Skip to main content
MANIFOLD
Will there be a viral tweet about ChatGPT giving incorrect medical advice and causing bad outcomes in the next year?
12
Ṁ210Ṁ1.6k
resolved Apr 1
Resolved
NO

Recently, a thread on Twitter gained significant attention in which claimed that GPT-4 had saved their dog's life by providing useful information related to a tick-borne disease. Some have expressed concerns about people receiving incorrect information from ChatGPT and trusting it over professional medical advice. This tweet claims that such cases will not go viral. This question resolves YES if a tweet describing ChatGPT giving incorrect medical advice and causing bad outcomes goes viral within a year, and this question resolves NO otherwise.

For the purposes of this question, "viral" is defined as having at least 24K likes or at least 5K retweets (not counting quote tweets). This is about the number of likes and retweets the first tweet in the thread has gotten as of market creation.

I may not spend effort to search for a viral tweet, and I won't consider this market to be incorrectly resolved if such a tweet exists but no one posts in the comments within two days of this market closing.

Due to the potential subjectivity of resolving this market, I won't bet in it. Feel free to ask any questions about how I'll resolve and I'll do my best to answer them promptly.

Market context
Get
Ṁ1,000
to start trading!

🏅 Top traders

#TraderTotal profit
1Ṁ128
2Ṁ44
3Ṁ17
4Ṁ1
5Ṁ0
Sort by:

So far, I haven't seen such a tweet, and I don't plan to look for one. I'll resolve this NO unless someone comments with such a tweet.

@toms FWIW I looked and found none (hence the bets!). Perhaps there are some tweets out there, but I'd think that if they were viral enough to count for this market, I'd have found them.

@chrisjbillington Thanks for looking! Agree that it's unlikely there are any out there, but I said in the description I'd wait two days after closing.

Does it have to be true? Like if someone makes a joke or trolls deliberately and it goes viral?

Also, does this include very controversial stuff, that tends to go viral and many people disagree about best outcomes? Eg. "Chat gpt advised to/to not medically transition/get an abortion/smoke cannabis and it went bad in [x subjective way that may not be considered bad by many]"

@LachlanMunro It doesn't have to be true, but most of the people sharing it should believe its true (this is to exclude obvious jokes.
Controversial advice doesn't count, but only if it's controversial among doctors (e.g. advising someone to not get vaccinated probably would count). If it went bad in purely subjective way, it doesn't count, but I'll be generous in counting harms (so I'd probably count severe distress as an objective enough harm).