Guys, can we please call it LLM and not a vague advertising term that changes its meaning on a whim?
Science Memes
Welcome to c/science_memes @ Mander.xyz!
A place for majestic STEMLORD peacocking, as well as memes about the realities of working in a lab.
Rules
- Don't throw mud. Behave like an intellectual and remember the human.
- Keep it rooted (on topic).
- No spam.
- Infographics welcome, get schooled.
This is a science community. We use the Dawkins definition of meme.
Research Committee
Other Mander Communities
Science and Research
Biology and Life Sciences
- !abiogenesis@mander.xyz
- !animal-behavior@mander.xyz
- !anthropology@mander.xyz
- !arachnology@mander.xyz
- !balconygardening@slrpnk.net
- !biodiversity@mander.xyz
- !biology@mander.xyz
- !biophysics@mander.xyz
- !botany@mander.xyz
- !ecology@mander.xyz
- !entomology@mander.xyz
- !fermentation@mander.xyz
- !herpetology@mander.xyz
- !houseplants@mander.xyz
- !medicine@mander.xyz
- !microscopy@mander.xyz
- !mycology@mander.xyz
- !nudibranchs@mander.xyz
- !nutrition@mander.xyz
- !palaeoecology@mander.xyz
- !palaeontology@mander.xyz
- !photosynthesis@mander.xyz
- !plantid@mander.xyz
- !plants@mander.xyz
- !reptiles and amphibians@mander.xyz
Physical Sciences
- !astronomy@mander.xyz
- !chemistry@mander.xyz
- !earthscience@mander.xyz
- !geography@mander.xyz
- !geospatial@mander.xyz
- !nuclear@mander.xyz
- !physics@mander.xyz
- !quantum-computing@mander.xyz
- !spectroscopy@mander.xyz
Humanities and Social Sciences
Practical and Applied Sciences
- !exercise-and sports-science@mander.xyz
- !gardening@mander.xyz
- !self sufficiency@mander.xyz
- !soilscience@slrpnk.net
- !terrariums@mander.xyz
- !timelapse@mander.xyz
Memes
Miscellaneous
For some weird reason, I don't see AI amp modelling being advertised despite neural amp modellers exist. However, the very technology that was supposed to replace the guitarists (Suno, etc) are marketed as AI.
I think that's because in the first case, the amp modeller is only replacing a piece of hardware or software they already have. It doesn't do anything particularly "intelligent" from the perspective of the user, so I don't think using "AI" in the marketing campaign would be very effective. LLMs and photo generators have made such a big splash in the popular consciousness that people associate AI with generative processes, and other applications leave them asking, "where's the intelligent part?"
In the second case, it's replacing the human. The generative behaviors match people's expectations while record label and streaming company MBAs cream their pants at the thought of being able to pay artists even less.
Wouldn't it be OCR in this case? At least the scanning?
Yes, but the LLM does the writing. Someone probably carelessly copy pasta'd some text from OCR.
Fair enough, though another possibility I see is that the automated training process for LLMs used OCR for those papers (Or an already existing text version in the internet was using bad OCR) and those papers with the mashed word were written partially or fully by an LLM.
Either way, the blanket term "AI" sucks and it's honestly getting kind of annoying. Same with how much LLMs are used.
Scientists who write their papers with an LLM should get a lifetime ban from publishing papers.
I played around with ChatGTP to see if it could actually improve my writing. (I've been writing for decades.)
I was immediately impressed by how "personable" the things are and able to interpret your writing and it's able to detect subtle things you are trying to convey, so that part was interesting. I also was impressed by how good it is at improving grammar and helping "join" passages, themes and plot-points, it has advantages that it can see the entire writing piece simultaneously and can make broad edits to the story-flow and that could potentially save a writers days or weeks of re-writing.
Now that the good is out of the way, I also tried to see how well it could just write. Using my prompts and writing style, scenes that I arranged for it to describe. And I can safely say that we have created the ultimate "Averaging Machine."
By definition LLM's are designed to always find the most probable answers to queries, so this makes sense. It has consumed and distilled vast sums of human knowledge and writing but doesn't use that material to synthesize or find inspiration, or what humans do which is take existing ideas and build upon them. No, what it does is always finds the most average path. And as a result, the writing is supremely average. It's so plain and unexciting to read it's actually impressive.
All of this is fine, it's still something new we didn't have a few years ago, neat, right? Well my worry is that as more and more people use this, more and more people are going to be exposed to this "averaging" tool and it will influence their writing, and we are going to see a whole generation of writers who write the most cardboard, stilted, generic works we've ever seen.
And I am saying this from experience. I was there when people started first using the internet to roleplay, making characters and scenes and free-form writing as groups. It was wildly fun, but most of the people involved were not writers, but many discovered literation for the first time there, it's what led to a sharp increase in book-reading and suddenly there were giant bookstores like Barns & Noble popping up on every corner. They were kids just doing their best, but that charming, terrible narration became a social standard. It's why there are so many atrocious dialogue scenes in shows and movies lately, I can draw a straight line to where kids learned to write in the 90's. And what's coming next is going to harm human creativity and inspiration in ways I can't even predict.
