Yeah, I don’t think that there’s going to be a viable future in distinguishing between LLM and human text either, certainly not for people who have an interest in defeating it. That being said:

• Some people are just going to use the most-common models.

• Some models are derived from other models and will share statistical similarities.

I still think that trying to detect LLMs is largely going to be a dead-end road, though.

There are also other wrinkles. I’d bet that an increasing amount of not-trying-to-maliciously-impersonate-a-human text out there is going to come from LLMs. Like, okay, Google Translate uses an LLM. I remember listening to an interview with a senior British EU translator (this was back before Brexit). This is going to be a quality human translator. He was saying that machine translation had gotten so good that these days, that now he always did his first pass with Google Translate and just cleaned it up by hand. Okay, so you figure that this is probably widespread, probably even more so now, right? So if most of the EU’s documents are coming from an LLM, even if you could hypothetically detect all LLMs, you probably wouldn’t want to trigger on some of them.

Not to mention that people adopt mannerisms from text that they read or hear spoken. If someone is using, say, automated browser translations, they’re seeing a lot of the world through an LLM, and you’d expect them to pick up characteristics of that.

Nagoxo

@steamdeckhq some technical feedback:
1. You have two cookie banners on your site. One full screen, one smaller afterwards.
2. You have some movie adverts that go full screen when scrolling. And it’s not clear, how to exit this.

Device: iPhone 15 Pro Max, iOS 26.7.1, Safari Browser

Yeah, I don’t think that there’s going to be a viable future in distinguishing between LLM and human text either, certainly not for people who have an interest in defeating it. That being said:

• Some people are just going to use the most-common models.

• Some models are derived from other models and will share statistical similarities.

I still think that trying to detect LLMs is largely going to be a dead-end road, though.

There are also other wrinkles. I’d bet that an increasing amount of not-trying-to-maliciously-impersonate-a-human text out there is going to come from LLMs. Like, okay, Google Translate uses an LLM. I remember listening to an interview with a senior British EU translator (this was back before Brexit). This is going to be a quality human translator. He was saying that machine translation had gotten so good that these days, that now he always did his first pass with Google Translate and just cleaned it up by hand. Okay, so you figure that this is probably widespread, probably even more so now, right? So if most of the EU’s documents are coming from an LLM, even if you could hypothetically detect all LLMs, you probably wouldn’t want to trigger on some of them.

Not to mention that people adopt mannerisms from text that they read or hear spoken. If someone is using, say, automated browser translations, they’re seeing a lot of the world through an LLM, and you’d expect them to pick up characteristics of that.

Ferengi Rule of Acquisition 13

"Anything worth doing is worth doing for money."

Source: Legends of the Ferengi (DS9 novel)

In this episode Uhura makes first contact with a race of sentient insects although actually is the reason why we love.

The Fediverse does require a little effort, as there is no algorithm spoonfeeding you content. Personally, I like that the barrier of entry requires some human effort as it feels like a natural filter for the sorts of elements we don’t want integrating here from platforms like Reddit.

RE: https://mastodon.social/@jimniels/117505288844524425

I have my phone in Do Not Disturb most of the time and have disabled most notifications entirely.

I actually think the entire idea of "phone notifications" is somewhat insane. Like, why do I even want that?? I don't, really.

The number of people on the entire planet who I want to hear from ASAP if they're desperately needing to get ahold of me I can count on one hand.

They called it the idiot filter back in the old reddit days

@mcc @virtuous_sloth No, my argument merely implements the requirements naively with a simple Boolean statement.

Sorry, this is not a maths argument in the slightest. I feel people let that part get out of hand here.

#Serbien, Belgrad (Slavija) jetzt.
(Ich werde euch vielleicht noch mit ein paar Bildern versorgen. Kann inhaltlich aber nichts mehr schreiben. Habe die letzten Tage kaum geschlafen und bin völlig ausgelaugt. Muss das alles erst verarbeiten. Das bedeutet so viel. Es ist schwer in Worte zu fassen.)

#Serbien, Valjevo jetzt.

Die Belgrader*innen haben den Park gegenüber vom Parlament wieder zurück(erobert). Er wurde über 1,5 Jahre lang als Fläche zum Parken von Schlägertruppen missbraucht. Als normale*r Bürger*in hatte man keinen Zutritt und musste neben der Verunstaltung von einem Erholungsraum Umwege in Kauf nehmen.

what I don't get is if this game was written by someone so aggressively cishet why was there a shark plushie on the shelf

@larsbrinkhoff @markhepburn yes. They had to do that for copyright reasons apparently.

Am I that predictable? 😀

I agree, but what do I know? I've only got a master's in UX.