Elon Musk's AI 'Grok' aggregates fake news posted by users such as 'Iran has launched an attack' and messes with official trends
Elon Musk founded the AI startup '
X's AI chatbot Grok created a fake trending headline about Iran attacking Israel | Mashable
https://mashable.com/article/elon-musk-x-twitter-ai-chatbot-grok-fake-news-trending-explore
Investors in talks to help Elon Musk's xAI raise $3 billion, WSJ reports | Reuters
On April 4, 2024, X updated the 'Explore' tab, which displays trends and search screens, and added a feature that displays 'Grok-generated easy-to-understand summaries of trends' in a news-like format in addition to the top related posts when clicking on a trend. However, it was discovered that Grok was summarizing unfactual information and displaying it as if it were a real event.
DEMO: Updated demo of the upcoming explore page.
— X Daily News (@xDaily) April 3, 2024
pic.twitter.com/4KeFuIyZpT pic.twitter.com/tBGiGxBP9n
On April 4, 2024, a headline trending on the site read 'Iran attacks Tel Aviv, Israel with heavy missiles,' but at the time of writing this article there was no evidence of this and it was deemed fake news.
When the headline was clicked on, numerous posts from verified accounts spreading fake news were found. Mashable, an overseas media outlet, said, 'We believe that X's algorithms, having noticed that fake news was spreading, created the trend, and Grok generated summaries that looked like facts.'
This isn't just the Iran-Israel issue, Grok has also been reported to treat jokes as real articles. The following post shows a summary such as 'NYPD mobilizes 1,000 police officers to prevent earthquakes, and even considers deploying Robocop,' which is obviously not true and is a compilation of jokes posted by users.
This is Twitter's trending news section, powered by Elon Musk's Grok AI. It pulled in a bunch of joke tweets about this and is treating it as if it's a real news story pic.twitter.com/kDMGN7mGhU
— Brett (@BrettRedacted) April 5, 2024
Several users have pointed out that one of the reasons for this is that Musk fired employees involved in content moderation. X user hasanabi criticized, 'As a result of Musk firing so many employees, Trends and Explore are no longer aggregating information correctly. The reason fake news like 'Iran is attacking Tel Aviv' is being promoted is because Musk fired the moderation team and is aggregating information using AI.'
earlier today i noticed the trending and explore tab was no longer aggregating hashtags like it has been since elon fired the content team... turns out it's ai aggregation and it's promoting false stories like iran striking tel aviv right now. pic.twitter.com/sDkwjiA3fD
— hasanabi (@hasanthehun) April 4, 2024
'Simply put, Musk is using Grok to create fake news articles and then using his platform, X, to amplify, spread, and hide them from fact-checkers,' behavioral scientist Caroline Orr Bueno wrote. 'Now you can see why Musk bought this platform. He wants to create his own reality.'
So what you're saying is that Elon Musk used his in-house AI tool (Grok) to create a fake news article, then used his own platform to amplify it, spread it, & shield it from fact-checkers?
— Caroline Orr Bueno, Ph.D. (@RVAwonk) April 6, 2024
Now you can see why he bought the platform. He wants to create his own reality. https://t.co/sJJvjqHfRK
Grok's developer, xAI, is reportedly in talks with multiple investors to raise up to $3 billion, which would bring xAI's valuation to $18 billion.
Related Posts:
in Web Service, Posted by log1r_ut