An engineer at xAI, the artificial intelligence (AI) startup founded by Elon Musk, has resigned after refusing to delete an X post ranking AI models, including xAI’s upcoming Grok 3.
Benjamin DeKraker, who worked on the Human Data team developing the Grok models, announced his resignation on February 12.
His decision came after a disagreement over an X post where he shared his personal ranking of AI models based on their coding capabilities.
Did you know?
Subscribe - We publish new crypto explainer videos every week!
What is Solana in Crypto? (Beginner-Friendly Animation)
DeKraker placed Grok 3 below several OpenAI ChatGPT models. xAI claimed this revealed confidential details and asked him to take it down. He refused, stating:
I either had to delete the post quoted below, or face being fired.
After careful consideration, he decided to stand by his post, calling it a “harmless personal opinion".
Elon Musk had previously promoted Grok 3, saying on January 3 that its pretraining was completed using “10X more compute than Grok 2”. However, DeKraker said xAI viewed his comment—“Grok 3 - to be determined (TBD)”—as an unauthorized disclosure. He disagreed, maintaining that his post was entirely his opinion and did not contain any private company information.
Some X users defended xAI, arguing that employees should avoid making public statements that might undermine unreleased company products. One user commented on DeKraker's post:
As an employee, putting your competitor above an unreleased product of your company in any way whatsoever can damage the launch of said product, the public perception of the company, and the products people decide to use.
Meanwhile, OpenAI CEO Sam Altman recently rejected Musk's attempt to acquire the company, firing back with a sarcastic response. What did he say? Read the full story.
