April 4, 2018
A few important things came to mind while I thought about this news. First, AI at Apple is so much larger than what people experience as ‘Siri,’ from the many other features like App Suggestions and Handoff which are marketed as Siri features, to CoreML, image recognition, and auto-focus on the Camera app. Saying they hired Giannandrea to “boost Siri” is like saying you charged your phone battery so you could send a tweet. There is much more than can (and should) be done here than just making the virtual assistant stop saying silly things when it doesn’t understand you.
Giannandrea will report directly to Cook, meaning he will have influence over the entire product and services line. This could be very good as far as making the wide uses of AI on Apple platforms better, but will also mean that his influence could be hard to identify in the first couple years. If this hire works it will be a long game, not a short one. Which also means that Giannandrea won’t be responsible if we actually do (or don’t) get the Siri improvements long desired at WWDC this year. It is far too late in the iOS 12 timeline for a new player to enter the game and meaningfully shake things up.
Finally, Google and Apple have vastly different cultures and ideologies around AI. I have to believe that Giannandrea wouldn’t have taken this job if he weren’t fully on board with committing to privacy and leaning on on device processing whenever possible. These are core tenants of Apple’s approach in this area, and it would be a damn shame if Apple changed course by bringing a Google guy on board. But even so, there is always a chance that big changes like this don’t end up working out (remember when Chris Lattner was a VP at Tesla?) Google and Apple are such different beasts here it’s easy to imagine that the transition doesn’t go completely smoothly. I’m hoping that’s not the case, mostly because it would show that Apple approached what is unquestionably a big ‘get’ in a rushed manner, but it’s always a possibility.