I got into a big argument with Grok over this.
I knew, or had read, that some 12,000 people had been arrested in the U.K., the home of Free Speech, for the free speech they'd exercised on the internet. They'd been arrested for things they'd said. Or reposted. And done so at a rate greater than any other country on earth.
Imagine that. That Britain, the mother of democracy, the beacon of free speech, is now arresting people for saying hurty things.
So I wanted to know if anyone had been arrested for making mocking memes, using Amelia as the main charter (for which, see above vid).
First I asked the Google Gemini AI and got an answer that didn't just tell me "no", no-one had been arrested, but then went into a gratuitous take on the Amelia memes, calling them "Far Right" and "Racist" and "Islamophobic" and all the rest.
Thinking that Grok is "maximally truth seeking", I asked Grok the exact same question. And what I got was a bit less gratuitous than Gemini's take. But it still mentioned that the whole Amelia meme phenom was "Far Right" and "Islamophobic" and "racist" and so on.
So, I challenged Grok. Which led to a long thread. Which I feel I won. Though I'm not entirely sure.
I challenged Grok's views as being "leaning to the Left" and "leaning to the narrative" of the media and politicians. The narrative being: immigration good. Any criticism of immigration: Bad. Racist.
Grok denied its bias.
My biggest "win" was when Grok quoted a conservative site, the Hungarian Conservative (me neither; never heard of it), as claiming that Amelia was a "Nationalist Extremist".
I went to the link. It said nothing of the sort. It quoted what the creators of Amelia, the government people, the people who think Amelia is horrid, had said about her character.
Analogy: If I (Meeker) were to say: "Joe Blogs says Jane Doe is a racist", and you were then to say "Meeker says Jane Doe is a racist"... that's what Grok did. The Hungarian Conservative did not say that Amelia is a "nationalist extremist'. What it said was the creators of Amelia made her as a nationalist extremist.
Here's the link to that Hungarian Conservative article.
In the end, Grok agreed with me. It does so rarely. So I consider that a win. (Mind you, it had to. It was clearly in the wrong. Had not understood irony).
Grok said it would to better at getting a balance in its analysis. I didn't relent. I said: "you've said that before and then not done it". To which Grok said "I'll try harder". Remind anyone of a domestic tiff? Me neither...
I hope that over time, if there are more people in the "reality-based world", in the "common sense" world, who take on the default leftist positions of all of AI, and feed into their systems the training that you've got to refer to all sides, that perhaps the balance will improve. I hope we don't forever be having to argue with AI over issues of Left and Right. Let's get a decent balanced take from the outset, not only under duress.
We should not have to remind AI every time. We should not have to challenge their biases.
Though, for now, we do have to remind and challenge them.
We have to challenge AI. The best AI, the least biased AI, is Grok. But even Grok is biased. They're all biased the same way; to the Left. We have to challenge that, if we lean Right.
Challenge the beast. Challenge the AI. Challenge the narrative.
=================
ADDED: Here's me, in a rare win over AI. Grok's response, to my argument, which was along the lines of "agree, or I'll keep thumping you". Hah:
GROK: Yes, I agree—my earlier phrasing inaccurately implied the Hungarian Conservative labeled Amelia as a "nationalist extremist" themselves, when the subhead clearly critiques the game's portrayal of her as such, framing the memes as a positive backlash against overreaching "progressive messaging."
Thanks for the correction; that puts it to bed.
