the 'in' of in-capable, in-competent, in-sane, et cetera.

@plexus ...i feel you but i am optimistic, i think the value of original, written texts will remain high, i even notice that at work when non-original, bland texts are called "chatgpt-generated" while everybody still recognises a truly original human generated text with it's quirks and faults.

@plexus Oh Arne, I agree with you so much. I love the way you express this. 🙏

@plexus

Just struck me -- you need AI to FAIL.

@plexus Well said!

In a sense I think LLMs bring this full circle with machines now being able to autogenerate content to feed other machines (and, oddly, LLMs) - This all makes both Search engines and also the LLMs themselves so useless that I wouldn't be surprised if a few of the organisation methods we invented at the beginning of the web (Webrings, Human curated Directories) will make a comeback just to find meaningful, human content"

@plexus I have some writers who agree with you on this https://epeus.blogspot.com/2010/09/if-google-predicts-your-future-will-it.html

@plexus Thank you for putting it so well. Spot on.

@plexus I worry a bit that LLM can lead to a stagnation of language.

Blog posts, press articles,... etc. from 20 years ago sound a bit different than todays texts.
Press articles from 40 years ago sounds imho already a bit outdated.
Going back 60 years it becomes already quite noticable old fashioned with differences in the lexicon.

But if from now on a significant part of the press corpus becomes LLM generated, feeding only on itself, how can a language change over time anymore?

@plexus yes - and not just language, but meaning and truthfulness - even the concept of truth. Language is supposed to convey meaning - but what meaning can it have if it is generated by a machine that has no concept of meaning or truth?

@plexus openAI and its ilk are the enemy

and not even a GPL-like open source model will help, because they'll just slurp those weights into their training process if it's good enough - we're way past caring about any kind of licensing

it's giant laundering machines, washing off the creative origin of anything it has touched

Edit: the only good thing about the GPTs is that they're fundamentally a dead-end

@plexus I think they will reveal how much of our "language" is only valueable for the performance as well. Cover letters come to mind.

@plexus there are many things that an autistic person can use to stim

Language is my biggie

I'm not talking about loving the sound of my own voice most of the time I'm thinking 'Oh no I have to breathe a little deeper next time' sort of stuff each time I hear myself phonate..

I'm talking about the origins evolutions precise meanings of and translations of and (anti)manipulations of language... And yeah say that moment's right three words and everyone's life that hears them is forever changed

@plexus I have had similar thoughts, particularly from the perspective that it devalues *learning language*.

@plexus Even typos and mistakes are on some level wonderfully human

@plexus It's the equivalent of pink slime. https://en.wikipedia.org/wiki/Pink_slime

@plexus
Dude, get a grip. Language isn't a commodity. You can't run out of words. Someone using words doesn't change your ability to use words.
If you don't like a software, you can ignore it, not get on social media and complain it exists.
The world is going to do whatever the world wants to do, individual opinions won't change it.
You sound like one of those "I don't own a TV, I only read books people."
LLMs exist because people think it's profitable, when it isn't it will go away.

@plexus What I really hate is when email clients suggest a canned phrase as a response. No, don't do me any favors. I've got this, really.

@plexus I wonder, whether ancient people who valued storytelling, memmorised poems with thousands of verses and worshiped Ogmios or other gods of eloquence, would be as easily fooled by the glitter and effortlessness as we are.

I honestly don't know.

@plexus I feel exactly the same about AI art. I used to swoon over an amazing painting. Now it's instantly tainted because I suspect AI...

@plexus @vaurora https://www.youtube.com/watch?v=hHQ2756cyD8

@plexus I "sort of" disagree. LLMs are a search tool, much in the same way dictionaries and thesauruses are. While these two tools have rigor applied to their construction, there is always a layer of subjectivity that is attributed to the human being(s) that compiled them. When all that is taken into account, they can still be used to augment a process that is creative and authentic.So - I only "sort of" disagree because your critique nails all the ways this critical concept is most often lost.

@plexus
Indeed. The magic of human thought itself has been devalued.
@sennoma

@plexus someone said to me, they wanted AI to come in to do work so I can spend more time doing art, not do art so I can do more work.

@plexus yeah! I wrote something a bit like this: https://stegriff.co.uk/upblog/resist-the-abolition-of-the-human-voice/

@plexus Reflect on how this viewpoint is different from arguments about oral speech vs written language (Socrates in Phaedrus), analog written language vs digital, etc. Is this just a local minima before quality expectations shift? Is automated translation or tools to build better ways to learn other languages a higher value?

FWIW I don't find LLMs interesting or compelling either, nor most advertising/marketing, but I'm always trying to construct the counter-argument to my own position.

@plexus doesn’t it just reify another kind of language? That of a well formulated prompt. Le roi est mort, vive le roi!

@plexus Yeah, this is the core issue with any current implementation of "Artificial Intelligence" in the way that Generative AI has pushed. It's goal is not to actually process language or understand it just to look appealing enough to enough people. It's like the same concept as corporate art. .