I stopped using ChatGPT and the likes for things such as writing ideas and whatnot

[Xeoz]

If it’s not too much personal info, I’m curious what your job is that requires its use.

Wow combining an ableist root suffix and whining about outmoderned jobs. And crying? You sound like a Republican YouTube troll. I expect better from people here on scribblehub. Well I’ve never done this before but if there’s a way of blocking your bigoted behind I’ll be doing it.

AI Content Research. SEO related stuff. But they're also trying to push me to do some AI content writing stuff, like humanizing AI text. Basically, rewriting it... It may look easy, but it's 23k words a day... around 16-20 articles. A quota type of job.

 

[ElijahRyne]

Something that took me a long time to realize it's incredibly unhealthy to constantly talking to an advanced chatbot. It ruined my writing cycle and oftentimes it will draw you into a worldbuilding rabbit hole or making yourself reading something that's actually not your style to write down.

While there's definitely nothing wrong to use ChatGPT or Claude or DeepSeek or even Proximity (basically any LLMs) to assess and refine or maybe doing further proofreading, there is a tendency to make you suddenly relying heavily on these tools, and soon it'll spread into the moment you just... said things you really want to write, but didn't wrote it down.

Again, there's nothing wrong on using LLM to help yourself in writing, it's just that you gotta use them moderately, otherwise you're as good as procrastinating and not writing anything.

Lastly, whoever read this piece of rant I wrote, thank you so much for taking your time. I know many won't care, but again, I appreciate all the attention you give with my first thread.

EDIT: I think I should move this to Writing General, though I don't know how

At most, I use ai to rate a section of writing. I have a section that I finished writing and ask 2-3 different LLM’s to grade that section using the same criteria. Then I judge if their critiques are relevant or necessary and go from there. This is at the final stages of editing where I have gotten rid of grammatical errors and the like and need to make sure that things are flowing nicely. IMPORTANTLY I do not ask the AI to rewrite any of my work, just areas that I am succeeding and failing at.

Having the AI rewrite a section for you is unnecessary and will harm your ability to improve and will leave you with an average result. Taking the LLM at its word, and not treating its suggestions with skepticism. will harm you ability to judge and discern where you need to improve in your writing, and harm your ability to improve that discernment.

Also having multiple LLM’s looking at the same section will allow you to cross reference the critiques and see common areas of interest. It does not necessarily mean that if two+ AI’s think an area needs improvements that you should improve that area, or improve it in the ways the LLM suggests. In the end they are just tools, you don’t bring a hammer to a construction site and assume it will be the only tool you will need to build a house. Nor should you expect the hammer to do all the work of building the house on its own.

 

[Humanistheart]

AI Content Research. SEO related stuff. But they're also trying to push me to do some AI content writing stuff, like humanizing AI text. Basically, rewriting it... It may look easy, but it's 23k words a day... around 16-20 articles. A quota type of job.

Oh man, that sounds like a lot! Do you ever get tired of it?

 

[Tetrahedron]

By the way, guys, it all began with me laying down an idea that's been stuck in my head to an LLM and see how it reacts.

Oh boy, it's so addictive. It feels like LLM might be the new source of dopamine for writers.

You can shit on me all you want ? but at the very least I'm becoming everyone else's cautionary tale to not lay down to it and JUST. DO IT. And don't talk to an AI/LLM before you get the job done

 

[Xeoz]

Oh man, that sounds like a lot! Do you ever get tired of it?

I mean yeah... I tried those humanizer bots, and it sucked. So I just wrote it from scratch.

 

[SouthernMaiden]

Good for you! Develop your talents .

Buuut, in my opinion the future is this, there will be tiers for content. Movies, books, shows, games ect.

The worst, but cheapest: AI slop. Think netflix where characters always say exactly whats happening, in case viewers are on tiktok. Just worse. Games will be gatcha garbo with AI waifus who talk to you. Books will be nothing but em dashes and " its not x, its y" phrasing again and again. All books will have the exact same tone and voice. Billionaire creeps will destroy all national parks to build datacenters. Coal power plants in our backyards to make power for more slop.
Good, but expensive: human made. Certified by organizations that guarantee that no AI was used. It will cost 10 times more than slop, and creators will have to pay to get their work certified. You'll be considered a snob and a leftist if you like human made content. Secretly, all rich people will laugh at the poors who watch slop. Remember i. 1984 how all the Proles consume AI generated art? Like that.

 

[autumnsugar]

Something that took me a long time to realize it's incredibly unhealthy to constantly talking to an advanced chatbot. It ruined my writing cycle and oftentimes it will draw you into a worldbuilding rabbit hole or making yourself reading something that's actually not your style to write down.

While there's definitely nothing wrong to use ChatGPT or Claude or DeepSeek or even Proximity (basically any LLMs) to assess and refine or maybe doing further proofreading, there is a tendency to make you suddenly relying heavily on these tools, and soon it'll spread into the moment you just... said things you really want to write, but didn't wrote it down.

Again, there's nothing wrong on using LLM to help yourself in writing, it's just that you gotta use them moderately, otherwise you're as good as procrastinating and not writing anything.

Lastly, whoever read this piece of rant I wrote, thank you so much for taking your time. I know many won't care, but again, I appreciate all the attention you give with my first thread.

EDIT: I think I should move this to Writing General, though I don't know how

W mans

 

[Emotica]

The only use A.I truly has in writing, at least for the very near future, is checking continuity, and making illustrations for those of us that can’t afford to pay an artist. A.I quiet literally isn’t made with the memory capacity it would need to even begin to give you anything other than regurgitated information—and the whole point of reading and writing is the creative aspect (for a lot of us anyway)

A.I will miss things if you write at any great length, even simple spelling and grammatical errors/typos. If you actually research how LLM’s work, it’d be really obvious. Imagine having termites in your house, and the exterminator clears 1 room, and then half of whatever else was relevant after that room. It’s not efficient.

Like I said though, for things like checking continuity, it can be great. I had to maintain almost 100 lore entries, and it was so much quicker to check for contradictions and update information with A.I. Yes, I can ctrl+f for a lot of things, but for the lore entries intrinsically tied to 2 or 3 other lore entries—let me just say that I haven’t had such strong continuity until I figured that out. I don’t have to worry about overcooking the lore books, or forgetting anything, because it’s easy to make mini-lore books from your bigger lore books when it’s relevant. I’m saying lore books, but it applies to outlines, canon, etc. I usually generate my notes and continuity into audio format and listen to them while either writing or multi-tasking, jotting down a quick note if something got twisted. (It’s hard remembering who the Queen’s sister’s husband was 100 years ago but it’s super relevant eventually)