Skip to main content
Meta Stack Exchange

Return to Answer

Post Timeline

Copy edits for clarity; add missing words, clarify sentences, etc.
Source Link
zcoop98
zcoop98
  • 11.1k
  • 3
  • 28
  • 62

I have big doubts about this part of presented data:

Yet, at the same time, actual GPT posts on the site have fallen continuously since release

Yet, at the same time, actual GPT posts on the site have fallen continuously since release

It's implied that postposts with a small number of Drafts are occurring less and less occur on sites, so it means that we have less chatGPTChatGPT answers.

I would argue that the only conclusion that can be drawn, is that we have less postposts that are blind copies from chatGPTChatGPT, with no reading or applyingedits to apply basic formatting.

I've executed a small experiment: I've gone to one of my latest answers, copied the question to chatGPTChatGPT, and attempted to create an answer with it'sits output. It took me three drafts, to paste and reformat answers,the answer so that it would be adequate to my personal standards.

I don't have the exact number of drafts required to createdcreate an initial answer, but since it consisted of three rather conservative in length paragraphs, and ready-made code copied from an IDE, I'm pretty sure that it took me less than 6 Drafts to create the answer.

I'm open to the idea that while this study was conducted there were some additional internal indicators of answers being AI made-made that were applied, but the published part, in my opinion, has major flaws onregarding this partpoint on drafts.

I have big doubts about this part of presented data:

Yet, at the same time, actual GPT posts on the site have fallen continuously since release

It's implied that post with small number of Drafts are less and less occur on sites, so it means that we have less chatGPT answers.

I would argue that only conclusion can be drawn, that we have less post that are blind copies from chatGPT, with no reading or applying basic formatting.

I've executed small experiment: I've gone to one of my latest answers, copied question to chatGPT and attempted to create answer with it's output. It took me three drafts, to paste and reformat answers, so it would be adequate to my personal standards.

I don't have exact number of drafts required to created initial answer, but since it consisted of three rather conservative in length paragraphs, and ready-made code copied from IDE, I'm pretty sure that it took me less than 6 Drafts to create answer.

I'm open to idea that while this study was conducted there were some additional internal indicators of answers being AI made were applied, but published part in my opinion has major flaws on this part.

I have big doubts about this part of presented data:

Yet, at the same time, actual GPT posts on the site have fallen continuously since release

It's implied that posts with a small number of Drafts are occurring less and less on sites, so it means that we have less ChatGPT answers.

I would argue that the only conclusion that can be drawn is that we have less posts that are blind copies from ChatGPT, with no reading or edits to apply basic formatting.

I've executed a small experiment: I've gone to one of my latest answers, copied the question to ChatGPT, and attempted to create an answer with its output. It took me three drafts to paste and reformat the answer so that it would be adequate to my personal standards.

I don't have the exact number of drafts required to create an initial answer, but since it consisted of three rather conservative in length paragraphs, and ready-made code copied from an IDE, I'm pretty sure that it took me less than 6 Drafts to create the answer.

I'm open to the idea that while this study was conducted there were some additional internal indicators of answers being AI-made that were applied, but the published part, in my opinion, has major flaws regarding this point on drafts.

Source Link
markalex
markalex
  • 1.8k
  • 1
  • 6
  • 16

I have big doubts about this part of presented data:

Yet, at the same time, actual GPT posts on the site have fallen continuously since release

It's implied that post with small number of Drafts are less and less occur on sites, so it means that we have less chatGPT answers.

I would argue that only conclusion can be drawn, that we have less post that are blind copies from chatGPT, with no reading or applying basic formatting.

I've executed small experiment: I've gone to one of my latest answers, copied question to chatGPT and attempted to create answer with it's output. It took me three drafts, to paste and reformat answers, so it would be adequate to my personal standards.

I don't have exact number of drafts required to created initial answer, but since it consisted of three rather conservative in length paragraphs, and ready-made code copied from IDE, I'm pretty sure that it took me less than 6 Drafts to create answer.

I'm open to idea that while this study was conducted there were some additional internal indicators of answers being AI made were applied, but published part in my opinion has major flaws on this part.