Making ai tags mandatory

[magic9mushroom]

"No, I did not mean to write that, I meant what it was the first time, gah you're not listening, ugh, I got to disable it on this part-" then proofreading highlights it, and tries to change it AGAIN.

Click to shrink...

Why don't you just turn it off? First thing I do upon getting a new computer.

 

[Wintercat]

Why don't you just turn it off? First thing I do upon getting a new computer.

Click to shrink...

I usually do, but I don't have a beta-reader so I try and use the proofreading option to at least look through for obvious typos, take a break, and return to read the story properly after a brief break, to catch any obvious typos before I post.

It doesn't always work, something slips past my checking, because I know how I wanted things to go, and I read it wrong, a mistake in my head turning to how I wanted it to be, and I don't catch it. Only to post, and then groan and edit later when a further look after a bit longer reveals "wait, that should be 'her' not 'him'" or something similar. I've also had some occasions with 'shirt' lacking the letter r in it, which doesn't trigger auto-correct, but definitely makes me feel embarrassed releasing a text that has such obvious mistakes in it still.

So its usually disabled, but writing stuff, and proof-reading et cetera, I do use the option to spot the obvious mistakes sometimes. Spelling error in some words can be noticed and corrected easily enough in such cases. I just dislike that instead of just underlining stuff in red, it still tries to rewrite when I edit the text.

I don't use Microsoft Word on casual writing, but LibreOffice, so occasional updates seem to catch me off guard when some option is back to how it 'should be' apparently.

Still, kind of veering a bit off-topic, but just figured making mistakes that awkward is something you expect more from a human author than an AI one, idle amusement. Humans and AI both make mistakes, just different sorts, and figure AI mistakes tend to be more a sampling of stories leading to weird things that shouldn't happen and don't make sense, resulting in odd stories in some attempts. Eh.

 

[NumberSix]

I usually do, but I don't have a beta-reader so I try and use the proofreading option to at least look through for obvious typos, take a break, and return to read the story properly after a brief break, to catch any obvious typos before I post.

It doesn't always work, something slips past my checking, because I know how I wanted things to go, and I read it wrong, a mistake in my head turning to how I wanted it to be, and I don't catch it. Only to post, and then groan and edit later when a further look after a bit longer reveals "wait, that should be 'her' not 'him'" or something similar. I've also had some occasions with 'shirt' lacking the letter r in it, which doesn't trigger auto-correct, but definitely makes me feel embarrassed releasing a text that has such obvious mistakes in it still.

So its usually disabled, but writing stuff, and proof-reading et cetera, I do use the option to spot the obvious mistakes sometimes. Spelling error in some words can be noticed and corrected easily enough in such cases. I just dislike that instead of just underlining stuff in red, it still tries to rewrite when I edit the text.

I don't use Microsoft Word on casual writing, but LibreOffice, so occasional updates seem to catch me off guard when some option is back to how it 'should be' apparently.

Still, kind of veering a bit off-topic, but just figured making mistakes that awkward is something you expect more from a human author than an AI one, idle amusement. Humans and AI both make mistakes, just different sorts, and figure AI mistakes tend to be more a sampling of stories leading to weird things that shouldn't happen and don't make sense, resulting in odd stories in some attempts. Eh.

Click to shrink...

Try using Grammarly, might work better than vanilla autocorrect for you.

 

[Vini666]

I wanted a way to have review comments like on FictionLive. That would help a lot with this AI problem; I can't stand reading a door described in more than eight words anymore, nobody deserves to go through that.

 

[Iridesce]

I wanted a way to have review comments like on FictionLive. That would help a lot with this AI problem; I can't stand reading a door described in more than eight words anymore, nobody deserves to go through that.

Click to shrink...

The door was a very light reddish brown with gold accents.

Damn, depending on how you count it that's actually not eight words.

 

[Wintercat]

Try using Grammarly, might work better than vanilla autocorrect for you.

Click to shrink...

Amusing as the idea is, especially with the topic of this thread, I think I'll pass for now.

I'd sooner deal with the mistakes popping up now and then and having the auto-correct being troublesome till its switched back off again, than play around too much with AI.

I'll admit it is fun watching a friend make images with it, but I've personally felt little desire to play around with the AI tools.

Still, thanks for the suggestion, might have to look on it with some scrap bit of writing that doesn't matter too much to see how that works when visiting a friend who plays around more with the AI, but in general I just try and proofread my work myself after a pause, or have someone help beta-read something on occasion.

 

[Vini666]

The door was a very light reddish brown with gold accents.

Damn, depending on how you count it that's actually not eight words.

Click to shrink...

I step closer and run a hand over the door. It's warm. Smooth, but not perfectly — there's a texture to the engravings. Like it's been handled before. Used. Not ancient, not new. Just… alive.
I thought 8 was a lot, but I realize now it's not that much. But counting, it's 24 words. And it's never just the door, it's a table, a wall, a tree, everything is important and needs a super mega detailed description.

 

[AztecCroc]

I step closer and run a hand over the door. It's warm. Smooth, but not perfectly — there's a texture to the engravings. Like it's been handled before. Used. Not ancient, not new. Just… alive.
I thought 8 was a lot, but I realize now it's not that much. But counting, it's 24 words. And it's never just the door, it's a table, a wall, a tree, everything is important and needs a super mega detailed description.

Click to shrink...

I mean, given the words in that sentence that aren't involved in directly describing the door, it sounds like it's pretty important to the story and worthy of such a description.

 

[Vini666]

I mean, given the words in that sentence that aren't involved in directly describing the door, it sounds like it's pretty important to the story and worthy of such a description.

Click to shrink...

It's just the room of requirement; the problem isn't just this specific case, but that almost everything is described this way. All AI stories are unnecessarily verbose.

 

[Iridesce]

I mean, given the words in that sentence that aren't involved in directly describing the door, it sounds like it's pretty important to the story and worthy of such a description.

Click to shrink...

Actually just door 37 of the dungeon. 48 more to go!

 

[Wintercat]

If there's tension in the air, such as opening the door to the principal's office after being caught doing something, a possibility of serious consequences hanging in the air, the tension might be ramped up a bit, noticing all manner of things while mind is kind of panicking, looking for some distraction.

Just leaving a bathroom and closing the door behind you? Yeah, less words is necessary.

Figure it is a matter of context, and figure an AI defaults to being a bit more verbose even when its unnecessary?

Then again, I've also seen some people write excessive descriptions on things.

.... One story i recall reading, a long while back, was Naruto fan-fiction that was mostly okay but it had one 'weird' Jiraya (self proclaimed 'Super Pervert') as the POV, and it went for 2 entire paragraphs about Tsunade's breasts, and how they were a religious experience to behold, and trying to convince someone else that seeing them is believing. It was over the top and frankly ridiculous... and with Jiraya, it felt in character.

Too bad that story got the axe during some purges, I'd guess mainly because it also included some pretty raunchy sex-scenes in the story.

 

[tacobowl8]

Yeah, this is devolving into general AI topic chatter. Locking the thread as the question has been answered.