“For more complex requests that require more computational capacity, Apple Intelligence can use Private Cloud Compute, which extends the privacy and security of your Apple products like iPhone into the cloud to unlock even more intelligence. Private Cloud Compute uses larger, server-based models powered by Apple silicon.“
I think the bigger issue is that even the "cloud" stuff seems pretty stupid. Meanwhile while I wouldn't really trust any of these companies very much in actuality, it's not like nobody else is doing "edge" AI: Google is basically treating Gemma models like A1 steak sauce and pouring them on everything. (And although I don't know that any of those use cases really are particularly interesting, Gemma itself can be pretty impressive IMO.)
https://support.apple.com/en-gb/guide/iphone/iphe3f499e0e/io...:
“For more complex requests that require more computational capacity, Apple Intelligence can use Private Cloud Compute, which extends the privacy and security of your Apple products like iPhone into the cloud to unlock even more intelligence. Private Cloud Compute uses larger, server-based models powered by Apple silicon.“