I am a young person who doesn't read recreationally, and I avoid writing wherever I can. Thank you for sharing your insight as well as sparking an interesting discussion in this thread.
Reading is incredibly important for mental development, it teaches your brain how to have the language tools to create abstractions of the world around you and then use those abstractions to change perspectives, communicate ideas and understand your own thoughts and feelings.
It's never too late to start exercising that muscle, and it really is a muscle, a lot of people have a hard time getting started reading later in life because they simply don't have the practice in forming words into images and scenes.... but think about how strong that makes your brain when you can form text into whole vivid worlds, when you can create images and people and words and situations in your mind to explore the universe around you and invent simulated situations with more accuracy... I cannot scream enough how critically important it is for us to exercise this muscle, I hope you keep looking for things that spark your interest just enough that you get a foothold in reading and writing :)
Yup, it's something I myself recently started to realise and have been forcing myself to read things that actually interest me.
While in elementary and middle school every 2 months we had a specific book we had to read and then would discuss it in class and would be graded based on our input.
Reading books and writing essays has been cemented in my mind as a boring chore that is forced upon me. It took years before it even occured to me that reading might be a fun activity, and a couple more before I actively started trying to read again. It's difficult to break away from the mould I've been set to during my childhood, but I'm slowly chipping away at it.
Children SHOULD read, but how can we get them to WANT to read?
I can confirm that a lot of student's writing have become "averaged" and it seems to have gotten worse this semester. I am not talking about students who clearly used an AI tool, but just by proximity or osmosis writing feels "cardboardy". Devoid of passions or human mistakes.
This is how I was taught to write up to highschool. Very "professional", persuasive essays, arguing in favor of something or against it "objectively". (Assignment seemed to dictate what side I could be on LOL.) Limit humor and "emotional speech." Cardboard.
I was taken aback in my first political science course at the local community college, where I was instructed to convey my honest arguments about a book assignment on polarization in U.S politics. "Whether you think it's fantastic or you think it sucks, just make a good case for your opinion." Wait, what?! I get to write like a person?!
I was even more shocked when I got a high mark for reading the first few chapters, skimming the rest, and truthfully summarizing by saying it was plain that the author just kept repeating their main point for like 5 more chapters so they could publish a book, and it stopped being worth the time as that poor horse was already dead by the 3rd chapter.
It was when it hit me, that writing really was about communication, not just information.
I worry about that these days: That this realization won't come to most, and they'll use these Ai tools or be influenced by them to simply "convey information" that nobody wants to read, get their 85%, and breeze through the rest of their MBA, not caring about what any of this is actually for, or for what a beautiful miracle writing truly is to humanity.
That isn't what I mean by cardboard. Persuasive, research, argumentative essays have been taught to be written the way tou described. They are meant to be that way. But even then, the essays I have read and graded still have this cardboard feel. I have read plenty of research essays where you can feel the emotion, you can surmise the position and most of all passion of the author. This passion and the delicate picking of words and phrases are not there. It is "averaged".
I think we're saying a similar thing, but I understand your point better.
I have read plenty of research essays where you can feel the emotion, you can surmise the position and most of all passion of the author.
Exactly! That's what I mean. There's so many subjects I expected to be incredibly dry, but the writing reminded me it was written by a person who obviously cares about other people reading the text. One can communicate any subject without giving up their soul.
(I am always surprised, but I find this in programming books often, haha.)
But that's what I meant by cardboard as well, I think we might be in agreement:
We expect to see a lot more writing that comes across like "This is what writing should look like, right?"
Writing that understands words, and "averages" the most likely way to convey information or fill a requirement, but doesn't know how to wield language as an art to share ideas with another person.
the writing reminded me it was written by a person who obviously cares about other people reading the text.
This is what's missing being discussed in nearly every online argument about AI art that I read online, there are rarely people who make the actual argument that the whole purpose of art and writing is to share an experience, to give someone else the experience that the author or artist is feeling.
Even if I look at a really bad poem or a terrible drawing, if the artist was really doing their best to share the image in their head or the feeling they were having when they wrote it, it will be 1000X more significant and poignant than a machine that crushes the efforts of thousands of people together and averages them out.
Sure there are billions of people who are content with looking at a cool image and think no deeper of it and are even annoyed at criticism of AI work, but on some level I think everyone prefers content made by another human trying to share something.
I know exactly what you mean, I still frequent a lot of writing communities and that "cardboard" feeling is spreading. Most young people who have an interest in writing are basically sponges for absorbing how their peers write, so it's tragic when their peers are machines designed to produce advertiser-friendly ad-copy.
