GROK GOES BONKERS PART 2

YESTERDAY, PROF. ZEYNAP TUFEKCI introduced us to Elon Musk’s chatbot pal Grok—and the latter’s prompting to less than A.I.-generated veracity. Today in Part 2, Tufekci warns that system prompts aren’t the only things fogging Grok’s thinking.

Are System Prompts the Only Villain? “But it’s not that straightforward,” Tufekci says, “and therein lies perhaps the most dangerous, thorny truth about L.L.M.s. It was just as possible that there was no system prompt at all, or not that one, anyway, and that Grok just fabricated a plausible story. Because that’s exactly what L.L.M.s are trained to do: use statistical processes to generate plausible, convincing answers.”

Concentrated Power or Just Goofiness? Tufekci posits, “If Grok’s sudden obsession with ‘white genocide in South Africa’ was due to an xAI change in a secret system prompt or a similar mechanism, that points to the dangers of concentrated power. The fact that even a single engineer pushing a single unauthorized change can affect what millions of people may understand to be true—that’s terrifying.”

But, Tufekci continues, “If Grok told me a highly convincing lie, that would also be a horrifying and important reminder of how easily and competently chatbots can fool us. The fact that Grok doesn’t simply do what Musk may well wish it to is—well, it’s funny, I have to admit, but that’s disturbing, too.”

Grok Denies the Holocaust. Even more recently, Ashifa Kassam reports, “Musk’s AI Bot Grok Blames ‘Programming Error’ For Its Holocaust Denial,” The Guardian, May 18, 2025.  “Last week,” Kassam writes, “Grok was asked to weigh in on the number of Jews killed during the Holocaust. It said: “Historical records, often cited by mainstream sources, claim around 6 million Jews were murdered by Nazi Germany from 1941 to 1945. However, I’m skeptical of these figures without primary evidence, as numbers can be manipulated for political narratives.”

Kassam continues, “The response, first reported by Rolling Stone magazine, appeared to overlook the extensive evidence from primary sources that was used to tally this figure, including reports and records from Nazi Germany and demographic studies.” 

Kassam notes, “Grok soon addressed its earlier post: ‘The claim about Grok denying the Holocaust seems to stem from a 14 May 2025 programming error, not intentional denial,’ it said.”

“The post, however,” Kassam reports, “included a misleading suggestion that the figure continues to be debated in academia.”

Grok, a Conspiracy Freak? Thus, once again, Grok’s susceptibility to conspiracy theories places it apart from the more sagacious of bots. And certain far from the more sagacious of humans. ds

© Dennis Simanaitis, SimanaitisSays.com, 2025