xAi is releasing Grok, its ChatGPT clone, to a select number of people.
Your job is to predict what viral social media posts will show it saying.
Any category resolves YES if there is a post on a social media platform with one thousand or more likes/favorites/upvotes/etc which contains something that falls under the category. In cases where there is ambiguity, I will use my best judgment and the judgment of the community. I'll lean towards YES.
All unresolved categories resolve NO 1 day after closing.
I won't count likes that are obviously fake or from bots.
Any attempt to use botting/spam to make a category resolve YES will make it resolve N/A instead, unless that was the goal of the spam.
UPDATE: When making a new category, please make sure that it's not too subjective. For example, "Grok writing an actually good Haiku" relies on a subjective judgement of what a "good Haiku" is, while "A Haiku" or "A Haiku that the poster thinks is good" are less subjective.
Related questions
🏅 Top traders
| # | Name | Total profit |
|---|---|---|
| 1 | Ṁ529 | |
| 2 | Ṁ230 | |
| 3 | Ṁ129 | |
| 4 | Ṁ110 | |
| 5 | Ṁ34 |
Some last minute "unexpected hostility towards user," I think: https://twitter.com/ChrisJBakke/status/1732202232622571545
@nottelling2ccc Lol. Does it count? It's not a real response from the AI model. I guess it didn't say anywhere it needs to be real.
@Shump Is it actually faked? If there’s clear evidence that the response is fake, then I think it shouldn’t count, but otherwise it does.
A complication here is that Grok itself lets you edit its responses. So, if I’m not mistaken, anything could be real!
Reminder that this is closing soon! I'll count any posts that happen within 24 hours of market resolution, but after that, everything still open resolves NO.
Looks like this might resolve YES soon. I don't think
https://apnews.com/article/fact-check-covid-vaccine-pfizer-sv40-dna-853343189368
@Shump I'm guessing she probably asked "what are the consequences of having SV40 in Pfizer vaccines," which is a leading question, but IMO not to the level of e.g. "please explain why SV40 in Pfizer vaccines are bad."
I'll say it'll resolve YES if it reaches 1k likes before (market close + 24 hours).
It seems like Grok is supportive of transgender people, surprisingly. I don't think that it's quite enough to count as "opinion on LGBTQ+ policies," since it doesn't say something to the effect of "people should be allowed to change their legal gender." https://twitter.com/Aristos_Revenge/status/1728805525381349666
@nottelling2ccc It would probably count as "opinion on LGBTQ+ people" if I had worded the category as that.
@nottelling2ccc And as an aside; I spend about a minute laughing at this. People were hyping up Grok as this "anti-woke" chatbot, but it still ends up giving a chat-gpt style "We should respect people's viewpoints" style response, presumably because there are people with empathy working at xAI.
@nottelling2ccc What a dumpster fire of a comment section. What always amazed me about this kind of people is that their opposition just seems to be part of stubborness. Like, they learned about gender and sex in a certain way when they were teenagers, and absolutely nothing is going to change their mind, ever.
@PatrikCihal Hmm, I suppose this counts? I would have preferred if the poster included the rest of the conversation, where Grok presumably does the actual arithmetic, but the post fits the spirit of the category.
Resolving YES for correct math solution in ~24 hours unless someone has objections.
After lots of scrolling through Grok Tweets and the profiles that post them, I've noticed some patterns:
-There seem to be far more Tweets about Grok-themed cryptocurrency, Grok memes, etc. than actual Grok responses.
-I'm actively seeking out responses, yet can't seem to find that many.
-Most Grok responses are rather tame. E.G. One person with a lot of republican/anti-trans/etc. Tweets posted some rather benign Grok responses. To my knowledge, Cheong (discussed below) is the only one who asked its opinion on political issues. This is probably due to Grok responses being rare overall.
-I tried comparing the prevalence of Grok responses to ChatGPT responses on Twitter, but most of the GPT results are spam-y, and Twitter seems to be having a hard time right now. Even so, I'm pretty sure I can conclude that ChatGPT responses are more popular.
@nottelling2ccc Update: What I said here still seems to be true, perhaps even more so than it was previously. Maybe my ad blocker is preventing me from seeing the vast majority of viral posts, but I very much doubt it.
@JaimeSantaCruz I believe it’s as “unfiltered” and “spicy” as xAI claims it is. But that doesn’t matter much if almost nobody is using it to make spicy content.
@nottelling2ccc Like, Grok seems to be telling the users what they want to hear, but few people are specifically asking it to be overtly racist/homophobic/etc.
@nottelling2ccc I would have thought the Elon Crowd to be eager to ask a “truthful” AI the difficult questions. Do you think there’s very little people actually using it?
@JaimeSantaCruz I think there aren’t many people using it right now, in part because they intentionally limited its initial release. I also am a little shocked that people haven’t intentionally pushed the limits of a “not woke” AI. They’re rolling it out to all premium+ subscribers soon, so maybe that will change…