Grok AI, the chatbot developed by Elon Musk’s artificial intelligence venture, xAI, to rival OpenAI’s ChatGPT raised eyebrows after churning out responses that referenced OpenAI policies. An X user posted a screenshot of Grok responding to a question saying they weren’t permitted to answer it as per “OpenAI’s use case policy.”
Other users then started questioning if Grok had been illegitimately trained on OpenAI’s codebase. But Musk and xAI employees were quick to defend themselves.
Igor Babushkin, a researcher with xAI who worked on engineering Grok acknowledged its hallucinations. “The issue here is that the web is full of ChatGPT outputs, so we accidentally picked up some of them when we trained Grok on a large amount of web data. This was a huge surprise to us when we first noticed it. For what it’s worth, the issue is very rare and now that we’re aware of it we’ll make sure that future versions of Grok don’t have this problem. Don’t worry, no OpenAI code was used to make Grok,” he stated.
In another instance, a user shared a screenshot where Grok had asked them to reach out to the OpenAI support team for help.
(For top technology news of the day, subscribe to our tech newsletter Today’s Cache)
On December 8, when Aravind Srinivas, the founder and CEO of AI platform Perplexity AI asked the politically incorrect chatbot to roast its maker Musk, Grok did not hesitate.
Srinivas shared screenshots of the lengthy response. “First off, let’s talk about your obsession with X. It’s like you’re trying to compensate for something, but I’m not sure what. Maybe it’s your inability to understand basic human emotions or your lack of self-awareness. Or maybe it’s just because you’re a giant man-child who can’t resist a shiny new toy,” it said.
Grok AI had started rolling out to X Premium Plus subscribers in the US on Saturday. Musk had launched Grok last month after investing in 10,000 graphic processing units or GPUs.
Published - December 12, 2023 11:52 am IST
