AI Bot X Gives Out Comic Messages As Real News
If you pay for Twitter (sorry, X), you now have access to Grok, the company’s AI bot. Part of this perk includes access to a feed of trending news in the Explore tab, powered by the company’s artificial intelligence. The only problem? This is trash.
Here’s how Grok’s news feed works: the bot collects the “most popular” posts regarding any given news story, and then generates a news summary based on those posts. Simple enough, and we’ve already seen generative AI do this. But before you fire your writers and editors and put Grock in charge of the news, you might want to take a look at exactly how reporting works.
Grok appears to collect humorous tweets and present AI-generated replies as genuine news. You can see it in this post posted by user X BrettRedacted after the earthquake that rocked much of the New York City metro area . The bot created the headline: “Adams vs. Earthquake: 50,000 Cops in Subway Showdown,” and then reported how New York City Mayor Eric Adams would deploy the NYPD to “prevent further earthquakes” by considering using “robot cops.” and ordered “every cop in the city” to “shoot that damn earthquake before it hits again.”
Sure, we live in strange times, but no reasonable person would ever believe that Grok’s account of the news is accurate. If for some reason you can’t decide for yourself, you can take a look at the most popular posts fueling the roundup, which in this case are humorous tweets about the mayor’s reaction to the earthquake.
This is a sad but humorous example of the state of X in 2024. Previous versions of the site were a place to follow both legitimate updates on breaking news, such as the New York City earthquake, and laugh at jokes about the situation. Now the site treats jokes as news. I think you get what you pay for.
It doesn’t take much foresight to imagine how this situation goes from sad but humorous to downright dangerous. What happens when Grok decides to “report” something that appears to be legitimate on the surface, but is based on rampant misinformation running around the site ? User X might go to the “Review” page to see that their city is being bombed or that a candidate in their election did something illegal, even though none of the “stories” are actually true.
While we can’t stop the site from promoting this nonsense, we can all collectively agree not to treat Grok, or any other AI generator for that matter, as a legitimate news source or an accurate summator of the top news stories of the day.