NYC Midnight : Creative Writing & Screenwriting Homepage
Forum Home Forum Home > GENERAL DISCUSSION > Creative Writing Corner
  New Posts New Posts RSS Feed - ChatGPT
  FAQ FAQ  Forum Search   Events   Register Register  Login Login

ChatGPT

 Post Reply Post Reply Page  <1 1112131415 19>
Author
Message
Silverlock View Drop Down
NYC Midnight Regular
NYC Midnight Regular
Avatar

Joined: 22 Jan 2022
Location: Brocéliande
Status: Offline
Points: 352
Post Options Post Options   Thanks (0) Thanks(0)   Quote Silverlock Quote  Post ReplyReply Direct Link To This Post Posted: 22 Feb 2023 at 5:48pm
Originally posted by taaaylor taaaylor wrote:

Clarkesworld just closed their submissions due to being inundated with AI story submissions. http://neil-clarke.com/a-concerning-trend/


My initial flippant thought was, "Oh no!  Who's going to reject my stories now?"

But jokes aside, this is really serious now - more serious than I realized.  Sheer volume of submissions is already a significant challenge for magazines, many of which are operating on a much smaller budget than Clarkesworld.  A great many sci-fi/fantasy mags are legally registered as nonprofits, because using volunteer labor is the only way they can stay afloat.  Some really good magazines might sink because of this new situation, regardless of how good or bad the AI-generated stories are.

And it might get worse.  Currently, Chat GPT does not pass the Turing Test (https://en.wikipedia.org/wiki/Turing_test).  But as the technology continues to be developed, I can't think of any reason to assume that it will definitely never do so.
"The third is the maker's drink, no bottom limit to quality guaranteed, and no top specified."
-- Silverlock by John Myers Myers
Back to Top
Zero11 View Drop Down
NYC Midnight Black Belt
NYC Midnight Black Belt
Avatar

Joined: 29 Mar 2022
Status: Offline
Points: 2914
Post Options Post Options   Thanks (0) Thanks(0)   Quote Zero11 Quote  Post ReplyReply Direct Link To This Post Posted: 22 Feb 2023 at 6:29pm
What percentage of CHAT-GPT involvement do you feel is acceptable? Given we can trust the results? I ran an excerpt from Kate Chopin’s The Story of an Hour through ChatZero and got 1.19% AI GPT involvement. Clearly, that was an error. 

So… what percent would just using an outline generated by CHATGPT get? What percent if you use the generated synopsis as a “prompt”?

Would 12% be acceptable? How about 28%
Back to Top
Suave View Drop Down
NYC Midnight Black Belt
NYC Midnight Black Belt
Avatar

Joined: 25 Jan 2015
Location: Thailand
Status: Offline
Points: 25028
Post Options Post Options   Thanks (0) Thanks(0)   Quote Suave Quote  Post ReplyReply Direct Link To This Post Posted: 22 Feb 2023 at 8:10pm
Originally posted by Zero11 Zero11 wrote:

What percentage of CHAT-GPT involvement do you feel is acceptable? Given we can trust the results? I ran an excerpt from Kate Chopin’s The Story of an Hour through ChatZero and got 1.19% AI GPT involvement. Clearly, that was an error. 

So… what percent would just using an outline generated by CHATGPT get? What percent if you use the generated synopsis as a “prompt”?

Would 12% be acceptable? How about 28%

And what would the result be if whoever is reading your story believes that yours is a AI construct because of said percentages? I can see where people's grades in universities and colleges could be affected by a percentage if the marker did not contact each writer before deciding to ding them some points. 
Back to Top
BarbaraFL View Drop Down
NYC Midnight Black Belt
NYC Midnight Black Belt
Avatar

Joined: 08 Jun 2006
Status: Offline
Points: 3785
Post Options Post Options   Thanks (2) Thanks(2)   Quote BarbaraFL Quote  Post ReplyReply Direct Link To This Post Posted: 23 Feb 2023 at 5:49am
Originally posted by Zero11 Zero11 wrote:

What percentage of CHAT-GPT involvement do you feel is acceptable? Given we can trust the results? I ran an excerpt from Kate Chopin’s The Story of an Hour through ChatZero and got 1.19% AI GPT involvement. Clearly, that was an error. 

So… what percent would just using an outline generated by CHATGPT get? What percent if you use the generated synopsis as a “prompt”?

Would 12% be acceptable? How about 28%

That is really interesting about the Chopin story - I wonder why it coughed up that percentage.

For me, zero is acceptable. An AI-generated outline means the AI is the writer. Writing isn't just the actual words. I'll die on this hill.
Back to Top
Zero11 View Drop Down
NYC Midnight Black Belt
NYC Midnight Black Belt
Avatar

Joined: 29 Mar 2022
Status: Offline
Points: 2914
Post Options Post Options   Thanks (0) Thanks(0)   Quote Zero11 Quote  Post ReplyReply Direct Link To This Post Posted: 23 Feb 2023 at 5:19pm
Originally posted by Suave Suave wrote:

Originally posted by Zero11 Zero11 wrote:

What percentage of CHAT-GPT involvement do you feel is acceptable? Given we can trust the results? I ran an excerpt from Kate Chopin’s The Story of an Hour through ChatZero and got 1.19% AI GPT involvement. Clearly, that was an error. 

