The innovation in the UI would be in making those core communications functions as simple as possible; and improving the voice and messaging experience. This doesn't need a massive screen, but it does invite improvements in navigating address books and message stores. Feature like Visual Voice would come from a canned Apple back end you'd co-develop with each of Lucent, Nortel and Ericsson to integrate with their existing voice switches, plus the usual suspects like Comverse for voicemail.

Everything else would be left out. All of it. Really. No wifi. No downloads. No Java. (I might even do a deal with the devil and license Qualcomm's BREW under the hood.) 3G only for voice spectral efficiency. Especially leave out video and TV -- they're going to be important, but not launch-critical, and are battery-killers. SMS is the killer app outside the USA.

You'd then eat your way up into the higher-end features over time. Oh, and you'd gun to make every developing country where PC and broadband penetration is low Apple-centric. They'd make the progression path one that cuts out Microsoft. New low-end Macs would complete the range -- basically a Mac Mini welded into an LCD screen. The real long-term threat to Apple is what Nokia is doing in India, but you can't see that driving along Highway 101.

Oh, and I'm going to be contrarian and say that the closed nature of the iPhone is a feature, not a bug. Makes support costs manageable, and ensures feature integration is flawless. The iPhone as launched is not a smartphone, it's a featurephone and fashion accessory. But the touch screen will turn out to be a liability: like programming a computer with only a 5V battery and piece of wire, in being able to do everything, you end up being excellent at nothing.

Martin turns the world on its head at Telepocalypse.