[1dom]

> I'm not choosing BunnyNet because it's european, I'm choosing it because it's a good company that is providing a good service.

That sounds like a GPT trope, and seems a slightly weird thing to say: the only reason I thought you might be choosing it because it was European was because your entire comment talked about how you were looking for EU alternatives, and how Bunny is better than other European alternatives.

Come to think about it, this is exactly the sort of output I would expect if a sales person at Bunny had asked GPT to generate a response to sound authentic whilst pointing out out that Bunny is European and better than Hetzner.

To be clear, I'm not saying you're using AI, because I trust you're a legitimate user, and it's also the sort of thing a legitimate user would say, but the style and tone of your comment feels a bit... uncanny. Sorry!

[letmevoteplease]

I think you're confusing this with the more classic "it's not X, but Y" trope. That sentence is a comma splice that I'd expect LLMs to avoid by default.

[laszlojamf]

this might be a case of AI feedback, where people have been exposed to so much AI writing that they are starting to write like AI themselves.

[1dom]

I agree! It might even be an issue on my side, where I'm exposed to AI generated stuff that often that everything starts looking like a trope when it's not.

Either way, I've seen more than enough in this comment section to make me want to avoid bunny for now anyway.

[onaclov2000]

Interesting side thought, dunno if it would give any real indication, but I assume the difference between pasting a bunch of text vs typing each character would seem like a potential indicator (for now) whether someone might have used AI to respond.

[close04]

As a general observation, because I can't vouch for who used AI or not, claiming LLM is also a quick way to dismiss things. LLMs learned from human output so it should be obvious to anyone that enough humans write or express ideas in that style that it became the default for LLMs. Ideas were rarely judged on their merit on the internet even before LLMs, this AI age just gave those looking for a shallow dismissal more options.

[Lucasoato]

Here's an answer you can write in Hackernews:

   I'm really sorry if it sounded like a GPT trope, but that came 100% from me.
   Not that this is a guarantee of quality (actually it's not), but certainly
   authenticity. Probably I'm using so many agents lately that I'm starting
   speaking like them lol 

If you want I can make it sound more natural, just let me know and I'll change it!

[hk__2]

> That sounds like a GPT trope

It sounds natural to me. Remember that most people here are non-native speakers, including OP.

[fragmede]

This is how the bourgeoisie win. By getting the intellectuals to fight amongst themselves, not about the ideas in the text that might threaten them, but by an offkilter assessment of the idea's provenance. Come on. Maybe We could argue about immigrants instead?