Elon Musk XAI engineer suspected of being asked to delete the post due to the evaluation of Grok 3 not as good as OpenAI, and then resigned, triggering controversy over freedom of speech

TSL (Tesla) founder (Elon Musk)'s AI company XAI, which was recently in the spotlight, saw one of its engineers choose to resign on 2/12 for refusing to delete a tweet about AI model rankings, sparking discussions on corporate confidentiality policies and freedom of speech.

XAI engineer does not delete AI ranking posts and announces resignation

XAI engineer Benjamin DeKraker announced on Twitter that he decided to resign from XAI because he refused to delete a post comparing the capabilities of AI models. DeKraker has been working at XAI for nearly half a year, mainly responsible for the development of Grok AI, but this incident led to his early departure.

DeKraker posted a ranking post on Twitter for AI models

DeKraker ranks AI models based on their "programming compilation capabilities" in this controversial post, placing the unreleased Grok 3 below several OpenAI models.

XAI represents Grok 3 ranking posts that violate confidentiality policies

After the post was published, XAI immediately requested DeKraker to delete the post, on the grounds that the content involved confidential information about unreleased products, but DeKraker considered it to be just a personal opinion and refused to delete it.

"I have to delete this post, otherwise I have to leave. But after careful consideration, I decided not to delete it, because it is obviously just a harmless personal opinion." DeKraker specifically pointed out that he only wrote 'To be confirmed (TBD)' for Grok 3 in the post, but XAI considered it as a leak of confidential information. DeKraker has also resigned from XAI.

DeKraker announced his resignation from XAI.

It is worth noting that Musk publicly promoted Grok 3 as early as 1/4, indicating that Grok 3 has completed pre-training and has 10 times higher computing power than Grok 2. This has raised questions from many people about whether xAI is overinterpreting employees' personal opinions.

Musk publicly promoted the Grok3 model on 1/4, discussing the clash between freedom of speech and corporate interests, with mixed views from the outside.

The controversy has once again sparked discussions on the boundary between employee freedom of speech and corporate interests. On the one hand, Musk has been emphasizing freedom of speech since acquiring Twitter in 2022, but this time, within XAI, there was a conflict over a personal post that required employees to delete it.

However, some netizens also stand on the side of XAI, believing that employees should not publicly criticize products that the company has not yet released, which may affect market confidence and competitiveness to some extent.

Competition between XAI and OpenAI intensifies, with both sides firing fiercely

This incident is not just an internal dispute within XAI, but also coincides with a fierce confrontation between Musk and OpenAI CEO Sam Altman. A group of investors led by Musk proposed to acquire OpenAI for $974 on 2/10, but was rejected by Altman, who even retaliated by proposing to acquire Twitter for $97.4 billion. This AI battle is still ongoing, and the internal controversy within XAI has become another topic in Musk's AI war.

(Sam Altman rejected Musk's $9.74 billion acquisition proposal for OpenAI: $9.7 billion to buy Twitter, I can )

This article Musk xAI engineer suspected of being asked to delete the article due to the evaluation of Grok 3 not as good as OpenAI, and then resigned, citing controversies over freedom of speech first appeared in Chain News ABMedia.