Anthropic's 2026 Public Benefit Report, released a few days ago, is interesting reading. First the foreword. This is something I brought up in my recent article on this. Language like this is, unsurprisingly, the same thing as Dario Amodei writes in his various essays. But the people in charge of Anthropic haven't chosen a life of Shakespeare and Tuesdays in the kitchen with family. They've chosen to dedicate themselves to changing the world, i.e. to lean in to thymos. That's why I find rhetoric like this unconvincing. Their revealed preference is not for a world like this at all. Do they see it as bravely sacrificing the possibilities for people like them to be satisfied for those of the average person?

To clarify - Anthropic project AI making human labour redundant. Therefore there's no way to humans to change the world any more.

Genuinely interesting question though whether people in the humanities are under threat from AI. Of course, AI can write a humanities paper. But people would have to value this more than they would human papers so much that the latter becomes valueless. Due to humanities being more subjective, and in many cases political, I'm skeptical this would happen. I think AI is likely to transform things so much that it's hard to predict how anything turns out, but in a world where we're still around, I think humanities will change less than more mathematical subjects.

Assuming Pangram keeps working, I would think that a strong, verifiable taboo on humanities papers being written by AI develops.

A sensible country would have something like an immigration safety board, which would employ anthropologists to determine which immigrant groups were likely to be the greatest risk to the native population, and restrict their entry. In this case it's not complicated. The graph reflects the same realities that a female tourist would experience if she travelled to say, Egypt vs Vietnam. There's a huge amount of anthropological work around into these cultural differences regarding sex and gender roles, and what this means for sexual offences. People have spent a lot of time over the last few years talking about 'rape culture' etc. It's obvious to anyone not ideologically blinkered that the most extreme examples of rape culture exist in the places at the top of this chart, and as we all agree that rape culture is a bad thing, perhaps we should avoid importing any more of it.

E.g. from the first article, the idea of locking away women for their safety.