I've seen the tickboxing or crossing in coloured squares in one or two other threads now.

Can anyone tell me if this is a result of AI output or something available in the new SF format or what please ?

Edited July 3Jul 3 by studiot

I had a similar thought. We’d had a few presentations that were very organized in outline form with similar icons. It felt very Microsoft-word to me (clippy resurrected as copilot) but I don’t have (nor would I use) such software. So I thought it could be a signature of one or more of the BS engines. I certainly take it as a red flag.

🔹 Summary:

Some new members, as well as some established ones, are using AI for a slick look that makes it seem like the poster really did some major work on this idea.

🔹 Why this matters:

This behavior opens the door to:

• lazy interpretations

• bad signal to noise ratios

• ignoring reading even your own citations

🔹 What I think of it:

I think it's spitting in the eye of rigor, study, and reason. Not a fan.

6 hours ago, studiot said:

Can anyone tell me if this is a result of AI output or something available in the new SF format or what please ?

✅ Check boxes are available in the emoji menu

🔷 As are a host of other icons

Thanks guys.

2 hours ago, Phi for All said:

I think it's spitting in the eye of rigor, study, and reason. Not a fan.

Even worse, it is literally just talking to a chatbot. What is the point if there is no person on the other end to talk to?🤖

8 hours ago, studiot said:

I've seen the tickboxing or crossing in coloured squares in one or two other threads now.

ChatGPT generates such lists (that is HTML ul, li tags with custom images).

Every time ChatGPT starts, you have to ask it not to generate this crap. This pisses me off a lot, and I tell him to stop listing the nonsense he writes.

After 1-2 repetitions you no longer have such crap.

17 minutes ago, CharonY said:

Even worse, it is literally just talking to a chatbot. What is the point if there is no person on the other end to talk to?🤖

What if, what kind of people would you want to talk to if you yourself are one of them?

This is pretty good series. AI gains self consciousness and things happen..

https://www.imdb.com/title/tt30444310/

Edited July 4Jul 4 by Sensei

Most forums allow people to do a bit of font/typeface messing about, but it's generally only, er, certain types of poster who feel the need to do much more than the basics. Seems to be their way to shout, which makes them more right^TM or something.

So non-default fonts has long been a box in my crankery bingo card.

Over the last few weeks or so there's been a huge up tick in "extremely formatted" posts; easy to spot, and a new box in my bingo card.

(Edit: closely related, the walls of text now generated. Sheer volume has always been an issue; people who take one small claim on a long journey then think the length of that journey is proof it was worthwhile. I wish they just stop at the first step, get that discussed before moving on.)

Edited July 5Jul 5 by pzkpfw

13 hours ago, pzkpfw said:

(Edit: closely related, the walls of text now generated. Sheer volume has always been an issue; people who take one small claim on a long journey then think the length of that journey is proof it was worthwhile. I wish they just stop at the first step, get that discussed before moving on.)

This is a peeve of mine too, especially when some of those walls of text derive down to, "In the future, we must look for ways to be more efficient, and be willing to embrace them if they help." I was told to stop that crap in no uncertain terms by my eighth grade composition teacher, and I've done my best since then to write meaningfully. The LLMs are SO much noise surrounding a signal.

I'll take this opportunity to point out that yes, bulleted lists with emojis are a feature of LLM output, and yes, LLMs have a distinctive writing style. Oddly enough this is one of my research areas now, collaborating with an English professor who knows how to quantify various features of writing style.

We wrote a paper (free preprint) where we generated a corpus of LLM and human writing. ChatGPT, for instance, really loves participial phrases and an information-dense style: lots of nouns, noun phrases, and nominalizations (things like "development", where the verb "develop" is turned into a noun). They're features that display authority.

More recently I looked at formatting, and while I don't have full results, the more recent models seem to particularly love using bold, bullet points, and sub-headings. I should add some code to count emojis.

15 minutes ago, Cap'n Refsmmat said:

I'll take this opportunity to point out that yes, bulleted lists with emojis are a feature of LLM output, and yes, LLMs have a distinctive writing style. Oddly enough this is one of my research areas now, collaborating with an English professor who knows how to quantify various features of writing style.

We wrote a paper (free preprint) where we generated a corpus of LLM and human writing. ChatGPT, for instance, really loves participial phrases and an information-dense style: lots of nouns, noun phrases, and nominalizations (things like "development", where the verb "develop" is turned into a noun). They're features that display authority.

More recently I looked at formatting, and while I don't have full results, the more recent models seem to particularly love using bold, bullet points, and sub-headings. I should add some code to count emojis.

It has always seemed to me the output of these things reads like a bad undergraduate essay, full of padding and long words to sound impressive. The sort of thing my tutor at uni used to really hate. Interesting to see you and others are starting to quantify the characteristics of their style.

On this forum the usual giveaway is a post that suddenly reads like a long piece from a textbook, quite different from the normal style of the person posting it.

9 minutes ago, exchemist said:

On this forum the usual giveaway is a post that suddenly reads like a long piece from a textbook, quite different from the normal style of the person posting it.

"We were sitting around the table talking about evolutionary pressures when this guy pulled out his podium and started a slide show!"

17 minutes ago, Phi for All said:

"We were sitting around the table talking about evolutionary pressures when this guy pulled out his podium and started a slide show!"

Ahem, yes. Or a sudden 8 para screed full of palaeontological terms about the Cretaceous…..

15 hours ago, exchemist said:

Ahem, yes. Or a sudden 8 para screed full of palaeontological terms about the Cretaceous…..

Titles, spacing and length are the usual giveaways.

When asked for references they produce a list of tentative, barely relevant or just do not exist citations.

6 minutes ago, pinball1970 said:

Titles, spacing and length are the usual giveaways.

When asked for references they produce a list of tentative, barely relevant or just do not exist citations.

Oh yes we had that on the recent not-really-a-creationist-honest-guv thread: a load of "references" without the detail needed to look them up and which, on researching them, seemed not to exist.

Edited July 6Jul 6 by exchemist

19 minutes ago, exchemist said:

Oh yes we had that on the recent not-really-a-creationist-honest-guv thread: a load of "references" without the detail needed to look them up and which, on researching them, seemed not to exist.

Yes. "Evolution has purpose and here are all the papers I have not read to prove it."

As most did not exist.