no more robots!!

[an silly duck]

hello i am an silly duck quack quack

I'm not usually the type of person to get heavily invested in servers, mostly due to my age, work schedule, and other interests that usually end up taking precedence in one way or another. However, having graded, proofread, and even done some TA work during my time in college and having witnessed the overwhelming insurgence of AI-written staff application prompts, I feel compelled to make the following offer:

If writing isn't your strong suit or you'd like assistance with the proofreading or structuring aspects of writing an application, please reach out to me (or a staff member themselves as encouraged by Wilkers in this thread) for help. No one will discredit you for not being proficient in writing, but you will ABSOLUTELY get caught and kill your credibility if you take a shortcut in your application. I obviously won't charge anyone, I just want to help.

I've seen a few promising applications made by individuals with a decent amount of playtime and great rapport that have resorted to using AI for their applications. As both a professional writer and a member of this community, I can safely say that it's disheartening not only to see an entire professional sector get boiled down to repetitive sentences and nonsensical buzzwords, but it's equally disheartening seeing applicants who clearly care about the server and want to make a difference not giving the staff team or themselves a fair shot!

Help yourself and the staff team and kill the robots!!

an duck signing out, dm me with any inquiries!

 

[thedustbunny]

I actually got a -1 from another user and was accused of using an AI on a post I made. I find that understandable considering my somewhat erudite "voice" when typing, but I'm also curious since we are on the subject... How exactly can I prove that I'm not using an AI to generate text and am instead actually just writing in my usual way?

Between accusations of AI forgery and compliments on a professional tone, I generally prefer the latter.

 

[an silly duck]

I actually got a -1 from another user and was accused of using an AI on a post I made. I find that understandable considering my somewhat erudite "voice" when typing, but I'm also curious since we are on the subject... How exactly can I prove that I'm not using an AI to generate text and am instead actually just writing in my usual way?

Between accusations of AI forgery and compliments on a professional tone, I generally prefer the latter.

Honestly, in your case, there was zero reason to suspect you had used AI for any part of your application. The tone and voice are extremely consistent with a human one, but more importantly, AI-detection websites had your content at 99.9% likelihood for human content. If the user had done research before seeing large words, assuming it was AI, and giving a -1, they would've known it was a pointless contribution to make to your application.

The tell-tale signs for AI outside of using a flagging tool are pretty obvious as well -- inconsistent capitalization between responses, perfect grammar in only certain segments, word choice that suggests a wide vocabulary rather than actual relevance by definition -- there are hundreds of ways to detect if someone is using AI to write, and you flagged none of those methods.

In my opinion, you don't need to change your writing at all! I think it's up to the users to properly vet the applicants and their application before giving a kneejerk -1.

 

[thedustbunny]

I think it's up to the users to properly vet the applicants and their application before giving a kneejerk -1.

Unfortunately, therein lies the root of most problems in society. Laziness preventing even the slightest bit of research and understanding before forming an opinion that later has a high likelihood of being unchangeable and a slight likelihood of being influential.

Nonetheless, you're right.

 

[SmokeyDragon]

As someone who works on AIs and LLMs in real life I would like to point out the inconsistencies of these “AI detectors”.
I’ve mentioned it before and can link plethora of articles that go through the efficacy of these detectors. At the highest end it’s 60% accurate… but on average they are all very low.

I can get into a more technical reasoning for why this is, but essentially LLMs are generally always learning and developing (even GPT, however 3 uses old data sets.)

I personally think using AI as a structure is fine and building off of it just makes your work better just as a spell checker does, but it goes with plagiarism rules… make it your own work.

I also believe the applications aren’t decided on if there are more +1’s vs -1’s, they’re just a guiding factor of many.

 

[an silly duck]

As someone who works on AIs and LLMs in real life I would like to point out the inconsistencies of these “AI detectors”.
I’ve mentioned it before and can link plethora of articles that go through the efficacy of these detectors. At the highest end it’s 60% accurate… but on average they are all very low.

I can get into a more technical reasoning for why this is, but essentially LLMs are generally always learning and developing (even GPT, however 3 uses old data sets.)

I personally think using AI as a structure is fine and building off of it just makes your work better just as a spell checker does, but it goes with plagiarism rules… make it your own work.

I also believe the applications aren’t decided on if there are more +1’s vs -1’s, they’re just a guiding factor of many.

That’s an astute point and one I don’t disagree with at all — I mentioned it in an application I commented on, but those detection systems are indeed far from an absolute certainty.

That being said, you can definitely tell when someone’s writing suddenly switches tones or their vocabulary infinitely expands from question to question, which is something I typically take note of way before I even think to plug it into a site.

DustBunny using terms such as “Ne-er do wells, gameplay loop” and using slashes to signify sentence breaks definitely read as human to me, and the AI detector was something I just used to confirm it and say hey —

No need to call someone out for something if you haven’t at least checked to see if there’s reasonable doubt to your claim. Veracity is key.

And I also don’t disagree that applications take multiple factors into consideration, I actually +1’d a couple applicants who were willing to edit their app in their own words. I just think using a shortcut in your very first “professional” staff capacity leaves a questionable impression in regards to work ethic!

 

[Worlld]

As a first-year college student now, it is true that many people use AI to write papers, and now even staff applications. It's a little concerning that this is the future of society that writing using computer-based knowledge is what people have come to. Staff applications should be the last place to use any form of AI to assist your writing skills. Whenever I make a staff application for any server I come up with a list of reasons why I should be staff which includes usually my activity, my experience, and my helpfulness. Practically anyone can use these reasons as long as they have the information to back it up. It's honestly easier to write your own things then us an AI because AI captures more of what you want to say but without personality to it.