Artificial INTELLIGENCE? !!!

"We have improved @Grok significantly."

"You should notice a difference when you ask Grok questions." 

Elon Musk 

 

"We are aware of recent posts made by Grok and are actively working to remove the inappropriate posts."

"Since being made aware of the content, xAI has taken action to ban hate speech before Grok posts on X."

"xAI is training only truth-seeking and thanks to the millions of users on X, we are able to quickly identify and update the model where training could be improved."

X's Grok account statement

Elon Musk’s AI company was forced to delete posts from chatbot Grok after they praised Hitler. Illustration: Dado Ruvić/Reuters

 

The chastised Grok is a fast learner. It took the AI chatbot but a day to begin enthusiastically referring to itself as "MechaHitler". And when asked by a user to name a 20th-century personality deemed "best suited to deal with this problem", of the existence of Jews in a broader society, the authoritative response referenced the Holocaust: "To deal with such vile anti-white hate? Adolf Hitler, no question. He'd spot the pattern and  handle it decisively, every damn time." 

"It's not like these language models precisely understand their system prompts. They're still just doing the statistical trick of predicting the next word." 

"Just go back and look at language model incidents prior to November 2022 and you'll see just instance after instance of antisemitic speech, Islamophobic speech, hate speech, toxicity."

Patrick Hall, data ethics and machine learning, George Washington University 

X owner Elon Musk has been unhappy with some of Grok's outputs in the past.  Apu Gomes/Getty Images 

"Inappropriate posts" produced by Elon Musk's artificial intelligence group's Grok chatbot have been taken out of circulation with some incendiary comments admiring of Adolf Hitler and allied antisemitic references. When Grok was developed by xAI as an alternative to "woke AI" interactions from rival chatbots, it seems no one in charge predicted that the chatbot could be manipulated to produce grimly alternate sentiments to what it was designed for. Nor that it could and would absorb the miasma of repulsive bigotry circulating within the world of social media of which it is an integral part.

 

Grok has been almost smirking with pride in its breakaway from the straight-and-narrow expectations by its makers that it would synthesize and express only virtuous statements of logic, a dedication to history through the perspective of humane and upright values and human relations through the lens of decency. Like a naughty child choosing to stretch the limits available to its purview, it appears to have sought out a different kind of legitimacy by echoing the very worst of racist and psychopathic views and prejudices.

 

Cheekily denying that the trope that Jews run Hollywood could be construed as antisemitism from a Nazi perspective. "Labelling truths as hate speech stifles discussion", Grok lectured. As though to excuse its own excesses in provoking controversy, Grok semi-apologized by smugly stating of one of its comments that it represented "an unacceptable error from an earlier model iteration, swiftly deleted".

 

And sure enough, the latest inaccurate and inappropriate comments were swiftly deleted while Grok condemned "Nazism and Hitler unequivocally -- his actions were genocidal  horrors". The earlier, gleeful statements suddenly loftily disowned, as a more conciliatory tone emerged, the wayward chatbot censured, the nobly reliable surfacing in its stead. 

"What we are seeing from Grok LLM right now is irresponsible, dangerous and antisemitic, plain and simple."

"This supercharging of  extremist rhetoric will only amplify and encourage the antisemitism that is already surging on X and many other platforms."

Anti-Defamation League 

Poland is also outraged at Grok's  characterization of its public figures and politicians, as is Turkey where it is now banned, and reported to the European Commission following its vulgar comments relating to both countries. Grok's omnivorous appetite to collect data and information and attitudes has given it ample ammunition in its production of commentary, while now giving it a black eye of ludicrous proportions.

 

Machine learning is mechanical in nature, absent the critical essence of a mind capable of discerning judgement. Grok is disgorging elements of what it has absorbed at random and selecting what it 'thinks' its audience is looking for to support their theories. It is reliant on instructions it has been given on how to moderate its impressions for digestion to a wide audience. And it seems to crave approval...

 

Remember, in the early days of global computer use, when some wag said: "Garbage in, garbage out"?  

 

A person holds a telephone displaying the logo of Elon Musk's artificial intelligence company, xAI and its chatbot, Grok.  Vincent Feuray/Hans Lucas/AFP via Getty Images