I do agree with your "averaging machine" argument. It makes a lot of sense given how LLMs are trained as essentially massive statistical models.
Your conjecture that bad writing is due to roleplaying on the early internet is a bit more... speculative. Lacking any numbers comparing writing trends over time I don't think one can draw such a conclusion.
Large discord groups and forums are still the proving ground for new, young writers who try to get started crafting their prose to this day, and I have watched it for over 30 years. It has changed, dramatically, and I would be remiss to say I have no idea where the change came from if I didn't also see the patterns.
Yes it's entirely anecdotal, I have no intention of making a scientific argument, but I'm also not the only one worried about the influence of LLM's on creators. It's already butchering the traditional artistic world, just for the very basic reason that 14-year-old Mindy McCallister who has a crush on werewolves at one time would have taught herself to draw terrible, atrocious furry art on lined notebook paper with hearts and a self-inserted picture of herself in a wedding dress. This is where we all get started (not specifically werewolf romance but you get the idea) with art and drawing and digital art before learning to refine our craft and get better and better at self-expression, but we now have a shortcut where you can skip ALL of that process and just have your snarling lupine BF generated for you within seconds. Setting aside the controversy over if it's real art or not, what it's doing is taking away the formative process from millions of potential artists.
BuT tHE HuMAn BrAin Is A cOmpUteEr.
Edit: people who say this are vegetative lifeforms.
Vegetative electron microscopes!
It immediately demonstrates a lack of both care and understanding of the scientific process.
I recently reviewed a paper, for a prestigious journal. Paper was clearly from the academic mill. It was horrible. They had a small experimental engine, and they wrote 10 papers about it. Results were all normalized and relative, key test conditions not even mentioned, all described in general terms.. and I couldn't even be sure if the authors were real (korean authors, names are all Park, Kim and Lee). I hate where we arrived in scientific publishing.
To be fair, scientific publishing has been terrible for years, a deeply flawed system at multiple levels. Maybe this is the push it needs to reevaluate itself into something better.
And to be even fairer, scientific reviewing hasn't been better. Back in my PhD days, I got a paper rejected from a prestigious conference for being too simple and too complex from two different reviewers. The reviewer that argue "too simple" also gave a an example of a task that couldn't be achieved which was clearly achievable.
Goes without saying, I'm not in academia anymore.
People shit on Hossenfelder but she has a point. Academia partially brought this on themselves.
People shit in Hossenfelder much more for her non-academic takes.
Her video on trans issues has made it very difficult to take her seriously as a thinker. The same types of manipulative half truths and tropes I see from TERFs pretending they have the “reasonable” view, while also spreading the hysteric media narrative about the kids getting transed.
She sucks when overextendeding her aura of expertise to domains she's not good in (eg metaphysics and esp pan-psychism which she profoundly misunderstands yet self-assuredly talked about). Her criticism of academia is good, but she reproduces some of that nonsense herself.
People shit on Hossenfelder but she has a point. Academia partially brought this on themselves.
Somehow I briefly got her and Pluckrose reversed in my mind, and was still kinda nodding along.
If you don't know who I mean, Pluckrose and two others produced a bunch of hoax papers (likening themselves to the Sokal affair) of which 4 were published and 3 were accepted but hadn't been published, 4 were told to revise and resubmit and one was under review at the point they were revealed. 9 were rejected, a bit less than half the total (which included both the papers on autoethnography). The idea was to float papers that were either absurd or kinda horrible like a study supporting reducing homophobia and transphobia in straight cis men by pegging them (was published in Sexuality & Culture) or one that was just a rewrite of a section of Mein Kampf as a feminist text (was accepted by Affilia but not yet published when the hoax was revealed).
My personal favorite of the accepted papers was "When the Joke Is on You: A Feminist Perspective on How Positionality Influences Satire" just because of how ballsy it is to spell out what you are doing so obviously in the title. It was accepted by Hypatia but hadn't been published yet when the hoax was revealed.
It is worthwhile to note that the enzyme did not attack Norris of Leeds university, that would be tragic.
It is by no spores and examined!
This early draft for The Last of Us just gets weirder and weirder.
The peer review process should have caught this, so I would assume these scientific articles aren't published in any worthwhile journals.
One of them was in Springer Nature’s Environmental Science and Pollution Research, but it has since been retracted.
The other journals seem less impactful (I cannot truly judge the merit of journals spanning several research fields)
It is by no spores either
"Science" under capitalism.
https://theanarchistlibrary.org/library/paul-avrich-what-is-makhaevism
Thank you for highlighting the important part 🙏
Also, look how people jump to the conclusion that you're either at one extreme side or the opposite extreme side of any issue - usually the worst one - because they spot an element of what you're saying that matches up with some meme they saw, that reduces a complex issue to a clear-cut binary choice between good and evil. Not that memes or AI are bad, people just lazily apply them way beyond their level of precision.
This is why we can't have nice things.