AI Fail: X/Twitter’s “Grok” Falsely Reports Tyler Childers Postponements

Once again, AI as replacement for actual journalism spectacularly fails. This time the culprit is X/Twitter’s “Grok” tool, which the social media giant recently started using to generate news stories native to the application based off of information aggregated from tweets.
On Thursday and Friday night this week (April 18th and 19th), Tyler Childers was scheduled to play two sold-out shows at Nashville’s Bridgestone Arena as part of his ’24 Mule Plow Tour. This is a landmark moment for the Kentucky native since it illustrates the commercial prowess of country music that is not supported by corporate country radio or the Music Row system.
Selling out the Bridgestone on one night, let alone two marks a critical achievement for any country music artist, and right in the belly of the corporate country beast.
On Thursday evening, Tyler’s show went off without a hitch, with 49 Winchester opening the show for Childers, and the tens of thousands of attendees walking away raving about the performance.
But according to X/Twitter, that show was postponed, along with Friday’s performance. In a trending news story generated by Grok, X/Twitter states,
Tyler Childers has postponed his highly anticipated Mule Pull ’24 Tour shows at Bridgestone Arena in Nashville, TN, originally scheduled for April 18 and 19. The postponement is due to unforeseen circumstances, and new dates are being determined. Fans who had purchased tickets for the original dates can rest assured that their tickets will remain valid for the rescheduled shows. Despite the disappointment, fans are eagerly awaiting the rescheduled dates, showcasing their unwavering support for the Kentucky poet.
But of course, this information is completely false. This false information is not just being served by X/Twitter, it’s populating in the right sidebar on Desktop in Twitter’s “Explore” feature, and in people’s feeds on mobile as a trending topic.
In smaller font than the story itself, there is a short disclaimer that reads, “This story is a summary of posts on X and may evolve over time. Grok can make mistakes, verify its outputs.”

Below the story are tweets from reputable sources such as The Nashville Scene, The Ryman Auditorium, and a gaggle of other Twitter users, but none of them speak whatsoever about any postponements. And since the false news story is being presented as a trending topic, there’s a good chance it very well could convince some users with tickets that the shows have indeed been postponed.
According to the news story, it had been up for at least 15 hours. There is also no way to report or attempt to correct the false information.
Such fake news stories can have significant real-world implications. As Mashable reports, earlier in April, Grok and X/Twitter falsely reported that Iran had attacked Israel. Cybernews also reported a fake news story by Grok claiming that New York Mayor Eric Adams was deploying 50,000 cops to the subway system to “shoot” a recent earthquake. The Grok news story was based off of joke tweets.
Grok’s failure is just the latest in a string of AI mishaps as tech companies scramble to keep up with what they believe will be the next paradigm in technology and media, while human handlers are nowhere to be found to at least give tacit oversight into the process. This only assures the exacerbation of distrust in all of media as the integrity of AI-generated news often is even worse than the human version.
April 19, 2024 @ 10:10 am
Sounds like Tyler Childers has an enemy who’s fking with him.
A fake concert postponement announcement is nasty, but relatively trivial. He should probably alert the police and take steps in advance, in case it escalates to fake 911 call about imaginary shootings and hostage-takings and at his house
April 19, 2024 @ 10:22 am
And a large number of people will unquestionably believe the ‘truth’ that AI and media propagates.
April 19, 2024 @ 1:40 pm
@savingcountrymusic.com – There is no substitute for Human intelligence, RJay, but an AI can be a writer’s ‘Copilot’ to quickly check for mistakes and a place to get a whole lot of “factual” information that can be used to emphasize a discussion point or make a statement with certainty. An AI depends on a Human to make it relevant. Like a singer-song-writer who sell his or her songs to a better singer-song-writer who then personalizes the words and make the words come alive in a song. Also, like the same old argument that states: “guns don’t kill, but people do” PS. don’t ask an AI that question if you want nothing but the ‘truth’.
April 19, 2024 @ 2:00 pm
Idiot.
AI, or not.
