Quote:
Originally Posted by
Dick Applebaum 
I think it is to Apple's benefit to get Siri on every capable Apple device ASAP... but consistent with an orderly rollout.
The iP4S only was, likely, made because there was limited availability of the 4S. I suspect that the next device that supports Siri will be the iPad 2... then iP4 and iPad 1.
I think this is was one of their main reasons for restricting it to the 4S, but I think they may have now decided to keep it as 4S and newer. Certainly restricting it to the 4S initially helped prevent a meltdown on the server side. I know, for Nuance, 4S launch day was their single largest (by a huge margin) day for requests. If the iPhone 4 and/or iPad 2 has been included at the same time, it is very possible they not have been able to scale well enough.
Quote:
Originally Posted by
Dick Applebaum
At some point, Apple will have learned enough from the current implementation so they can do some/much of the work (of the current implementation) on the device itself without going to the servers except to handle exceptions and requests that require data or services not available on the device.
I don't think the server side will even be so reduced. There is a reason Nuance has transitioned to server supported voice for mobile instead of running native. No one has more experience doing voice analysis than Nuance and if they feel locally running analysis is not feasible for mobile (at this time) I would have to concede to their wisdom and experience.
Quote:
Originally Posted by
Dick Applebaum
It's kind of a continuous process where new Siri capabilities are added to Apples Servers (to understand and refine) -- and older capabilities are migrated to the iDevices (to improve performance and reduce server load).
Possibly, but I just don't see it being offloaded entirely until mobile processors are on par with current desktop processors. The lag caused by the network requests is nothing compared to the lag of a smaller CPU trying to do real-time analysis and AI.
Quote:
Originally Posted by
Dick Applebaum
So, I believe we'll see staggered rollout of new languages and features to an ever increasing number of devices. And as Siri features become well-understood (by Apple) and refined -- more and more of the processing will migrate to the devices, themselves.
Almost certainly, but it can never be completely self-contained nor would it greatly benefit the process to migrate the majority of the analysis to the device.
