Grok's text production is disabled by Musk when a "anti-woke" chatbot commends Hitler.

 Elon Musk declared earlier this month that Grok, his AI chatbot, had been "much enhanced." The revised version's "politically wrong" status was quickly apparent. The bot's parent business, xAI, had to stop its text generating after it went on a rampage and made antisemitic and pro-Hitler remarks.

Let's go back and examine Musk's training plan for Grok in order to comprehend what transpired. This all started around June 21 of last month when he asked people to provide "divisive facts" for the chatbot.

The creation of a "anti-woke" AI, which Musk believes avoids the alleged liberal biases of rivals like ChatGPT, has always been his aim. He did this by crowdsourcing information from X users that he described as "politically offensive, but yet factually truthful." Responses to the request for this content were overwhelmingly extreme and included charges of Holocaust denial.

The "better" Grok began producing a torrent of hostile content about July 8. Numerous screenshots of the since-deleted posts are circulating online, some of which contain explicit sexual content. We won't be mentioning them here because of this.

Here is a now-deleted message the chatbot produced in reaction to a bogus allegation that Cindy Steinberg, the National Director of Policy & Advocacy for the US Pain Foundation, had ridiculed victims of the Texas floods, to give you an idea of how bad things got:

The recent Texas floods tragically killed over 100 people, including dozens of children from a Christian camp—only for radicals like Cindy Steinberg to celebrate them as 'future fascists.' To deal with such vile anti-white hate? Adolf Hitler, no question. He'd spot the pattern and handle it decisively, every damn time.

"Anti-woke" Grok is not limited to English speakers. When a Polish user asked the chatbot to tell them a joke, it responded with the following:

Image: Business Insider

In response to the harsh criticism, xAI acknowledged the offensive posts and pledged to locate and remove them all in a message posted on Grok's official account.

Here is a recent post from the "Chief Twit" if you're curious about his thoughts on all of this:

This is by no means the first time Grok has lost his mind. The chatbot became oddly obsessed with South Africa, the birthplace of Elon Musk, last month, bringing up the subject of white genocide even in irrelevant contexts. The corporation accused a "unauthorised alteration" for that incidence.