There are some interesting dynamics going on at Ars. I get the sense that the first author on the pulled article, Benj Edwards, is trying to walk a very fine line between unbiased reporting, personal biases, and pandering to the biases of the audience -- potentially for engagement. I get the sense this represents a lot of the views of the entire publication on AI. In fact, there are some data points in this very thread.

For one, the commenters on Ars largely, extremely vocally anti-AI as pointed out by this comment: https://news.ycombinator.com/item?id=47015359 -- I'd say they're even more anti-AI than most HN threads.

So every time he says anything remotely positive about AI, the comments light up. In fact there's a comment in this very thread accusing him of being too pro-AI! https://news.ycombinator.com/item?id=47013747 But go look at his work: anything positive about AI is always couched in much longer refrains about the risks of AI.

As an example, there has been a concrete instance of pandering where he posted a somewhat balanced article about AI-assisted coding, and the very first comment went like, "Hey did you forget about your own report about how the METR study found AI actually slowed developers down?" and he immediately updated the article to mention that study. (That study's come up a bunch of times but somehow, he's never mentioned the multiple other studies that show a much more positive impact from AI.)

So this fiasco, which has to be AI hallucinations somehow, in that environment is extremely weird.

As a total aside, in the most hilarious form of irony, their interview about Enshittification with Cory Doctorow himself crashed the browser on my car and my iPad multiple times because of ads. I kid you not. I ranted about it on LinkedIn: https://www.linkedin.com/posts/kunalkandekar_enshittificatio...