So… what percent would just using an outline generated by CHATGPT get? What percent if you use the generated synopsis as a “prompt”?

Would 12% be acceptable? How about 28%

And what would the result be if whoever is reading your story believes that yours is a AI construct because of said percentages? I can see where people's grades in universities and colleges could be affected by a percentage if the marker did not contact each writer before deciding to ding them some points. 

I’m not sure how accurate the AI detectors are yet. Maybe combined with plagierism detectors?
Back to Top
Blue Mass View Drop Down
NYC Midnight Groupie
NYC Midnight Groupie


Joined: 01 Feb 2018
Location: California
Status: Offline
Points: 113
Post Options Post Options   Thanks (0) Thanks(0)   Quote Blue Mass Quote  Post ReplyReply Direct Link To This Post Posted: 23 Feb 2023 at 7:58pm
My thoughts on chat GPT are that it is yet another tool that give the haves an edge over the have nots. Think of it this way: things such as chat GPT have been being used by the few in contest such as this one before the release of chat GPT. And I am sure things like this have been used to get major deals and the people and companies committing to these people have unwittingly degraded the quality of the stories that will be put out. That how craftsmanship dies or makes it more expensive and exclusive to a few willing to put in the effort. 
Back to Top
Suave View Drop Down
NYC Midnight Black Belt
NYC Midnight Black Belt
Avatar

Joined: 25 Jan 2015
Location: Thailand
Status: Offline
Points: 25028
Post Options Post Options   Thanks (0) Thanks(0)   Quote Suave Quote  Post ReplyReply Direct Link To This Post Posted: 23 Feb 2023 at 11:02pm
Originally posted by Zero11 Zero11 wrote:

Originally posted by Suave Suave wrote:

Originally posted by Zero11 Zero11 wrote:

What percentage of CHAT-GPT involvement do you feel is acceptable? Given we can trust the results? I ran an excerpt from Kate Chopin’s The Story of an Hour through ChatZero and got 1.19% AI GPT involvement. Clearly, that was an error. 

So… what percent would just using an outline generated by CHATGPT get? What percent if you use the generated synopsis as a “prompt”?

Would 12% be acceptable? How about 28%

And what would the result be if whoever is reading your story believes that yours is a AI construct because of said percentages? I can see where people's grades in universities and colleges could be affected by a percentage if the marker did not contact each writer before deciding to ding them some points. 

I’m not sure how accurate the AI detectors are yet. Maybe combined with plagierism detectors?

I don't see how any detectors will be of help if there is not a screening of the results with the perp - it will make so much extra work for the teachers. 
Back to Top
NorthernPixels View Drop Down
NYC Midnight Addict
NYC Midnight Addict
Avatar

Joined: 28 Jan 2020
Location: Toronto
Status: Offline
Points: 647
Post Options Post Options   Thanks (2) Thanks(2)   Quote NorthernPixels Quote  Post ReplyReply Direct Link To This Post Posted: 24 Feb 2023 at 1:41pm
I just found this thread.

Friends, if you're a writer who cares about the craft, then you will not use AI for any purpose. There are other tools for helping with spell-check grammar/etc. 

AI generating ideas, writing prose, etc is an insult to yourself, your intelligence, and your ability to create. 
The process of creating is vital to the work. Don't dilute it with this garbage. There are better, more societally beneficial and with more monetary potential for AI/ML than using it to circumvent a primal form of human expression. There's less than zero value in AI media. 

Edited by NorthernPixels - 24 Feb 2023 at 1:42pm
Back to Top
Zero11 View Drop Down
NYC Midnight Black Belt
NYC Midnight Black Belt
Avatar

Joined: 29 Mar 2022
Status: Offline
Points: 2914
Post Options Post Options   Thanks (0) Thanks(0)   Quote Zero11 Quote  Post ReplyReply Direct Link To This Post Posted: 24 Feb 2023 at 3:17pm
Found this game linked on an MIT Technology Review article. I’ll cite both, so you can trust the link. But it is a game to train people to spot AI involvement, where you have to guess if something is AI or human generated. Interesting in how difficult it can be!




Back to Top
Suave View Drop Down
NYC Midnight Black Belt
NYC Midnight Black Belt
Avatar

Joined: 25 Jan 2015
Location: Thailand
Status: Offline
Points: 25028
Post Options Post Options   Thanks (0) Thanks(0)   Quote Suave Quote  Post ReplyReply Direct Link To This Post Posted: 25 Feb 2023 at 9:27pm
May be an image of text that says Hey you goin to sleep? Yes, now shut up Every c in &quot;Pacific Ocean&quot; is pronounced differently
Back to Top
 Post Reply Post Reply Page  <1 1112131415 19>

Forum Jump Forum Permissions View Drop Down

Forum Software by Web Wiz Forums® version 12.05
Copyright ©2001-2022 Web Wiz Ltd.

This page was generated in 0.328 seconds.