April 19, 2024 @ 2:36 pm
I have no problem with AI being used as a tool competent humans you describe. The issue here is that AI seems to be roaming free range and spewing loads of garbage with no checks or balances from a competent human.
April 19, 2024 @ 10:36 am
This is the same feature that reported yesterday Klay Thompson was being investigated by Sacramento police after he vandalized a number of homes by throwing bricks at them. This is because of people tweeting about klay Thompson missing a bunch of shots in an nba game. Turns out AI sucks
April 19, 2024 @ 10:39 am
He played last night to a packed house in Bridgestone and was fantastic. Great, great show. Doing it again tonight as well. Amazing how a guy with no radio airplay can build such a huge and enthusiastic following.
Trigger, you reporting from Two-Step this year?
April 19, 2024 @ 10:49 am
I am currently planning to be at the Two Step Inn, but the weather forecast right now is VERY dodgy. A good chance some or all of it doesn’t even go down. We’ll see.
April 19, 2024 @ 12:35 pm
This is not an AI problem as much as it is a Twitter problem. AI can probably do a pretty good job summarizing real news. But when you take the amalgamation of stench, hate, lies, and deep fakes that comprise the majority of Twitter, this is what you get.
April 19, 2024 @ 12:52 pm
AI has a very dubious track record composing news. We saw this with newspaper giant Gannet’s effort to have AI write sports coverage that was so laughable, they had the bury the project. Then you had Google’s Gemini get basic facts wrong like showing the US founding fathers as Black guys to push a diversity agenda.
I do agree though if your point of aggregation is Tweets, it’s going to fail. Twitter is the home of misinformation to begin with, not to mention the sarcasm, shitposters, and fake accounts. AI can be useful at creating summations of deeper news stories, but only if the underlying facts and information is accurate. The issue with that is this practice is illegal with copyrighted material. I guess Musk believes the tweets are his property, but this and other examples show there’s just not enough truth on Twitter to make the practice viable.
April 19, 2024 @ 3:51 pm
“But when you take the amalgamation of stench, hate, lies, and deep fakes that comprise the majority of Twitter, this is what you get.”
Nah, they got rid of most of this when Musk bought and the majority of leftists deleted their accounts (after having a huge cry-in first).
April 19, 2024 @ 3:58 pm
You sound like someone who pays $8 a month for twitter, if you think it’s any better now.
April 19, 2024 @ 4:17 pm
Yeah, that’s just not true. As someone who is on X/Twitter many times a day monitoring for news, I would say that at any given time, the majority of content on the platform is false, because the more false something is, the more viral it goes. It reached a fevered pitch recently with the release of Beyonce’s “Cowboy Carter” to the point where the entire world believed it was a country album despite Beyonce saying the contrary.
False and misleading reporting comes and goes all the time for a host of reasons, and actual news stories usually get corrected or pulled within hours. This trending topic has now been up for 21 hours, and it’s so commonplace on Twitter, nobody even seems to give a shit.
I can’t tell you how much this stuff literally makes me lose hope for the future of humanity. It’s not just the fake news. It’s that nobody can even find the energy to give a shit anymore. Years ago, something like this would be a national news story itself.
April 19, 2024 @ 6:28 pm
X/Twitter, CNN, no difference.
Save yourselves the meltdown(s).
April 19, 2024 @ 8:01 pm
Musk has been great for X.
April 19, 2024 @ 1:43 pm
I don’t know if it is AI fake news or just plain fake news, but there’s a lot of stupid “articles” floating around social media claiming that this country legend or that one “opposes Beyonce’.”
April 19, 2024 @ 1:47 pm
I love X, and Substack. All the people whining about them are “govern me harder daddy” crybabies.
Using X posts to generate news though…doesn’t take a genius to know that may not work out so well.
April 19, 2024 @ 1:59 pm
I’m not necessarily bagging on X. But if the idea is to replace the need for actual journalists with tweet-based AI that serves false information, that seems like a lunge towards dystopia that won’t be healthy for anyone.
