Curious: do you think AI assistants will (over the next few years) make human intelligence more or less diverse / high variance? A case for more diverse: they make it much easier to develop unique combinations of understanding and expression, a combinatorial explosion A case for less diverse: it means human thinking will be mediated by a tiny handful of centralized artifacts. Choices made by the frontier AI companies are going to strongly affect the way billions of people think, what they think about, how they think about it