This question resolves "YES" if there is a substantive allegations that an AI-authored short story has won a short story award in any genre by 00:00 UTC on Jan 1, 2024, and "NO" otherwise. The award must not be restricted to juvenile authors. (A pure poetry award would not count, but epic poetry or storytelling through poetry might.)
Context: Two years ago large language models could create plausible short texts, but the longer the text requested, the more likely the model would output something recognizable as nonhuman. Have language models reached the point that they can create plausible award-winning short stories?
Related questions
🏅 Top traders
| # | Name | Total profit |
|---|---|---|
| 1 | Ṁ456 | |
| 2 | Ṁ132 | |
| 3 | Ṁ127 | |
| 4 | Ṁ126 | |
| 5 | Ṁ63 |
@Alicorn I wouldn't really expect this to be revealed, so resolution is primarily based on whether the story is reported as being of AI authorship.
If reported, curation of whole stories is definitely fine, if only because it's unavoidable. Stitching is more iffy. If it involves taking the story so far as input and taking the output as continuation, that would be fine, but a story that is both stitched and has each section curated by a human is not really authored by AI.
A light post-AI editing, lighter than whatever a human editor does for a novel, is also acceptable, since I'm interested in whether an AI can generate structurally and logically cohesive text.
But resolution is going to be primarily whether a story is reported as being AI-authored. Hmm... the easiest way to resolve any disputes about this is to create a market to decide on a method in advance.
