Top
Best
New

Posted by mukmuk 5 hours ago

AI content is everywhere on social media, especially LinkedIn(www.pangram.com)
133 points | 119 commentspage 4
PacificSpecific 4 hours ago|
Honestly I can't tell the difference with LinkedIn. Feels the same as it's ever been
volkk 4 hours ago||
Was just thinking today, -- happened to login to LinkedIn, open it up and the entire front page is just AI slop being applauded and liked with people seriously interacting with it as if it's somebody didn't just shit it out in 20s with zero effort. The whole thing needs to die so badly.

On Instagram, I'll get fed "real" content, but you read the description and it's this giant 3-4 paragraph thing that I don't bother to read because I know with certainty that it's AI slop. Before AI, the descriptions of sports videos or meme videos were 2 sentences, now they're entire theses.

The only people left reading this crap are people that still haven't caught up with the concept of AI slop

mattas 5 hours ago||
It's like a burgoo [1]. A steaming cauldron of community slop.

[1] https://en.wikipedia.org/wiki/Burgoo

noumenon1111 4 hours ago|
Hey, at least a bowl of burgoo tastes good, unlike some community slops.
javier123454321 5 hours ago||
The question is not whether something is AI generated. That's the default state now. Question whether it is human, the economics are exceedingly in the favor of this new normal. https://javiergonzalez.io/blog/the-economics-of-slop/
hasteg 5 hours ago|
Same here, I hate it. Instead of just reading I find myself also dedicating brain power trying to decide if it's worth reading or not based on the first few sentences... if someone can't put in the effort to actually write something themselves I absolutely do not want to put in the effort to read it.
yegle 4 hours ago||
Now let's stir the shit further, on LinkedIn group the posters of AI content by country/company.
estetlinus 4 hours ago||
LinkedIn has become an AI-slopped wasteland. It’s like the opposite of when boomers found Facebook, which was the weirdest melting pot of zero-integrity posts and comments.

Now we have these tech-savvy people generating worthless images and producing generic, emoji-infested takeaways.

happywanderer 4 hours ago|
I read through LinkedIn posts and it's AI slop all the way down, it's horrendous. Every post is either written entirely by AI or mostly written by AI.
ButlerianJihad 3 hours ago||
LinkedIn is really the first place I experienced LLM-generated content at scale. I did not know it at the time. I just figured, because the corporate world tends to adopt standards/templates and coalesce around a certain lingo, that this was the way people were posting to LI social media. But indeed as I later found out, there were many classic LLM tells built into all those posts.

Of course, I invariably found those LLM posts to be vacuous and pointless and so it was very easy to begin skimming right past them as I figured out they were cut from the same cloth.

To LinkedIn's credit, though: in 2020 I landed "my pandemic job", a dream job that was 100% remote (thanks to the lockdowns) and had a 100% flexible schedule and I had a really great time with the enjoyable work I was doing.

This job was somehow landed through LinkedIn. To this day, I cannot recall how or when I applied to the company. But I know I was filling out "1-Click Apply" forms there, and I had splashed up my involvement with the community college (though I would not actually graduate from college for another 3 years) and eventually a recruiter telephoned me to ask if I was interested. And as I was juggling some crazy issues in my personal life along with the pandemic lockdowns, I had to assure the recruiter multiple times: yes I'm interested, no please do not hang up, yes please let us continue with your process!

And it was that sort of tenacity that landed the initial job, and helped me hang on to this employer through M&A and multiple job-role changes. They formally terminated me about 4 times, but I was hired 5 times so it sort of evened out, I suppose.

If it were not for LinkedIn or community college, that recruiter never would've found me. I never would've got that job. My life would be so different, especially my pandemic-lockdown life! So grateful.

Ironically, my employer (EdTech industry) began to 100% embrace LLMs for their students and even offered a front-page LLM for students to ask about their homework, and other assignments. It was definitely crazy times for us, as we were the ones tasked with detecting plagiarism and other types of cheating in that homework, while students were being actively encouraged to tap into LLM-based resources for answers...

subygan 4 hours ago||
almost all platforms are like this now. Every active player in each of these platform is trying to get the most eyeballs in their content / profile. and the silent scrollers are not contributing anything else anyway.

We've societally come to the consensus that, we want to reward a race to the bottom slop. passive scrollers by not doing anything about it, active posters by contributing to it.

but there is no way else to win in this game.

A friend of mine writes the most human curated thoughtful newsletter about AI, spending 100 hours. and maybe 200 people know of its existence.

dvt 4 hours ago||
Pangram doesn't work, and I wish people would stop treating it as gospel (but the AI/anti-AI grift is real). Here's a fun paradox: I can literally tell ChatGPT: "Say X" and it will say "X"—so that's a case where content is both AI generated and not. What if it changes a few words? Moves some sentences around? Where does something go from human- to AI-generated? (This is the classic Sorites paradox.)

Pangram tries to look for common patterns (rule of three, em dashes, etc.) but these are heuristic methods and not to be taken as gospel. There is no provable method to make a distinction between AI and human-generated other than the fact that AI-generated text tends to reek of pseudo-intellectual undergrad with a thesaurus.

wgd 4 hours ago||
Pangram does work, in the specific sense that when it says something was AI authored it is vanishingly unlikely that it was written by a human (who was not deliberately trying to write like an AI), and IMO getting people to recognize that we actually do have a decent solution in this space now is pretty important if we want the Internet to remain a place for humans and not just bot swarms.

> rule of three, em dashes, etc

You appear to be misinformed about how Pangram specifically works, it is not based on pattern detection of that sort. I recommend reading their whitepaper, it's a pretty understandable explanation of exactly how they trained their classifier.

timpera 4 hours ago|||
I just tried it, created an account, and wrote a few sentences about how my day went. These sentences got classified as AI assisted, so clearly their classifier doesn't work that well.
dvt 4 hours ago|||
> Pangram does work

It's trivial to see how many people think Pangram is absolute trash[1] (because it is).

> You appear to be misinformed about how Pangram specifically works, it is not based on pattern detection of that sort. I recommend reading their whitepaper, it's a pretty understandable explanation of exactly how they trained their classifier.

I did read their paper (which is, by the way, very scant on details), and they trained their classifier in the laziest way possible: here's a chunk of "human-written" text and here's a chunk of "AI-written" text, put them in the right bucket, and do this a zillion times. Literally zero sophistication. Also: what do you think "pattern recognition" is, if not a "classifier"?

[1] https://www.reddit.com/r/academia/comments/1rm11rs/pangram_c...

_dwt 4 hours ago||
[dead]
dukeofdoom 4 hours ago|
It's still pretty bad at pixel art, and just has this generic look visually. I remember watching this video of this indy game developer that tried to hire an artist for some visuals for cover art for his game, and kept getting sent AI generated stuff by scammers. Finally he did find a real artist, and the cover was really good. But expensive.
More comments...