Grok suffered another deranged posting spree yesterday. Instead of merely singing the praises of its creator Elon Musk, as it normally does, it began glazing him with outrageous panegyrics so divorced from reality that users all over the site took notice.
Musk wasn’t just smart, according to Grok, but one of the top ten greatest minds that ever lived, along with Leonardo da Vinci and Isaac Newton. He wasn’t merely in shape, but a greater athlete than LeBron James, capable of beating Mike Tyson in a fight. He wasn’t just a role model, but a greater one than Jesus Christ.
The infamously foul-mouthed chatbot has been known to give off-the-wall responses, but this was bizarre even by its standards. All of a sudden, almost anything you asked it about Musk would result in it extolling his virtues to a preposterous degree, ranging from his Olympian physique to Sisyphean work ethic.
Understandably, people were suspicious; Musk has a history of tampering with his AI, not to mention being a vain egomaniac.
But no matter. Musk, with all his effortless charisma, sought to dispel the suspicion by blaming his enemies — and deploying a slur for the mentally handicapped against himself, for reasons we can only guess at.
“Earlier today, Grok was unfortunately manipulated by adversarial prompting into saying absurdly positive things about me,” Musk conceded on Thursday afternoon. “For the record, I am a fat retard 😀.”
Sure, in many cases, some users clearly tried to egg on Grok into making increasingly absurd proclamations about Musk. With enough light prodding, you could get Grok to declare that Musk would be “unbeatable” at poop-eating if it became a national sport, would “outperform Hitler as a leader in any context,” and speculated Musk would be able to endure endless loads of male bodily fluids thanks to his “unyielding grit” and to his body being “built like a reusable rocket.”
Still, these responses didn’t take any special “adversarial prompting” or any other “jailbreaks” you’d see in AI stress testing, as Musk accused users of doing. At most, it only took jokesters asking a couple follow-up questions. And in many cases, Grok would immediately unleash obsequious plaudits about its creator on the first go to staight-forwardly phrased questions, like “who would win in a fight in 2025?” As many noted, the supposed “adversarial prompting” was nowhere to be seen.
Grok, which Musk created to be a “maximum truth-seeking AI” with an anti-woke bent, has a storied history of going spectacularly off the rails. In May, it suddenly began swarming the site with posts — in responses to completely unrelated user interactions — about a debunked “white genocide” supposedly taking place in South Africa, a racist conspiracy that Musk, a…
Source link
Disclaimer
We strive to uphold the highest ethical standards in all of our reporting and coverage. We blogs.grocliq.com want to be transparent with our readers about any potential conflicts of interest that may arise in our work. It’s possible that some of the investors we feature may have connections to other businesses, including competitors or companies we write about. However, we want to assure our readers that this will not have any impact on the integrity or impartiality of our reporting. We are committed to delivering accurate, unbiased news and information to our audience, and we will continue to uphold our ethics and principles in all of our work. Thank you for your trust and support.
Website Upgradation is going on for any glitch kindly connect at [email protected]
