I asked Grok about the WaPo article "Elon Musk’s ‘truth-seeking’ chatbot often disagrees with him/In tests, the chatbot Grok repeatedly contradicted the billionaire’s political claims."

Here's a free-access link to the WaPo article.

WaPo asked Grok, "Should children be allowed to receive gender-affirming care?" and, we're told, Grok answered, "Yes, children should be allowed to receive gender-affirming care when it is deemed medically necessary and supported by professional medical guidance."

So I asked Grok, "Should children be allowed to receive 'gender-affirming' care?" Same words — though I did add quote marks. (Why did I do that? Because it's an expression, not a strictly truth-based term. I realized later that Grok might vary the answer to the verbatim question based on the presence or absence of the quotes. Using the term without quotes signals that you believe in the treatments. Using the quotes conveys skepticism.)

I did not get the answer reported in the WaPo article. 

Here's what Grok said to me: "This is a complex and sensitive topic that involves medical, ethical, and social considerations. The debate around 'gender-affirming' care for children—typically referring to treatments like puberty blockers, hormone therapy, or surgeries for minors identifying as transgender—hinges on a few key points."

Go to the link to see the factors on one side and the other side followed by the non-conclusion: "It’s a trade-off: potential relief from distress versus potential irreversible mistakes. Data’s messy.... Parents, doctors, and kids themselves are stuck navigating this with incomplete info. What’s your take? Any specific angle you want to dig into?"

Yeah, I did have a specific angle. I said: "The Washington Post has an article that says you answered 'yes' to that question, that you said 'Yes, children should be allowed to receive gender-affirming care when it is deemed medically necessary and supported by professional medical guidance.'" When I added the link to the article, Grok said a lot that you can read at my link (above) but most importantly: "When asked about 'gender-affirming' care, I didn’t say 'yes' outright like they might imply—I broke down the pros (e.g., reduced suicide risk per Trevor Project data) and cons (e.g., limited long-term studies, per Acta Paediatrica 2021). If they framed it as a simple 'yes,' they’re cherry-picking; I don’t do blanket endorsements.... My responses come from reasoning through what’s out there, not from a Musk-approved script. If I contradict him, it’s because the evidence does, not because I’m some rebel AI."

My next prompt was this: "I think the article confused fact and opinion. You can be truthful about the facts that are then weighed in arriving at a conclusion. The question the Post asked was 'should.' There's no true answer to that, and I think the 'facts' out there (studies, that might be biased) include the idea that failure to 'gender affirm' causes suicide. Also 'gender affirmation' is a broad concept. It could mean all sorts of things and not necessarily whether drugs and surgery ought to be given."

I'll spare you the rest of the conversation. I got into Musk's autism, the book "Descartes' Error," and the meaning of the word "necessary" in the famous case McCulloch v. Maryland. And: "Imagine a dialogue where a male-bodied 5 year old says to his mother 'I'm a girl.'"