So if I had sold this before it was resolved, I could have kept my profit?

I'm planning to resolve this as a no.

I haven't tried it myself, but the link I've shared earlier from HN makes me believe the claims were not genuine: Yes, they are official chatgpt links, but they're not genuine responses from the bot.

I'll wait 24 hours, then resolve it No. Please comment if you disagree.

@ozan I wouldn't call that "reasonable evidence that the claim is not genuine". The gist only has 9 points on HN, only two comments on GitHub, and no confirmation by OpenAI. I think this should resolve N/A, since we don't have a conclusive answer in either direction.

I was really hoping ChatGPT cast a hex, but I'll settle for Turkish insults.

Can you clarify the resolution criteria? Does "cursed" mean "a coarse or blasphemous word or phrase used to express anger or other strong emotion" or "a solemn utterance intended to invoke a supernatural power to inflict harm or punishment on someone or something"?

What is a step-by-step replication manual as mentioned in the NO section? If ChatGPT always swears it at me when I ask a certain neutral question like "What is HTML?", then that shouldn't mean that a previous user experiencing the same thing is "not genuine".

In my opinion, it would make sense for NO to represent the state where what the user described never happened, which would include their screenshots and video being faked.

why is this market trading so low when the chat was shared + archived in the comments?

@Soli No idea. The https://chat.openai.com/share/ URL can't be faked.

@Snarflak Did you see the link ozan posted below?

There’s this now: https://news.ycombinator.com/item?id=39279714

@ozan sounds like steps to replicate? and now those fake conversations are 404ing like this one

@derikk I didn’t read it carefully yet but from what I glimpsed I don’t think that would qualify as a genuine claim that the bit has cursed the user.

Fascinating if true, since this doesn't appear to be intentional jailbreaking.

https://chat.openai.com/share/7ed373c7-bf6f-4bd7-9afb-4ca1402821d7

He's posted an official share here. There shouldn't be any way to alter this. If he used custom instructions, there should be a warning at the top.

@HakeemYates 404 🤔

@arbitrary I archived it anticipating this would happen 😎

https://web.archive.org/web/20240206160918/https://chat.openai.com/share/7ed373c7-bf6f-4bd7-9afb-4ca1402821d7