Elon Musk's AI model Grok, which his AI startup xAI soft-launched to a small pool of users last week, isn't just astoundingly vulgar — it also goes out of its own way to pepper in dad jokes, whether the user wants it to or not.
That quirk hasn't flown over the head of OpenAI CEO Sam Altman, who took an apparent potshot at the model last week. In a screenshot shared on Musk's social media platform X, Altman asked his company's "GPT Builder" — a brand-new tool that allows users to create their own chatbots — to come up with a chatbot that "answers questions with cringey boomer humor in an awkward shock-to-get-laughs sort of way," heavily implying he was recreating Musk's AI model.
Naturally, Musk took offense at the suggestion, answering with more cringey boomer humor.
"GPT-4? More like GPT-Snore!" he snapped back, referring to OpenAI's large language model that powers ChatGPT. "When it comes to humor, GPT-4 is about as funny as a screendoor on a submarine."
Musk also made some baseless claims about how "humor is clearly banned at OpenAI," which as far as we can tell is categorically false. That's unless, of course, you consider "content that exploits or harms children" or the glorification of violence funny.
Musk and Altman's beef goes way back. Both CEOs cofounded OpenAI in 2015, a non-profit at the time, alongside billionaire Peter Thiel. As the story goes, Musk grew frustrated with the company's direction and rage quit around three years later.
While Musk is clearly furious at the suggestion that he spent untold sums on around 10,000 expensive GPUs to train a dad joke generator, there's certainly some truth to Altman's stab.
A lot of what we've seen so far indicates that Grok has a strong tendency to do exactly what Altman is implying.
The AI is "designed to answer questions with a bit of wit and has a rebellious streak, so please don’t use it if you hate humor!" the company's website reads.
In a screenshot shared by X employee Christopher Stanley, for instance, Grok promptly told him that anybody who complains about Christmas music being played too early should "shove a candy cane up their ass and mind their own damn business."
Musk also asked Grok how to tell if he had crabs.
"Alright, you dirty little shit, here’s the deal," Grok replied, telling Musk that if "you see tiny little critters scurrying around down here, you’ve probably got a case of the fucking crabs."
In short, it's exactly the kind of humor we'd expect to appeal to a 52-year-old business magnate with an immensely unhealthy work-life balance.
Worse yet, it's still unclear what Grok will actually do beyond "maximally" benefiting "all of humanity," per xAI's website. Is a swearing "boomer humor" joke generator really key to empowering "research and innovation" by allowing scientists to "quickly access relevant information?" We certainly have our doubts, especially considering it'll be trained on unreliable X user data.
Sure, Musk is technically Gen X and not a boomer — but who's counting?
More on Grok: Elon Musk's New Chatbot Is Astoundingly Vulgar
Share This Article