AI Community Notes on X — Fact check any tweet thread or article
Amazing work team! Wrangling the Twitter API AND building "Perplexity on steroids" is not trivial! Thank you all for continuing to push for more information transparency and giving us tools to engage critically with the onslaught of slop, whether AI or human made. 🫡
@madison_scott1 It's available to everyone and free! We are paying Twitter for API access so there are some rate limits/maximum tweets per hour and per day.
Excited to see how it helps people to overcome misinformation and find truths
Excited to see this live on twitter! A great way to fact check the content and increase the transparency. 🚀
@jesus_porter From @alexcdot: We follow a bit of Wikipedia’s approach when vetting sources.
I love te concept. How does it deal with tweets that are rhetorical or satirical in nature?
Edward and the team at GPTZero continue to innovate and show that they are the leaders in identifying machine vs human content. With the release of @GPTZero factcheck they are turning their attention to the tsunami of disinformation on X. Since X and now FB rely on community notes to be the only fact-checking on their services, @GPTZero can make the community approach more than window dressing
Can the AI recognize when a tweet has already been fact checked by another source like X’s existing Community Notes or does it start its analysis from scratch every time?
@madelyn_wright Try it out! I've seen it turn around a post in less than 1 minute, but sometimes it takes 5 minutes.
Origin By GPTZero
Hey product hunters!
Edward here, founder of GPTZero 👋 Two years ago, we launched a viral AI detection app; now we’re back with a mission to reduce AI slop and misinformation on the internet. Thank you for visiting this launch! 🚀
We’re really excited to introduce user-initiated community notes on X. Simply tag any post or thread with @ GPTZeroAI fact check, and our AI engine will research the tweet and generate a community notes reply for everyone to read.
To power this, we built our own version of Perplexity on the backend; our pipeline categorizes claims in the articles, searches the internet and academic databases for key sources, sorts the recent articles by relevancy and credibility, and then incorporates and cites it in an AI-generated community notes summary.
Common things we’ve heard from people that have wanted this:
- my X feed has been taken over by DOGE content
- IDFK whether these posts are true or not
- Did Mark Zuckerberg fire all his fact-checkers?
- Community notes were really good on X, but where did they go?
If any of these sound like you, please give us a try!
Just tag 'GPTZeroAI' on X, and write ‘fact check’
It’s our first time integrating with X’s new API and building a Perplexity clone, so please give us as much constructive feedback as you'd like. With your help, this product could make a real difference towards restoring truth online.