Exactly, Apple is entirely too conservative to shine with LLMs due to their uncontrollability, Apple likes their control and their version of "protecting people" (which I don't fully agree with) which includes "We are way too scared to expose our clients to something we can't control and stop from doing/saying anything bad!", which may end up being prudent. They won't come close to doing something like OpenClaw for at least a few more years when the tech is (hopefully) safer and/or the Overton Window has shifted.

And yet they'll push out AI-driven "message summaries" that are horrifically bad and inaccurate, often summarizing the intent of a message as the complete opposite of the full message up to and including "wants to end relationship; will see you later"?