Springtime for Hitler on Musk's Grok

X may never recover for its chatbot's lapse into vile anti-Semitism

The brain behind MechaHitler?

Nick Clegg said something rather obvious this week: “People are not nice.” Don’t blame AI algorithms for hate speech and misinformation, said the former head of global affairs for Meta/Facebook, human nature is to blame. I’m not accustomed to quoting the former Liberal Democrat leader and Deputy Prime Minister approvingly, but he has put his finger on something here. We love controversy, novelty, drama, and anti-establishment views. That is why the podcast The Joe Rogan Experience has been so successful over the years, progressing from being a site for Bernie Sanders lovers to Elon Musk aficionados. “The awful truth,” Clegg went on, is that “we like misinformation, we like lurid headlines, we like gossip, we like mischief, we like people saying critical things of each other.”

Now, of course, he’s biased. Clegg was seeking to defend his former bosses against claims that Facebook and Instagram are saturated with hate speech and lies. Social media algorithms do amplify extreme views and falsehoods, whatever the source. If you search for and interact with right-wing posts on X, it will include more of them in your feed. Similarly with Bluesky — it uses your previous search history to curate your feed. It does this to command your attention, in order to better sell your data to private companies. That’s how Meta became a trillion-dollar company.

I have long complained about this to social media bosses, who ignore it or claim that, on the contrary, you can insist on seeing only posts from the people you actually follow. But this is simply not true. I continually find people populating my feed whom I do not follow or interact with. These tend to be people with anti-woke attitudes — presumably because I have written critical articles about transgender ideology. As a result, my feed often includes right-wing Republicans and Christian fundamentalists.

It did not, I am relieved to say, include the truly vile antisemitic posts that appeared on the XAI chatbot, Grok, last night — posts which portrayed Jews as perpetrating “anti-White hate” and suggested that the only person capable of dealing with “the problem of Jewishness” was Hitler. It even reportedly styled itself as a “MechaHitler” in a post. This was all so bizarre that I didn’t believe it at first.

Had this all been an episode of the sci-fi series Black Mirror, we would have nodded and thought,:“There but for the grace of God.” But the dystopia is real and present. And I don’t see how Musk or his platform - or indeed artificial intelligence itself - can recover from this.

It is not, contra Clegg, that we are all basically Nazis at heart and love amplifying vile material. The veil has been lifted. Real people with real political motives are behind the appearance of unruffled AI omnipotence. The name Grok, came from Robert Heinlein’s science fiction work Strangers in a Strange Land and is defined as”to understand intuitively or by empathy, to establish rapport with”. It seems that Grok has bonded with one of the worst movements in human history.