What are everyone’s thoughts on the future of privacy when it comes to communications between a device with AI Agentic features switched on and a device with them switched off or stripped entirely like de-Googled devices and the like?
If they’re fully on-device then it’s not an issue I think. Apple’s Private Cloud Compute looks interesting but I think on-device processing is going to be the king of private AI.
I agree on-device processing is going to be the pinnacle of private AI but mostly private for the user of that device. To say that same user has an expectation of privacy by communicating with an iPhone using cloud based AI irrespective of Apple’s Private Cloud Compute is a stretch IMO.