Perhaps just as troubling as this story is the fact that nobody really seems to give a shit about it. There is a very good chance folks get misled by this. If this was being pushed for an event I was either promoting or a part of, I would probably sue, especially since it’s been up for going on 24 hours. Strong grounds for damages.
April 19, 2024 @ 6:49 pm
Question though..how many people get their news from Grok? How many people missed the show due to this?
I don’t really care because there’s much worse things to care about, and I know that people won’t trust the AI, especially if it does stuff like this or is insufferably woke like Gemini. Its not doomsday…for a little while at least.
The medical / pharma industry and the government are far more concerning pushers of misinformation than any one or thing else.
April 19, 2024 @ 7:34 pm
A ton of people get their news from X, and Grok is sharing the information as a trending topic, meaning it’s serving it to hundreds of thousands of people. Most people who interact with the information probably have no idea it’s being populated by Grok. They just see it on Twitter.
That said, I don’t know how many people are being mislead by it, resulting in an actual disruption to their life. My guess is those numbers would be low, if only because the next thing folks would do is try to find out where to get a refund, and they’d find out the shows are actually not postponed.
But I’m not even sure if that’s the point. The point is this is no way to disseminate news. The reason X and other social media platforms are trying out this technology is they don’t want folks going off their platforms for further news. They want it all served in one place. That’s especially insidious if that news is false.
April 22, 2024 @ 5:47 pm
“But I’m not even sure if that’s the point. The point is this is no way to disseminate news. The reason X and other social media platforms are trying out this technology is they don’t want folks going off their platforms for further news. They want it all served in one place. That’s especially insidious if that news is false.”
We definitely need to include Zuckerberg in this conversation.
Worst offender of all.
April 19, 2024 @ 5:20 pm
My health insurance is cheaper if I agree to use an AI doctor. It’s got me scheduled for a prostate exam and a hysterectomy next week.
April 19, 2024 @ 5:52 pm
This article represents why real journalism by real journalists, just like SCM, will never fully die or go away. People will inevitably gravitate back to real human beings for their news and information. Fuck AI.
April 21, 2024 @ 3:41 am
AI = Artificial Ignorance
April 21, 2024 @ 6:11 am
I was wondering what Grok was. I use twitter to follow sports and advertisements to join Grok pop up all the time. I pretty much ignore it. I don’t see anyone talking about it, so I assumed most others do as well. Guess I’ll continue to ignore it.
April 22, 2024 @ 5:51 am
Lets stick to people being journalists, okay?
This AI stuff is out of hand.
Problem is too many people support it.
It’s too late.
April 22, 2024 @ 10:22 am
Amazing that Trig posted 4 examples of AI generated fake news ranging from the relatively harmless (Childers) to potentially destructive (Iran/Israel), yet the usual suspects on this board have bent over backwards because if Elon Musk is doing something to “own the libs” then no amount of misinformation is worth distracting from that goal.
The thing about the harm of AI is that it cares not for left or right (though certainly it can be trained to have the biases of either), but because it is by definition artificial and incapable of understanding real world events. Today it falsely reports a canceled concert, tomorrow it could detect a non-existent nuke on its way from Russia and start us down a path of total annihilation.
April 22, 2024 @ 12:23 pm
Elon Musk’s leadership at X has been very uneven, very mercurial, and at times, damaging to the platform. I do agree that some of his free speech moves have been important. But then he’ll be very thin skinned and remove content or people that affect him personally. Then we have this, which is impossible to justify. This truly makes me fear for the future of humanity. Musk himself has spoken out abut the concerns with AI, and here he is being a party to it.
April 23, 2024 @ 9:25 am
Owning the libs is worth some minor fake reports.
Musk’s ownership has been positive.
April 23, 2024 @ 11:53 am
Careful what you wish for
April 23, 2024 @ 1:18 pm
Oh man, a cryptic threat.
It doesn’t matter. Congress just waved Ukrainian flags while passing a bill that offered no money for border security.
Fake AI reports are the least of this nation’s concerns while fentanyl floods across the border.