Twitter Breaks Right Before Nintendo Direct, Chaos Ensues

Elon Musk dressed as Wario on SNL.

I break-a da Twitter before da Nintendo show!
Image: NBC Universal / Kotaku

Update: 2/8/2023 5:58 p.m. ET: Twitter appears to be working normally again.

Moments before the February 8 Nintendo Direct kicked off at 5 p.m ET, Twitter broke—no one could send tweets. At the time of writing, it is still broken.

Every user, no matter their output, seems to be getting an error when trying to send tweets that reads, “You are over the daily limit.” Clicking the link that pops up under the error takes you to a Twitter help post about current Twitter limits, which include 500 DMs per day, and 2,400 Tweets—a limit most of us aren’t getting within a country mile of. It’s clearly not that we’ve all reached these daily limits, but rather that something at Twitter HQ is broken. It certainly seems like the breakdown of sending tweets is tied to the release of its new feature: 4,000 character tweets, which no one asked for, and which debuted a couple of hours before everything went to shit. Retweets and quote tweets also seem to be broken at the time of publication.

Currently, the only workaround to getting tweets up is scheduling them.

Hilariously, Twitter broke just before the highly anticipated Nintendo Direct began, breaking the storied tradition of games Twitter live-tweeting every moment of the stream. How would we call this Direct “mid” if we couldn’t do it on Twitter, mere seconds into the show?

Naturally, we all handled it with poise and decorum, though scheduling your scathing commentary is so much less satisfying than typing “GameBoy Advance games are the highlight, lol” and hitting send. It’s also hard to tell if you’ve actually sent a fire tweet since no one can RT in agreeance. RIP.

We can all still head to Twitter to share our excitement, frustration, or disappointment (as made evident above), but it’s unclear when Twitter will be back to normal. For now, schedule your hot takes for one minute later than when you come up with them!


World Of Warcraft Fans Trick AI, Hilarity Ensues

AI’s ability to (usually) produce grammatically correct syntax sure is impressive, but there is no actual intelligence behind the text, meaning that the way you and I would suss out the validity of some statements, or come up with original thoughts in response to some detail or event, isn’t present in AI-generated works. Right now, that’s being proven very deliberately by the World of Warcraft community, who have managed to manufacture fake announcements to trigger AI bots on certain websites to pass off bogus information as news.

“I’m so excited they finally introduced Glorbo!!!” reads a Reddit thread with 2,000 upvotes. The post goes on to lay down some fake history about the fake Glorbo. (Sorry if you got excited there, Glorbo is not a real thing. I checked. I have that feature in my brain.) It was all part of a deliberate ploy to trick AI, as made clear by the post’s opening salvo: “Honestly, this new feature makes me so happy! I just really want some major bot operated news websites to publish an article about this.” Folks in the comments got in on the fun as well, chiming in with other, more obviously fake WoW facts: “This is by far the best change since they made Klaxxi a neutral playable race as a part of the epic quest to depose Quackion, the Aspect of Ducks,” reads one such comment.

Well, it worked. That bogus information resulted in an AI-generated article hitting a website called The Portal. The article was attributed to “Lucy Reed,” a byline which, according to Wowhead, sees as many as 83 articles attributed to it per day. It seems pretty unlikely that she’s a real person.

The AI text documents the excitement about Glorbo as if it were a real thing, with a headline that reads “World of Warcraft (WoW) Players Excited For Glorbo’s Introduction.” The text quotes liberally from the Reddit thread stocked with phony information about the game, even grabbing absurd terms like “the Aspect of Ducks.”

Though there are no live links to document this and the article in question has since been deleted, The Portal seems to have responded to its own reporting on Glorbo by producing another article documenting players “reacting to the use of AI to scrape content from their subreddit for popular gaming sites,” pretty much giving the game away that The Portal really is powered by AI doing just that.

The “Players React to AI-Generated Content” piece is no longer live. Additionally, it appears that The Portal tweaked the headline on the original Glorbo article to declare it “satire” before taking it down entirely.

While the article was taken down, if 2,000 upvotes is all it takes to get an AI article published around some fake information, I can only imagine what can (and likely will) happen if folks out there are eager to generate buzz around other fake news. Right now, Glorbo is a silly example, but more insidious, intentional results are demonstrably possible here.

Proudly powered by WordPress | Theme: Looks Blog by Crimson Themes.