Grok is an AI chatbot created by Elon Musk that was launched in 2023 to compete with other AI chatbots like Open AI. Given Musk's current embrace of right wing politics, it was no surprise when he claimed that it would be “maximally truth-seeking … even if that truth is sometimes at odds with what is politically correct.” He also said in an interview on Fox News that OpenAI had been “training the AI to lie” by incorporating human feedback that directed the chatbot “not to say what the data actually demands that it say.” Implying, of course, that Grok would not report such lies.
But, like any chatbot, Grok searches the internet for reliable, factual information, and, as Stephen Colbert once pointed out, "the truth has a liberal bias." In other words, Musk's own opinions, based as they are on a rightwing ideology that believes what Donald Trump and the right wing media says without question, is often more about feelings and desires than reality. (For example, those Haitian immigrants never actually ate dogs, despite what Trump said in the 2024 election, but his followers wanted it to be true, so they perceived it as such.) But feelings and desires are not supposed to drive the gathering and distributing of facts and information, as a few pointed questions I made at Grok's free website proved:
My first question was about Musk's Department of Government Efficiency. QUESTION: What did Elon Musk accomplish with his DOGE Government initiative? While the answer given was long, here's what was given as a "critical perspective" at the end: "While DOGE achieved measurable cuts in spending and workforce, its impact appears overstated, with savings far below initial promises and significant collateral damage to government services. The initiative’s aggressive approach—modeled on private-sector cost-cutting—disregarded the complexity of public administration, leading to legal challenges, dysfunction, and public discontent. Musk’s dual role as a business magnate and government advisor raised ethical questions, with some viewing DOGE as a vehicle to advance his corporate interests. The lack of transparency and verifiable data further undermines claims of success, suggesting DOGE was more disruptive than transformative." Those last few words are pretty striking, "more disruptive than transformative". Elsewhere in the answer it also states that "DOGE targeted agencies investigating Musk’s companies, raising concerns about political weaponization" and that it promised 2 trillion dollars in savings and only accomplished one hundred and fifty billion. It even pointed out the errors DOGE made on its own website, like claiming an $8 billion ICE contract cancellation that really only cost $8 million. In other words, even Musk's own chatbot knows that then promises of DOGE were overblown, that Musk himself was the wrong person for the job given his own conflicts of interest, and that the results were mixed at best. So, basically everything that anyone that isn't a Musk fanboy should know.
Wanting to dig deeper into the DOGE issue, I asked a darker question about Musk's heavy cuts to the USAID program, which only accounts for less than one percent of our total government spending : QUESTION: Will the cuts to USAID made by DOGE result in the death of children? Grok did not mince words here: "Studies and internal USAID memos estimate significant increases in child mortality due to the disruption of lifesaving programs. For instance, researchers cited in posts on X suggest that DOGE’s cuts could lead to approximately 148,000 child deaths, part of a total of 219,000 deaths attributed to the aid reductions." It then explained that child deaths would increase due to spending cuts on USAID programs concerning malnutrition, infectious diseases and HIV/AIDS. While it did mention that "some argue the cuts are necessary to eliminate inefficient or politically driven programs, as claimed by President Trump and Elon Musk, who assert USAID promotes a liberal agenda and wastes money. There’s also the possibility that some terminated programs were mistakenly cut and could be reinstated, as suggested by internal communications." But it then adds, " the lack of transparency on which programs remain active and the halting of payment systems even for approved waivers undermine claims that lifesaving efforts are being spared." It then ends with these dark words, "In summary, while the precise death toll is uncertain, the cuts to USAID’s programs are highly likely to result in thousands, if not millions, of additional child deaths due to the loss of critical health and nutrition support. The evidence from past aid impacts, current program terminations, and expert projections strongly supports this conclusion." This is truly overwhelming and depressing to consider, especially after Musk bragged about throwing the USAID program into "the wood chipper.", essentially joking about what could be called genocidal neglect.
Ok, here's another, less dark one: QUESTION: Has the Cybertruck been a success for the Tesla company? Again, the answer shows far more balance than Musk himself ever would: while Grok does mention that the initial sales and attention were strong after the Cybertruck's first release, it also shows how much those sales declined. For example, it mentions that "despite initial success, sales fell short of Elon Musk’s ambitious goal of 250,000 units annually, with only 38,965 sold in 2024—84% below projections." It also added that "by April 2025, Forbes labeled the Cybertruck the auto industry’s 'biggest flop in decades' due to its sales miss and polarizing reputation." Grok doesn't forget the eight recalls that have been made since the vehicle was released, some of them for serious issues like body panels detaching, and other complaints made about it like rusting panels and getting stuck in the snow. Finally, it said in conclusion that "The Cybertruck has been a partial success for Tesla. It achieved strong initial sales, profitability, and market dominance among EV pickups, while its bold design kept Tesla in the spotlight. However, it fell far short of Musk’s lofty sales targets, faced significant quality issues, and struggled to win over traditional truck buyers due to its niche design and high price. Its long-term success hinges on Tesla addressing quality concerns, expanding affordability, and navigating Musk’s polarizing public image. For now, it’s a high-profile win with notable flaws, not the game-changer Tesla hoped for." While I find that conclusion a bit generous towards Tesla, it still ends with it saying that Musk's dream of the Cybertruck replacing regular pick up trucks in this country just isn't happening.
Now let's try a question that is a matter of opinion: QUESTION: Is Elon Musk a good father? Right away Grok calls this a "tricky" question, which is understandable. It then lists attributes both good and bad about Musk's parenting: On the one hand, it says that he is seen spending time with some of his children, that he encourages them to be free thinkers and that he obviously has enough wealth to take care of them. On the other hand, it mentions that he spends most of his time working, is estranged from his trans daughter Vivian, and that "Posts on X and media reports often portray Musk as detached or inconsistent as a father. For instance, his relationship with Grimes ended in 2021, and she later posted on X about co-parenting challenges, hinting at Musk’s limited involvement." It also points out that giving children odd names like "X Æ A-Xii", "Exa Dark Sideræl", and" Techno Mechanicus"will be difficult for them in the future. While rating one parent as better than another is generally difficult, the openness with which Grok was willing to concede that Musk may not be the ideal dad once again shows that the chatbot has more adherence to the facts than Musk himself does.
And for the last one, I thought I'd wade into a current right wing obsession that Musk himself shares: QUESTION: Should trans children get gender reaffirming care? It's answer is balanced, listing both sides of the question before saying that "Decisions should involve thorough psychological assessment, informed consent (from both child and parents), and reversible steps where possible. Blanket policies—whether fully affirming or outright banning—ignore the nuance of each case. Families and doctors need flexibility to navigate this, grounded in the best available science, not ideology." Given that Musk has said that his own trans daughter was "killed by the woke mind virus”, clearly Grok's even handed answer is not part of his personal belief system, or the one given by the right wing media's scare tactics over the issue. Once again, Grok is more open minded than Musk himself.
While I'm sure a massive narcissist like Musk will never ask Grok questions like this, the fact that his own, not "politically correct", chatbot was willing to give such evenhanded and often negative answers about him and his actions shows that people like him and Trump live in a fantasy world, surrounded by sycophants, in which they see themselves as masters of the universe. But, according to Grok, the truth is far more complicated.