First things first

A few months after hiring John Giannandrea away from Google to head its up its AI strategy, Apple has announced Giannandrea will now lead a newly created team comprised of the Siri and Core ML teams. Core ML, if you're unfamiliar, is Apple's toolkit for iOS developers who want to add artificial intelligence / machine learning (hence the "ML") into their applications. Siri, obviously, is the digital assistant that started this whole craze that's now largely dominated by Amazon's Alexa and Google's nameless voice from the machine (technically, Google Assistant).

I wrote at the time (in the first link above) that Apple probably hired Giannandrea for his skills in the knowledge base/graph area (he came to Google from Metabase) more so than for his artificial intelligence knowledge, and this news really cements that viewpoint. The logic there being that Google's prowess in search, which was bolstered by Freebase database/knowledge graph it acquired along with Metabase, is arguably more important to the success of Google Home and the entire Google Assistant product set than accurate speech recognition. By putting Giannandrea in charge of seemingly more product-focused AI team, Apple is acknowledging that it has some catching up to do.

You could also argue that Apple is at a disadvantage with Amazon and Google because it has fewer developer hooks. Yes, iOS is extremely popular, but so is Android and, if you're building for the smart home, so is Amazon Echo/Alexa (Apple's Home Pod does not appear to be a major player). However, Google and Amazon also have cloud computing businesses that push developers to their platforms by giving them some pretty powerful AI tools to play with. They also have a lot of data from their various businesses.

As I wrote in one of my first-ever posts for ARCHITECHT (way back in January 2017), those two companies are working very hard to develop flywheels where cloud services, data and consumer devices can continuously fuel each other.

If this is the way that consumer technology and consumer-grade AI is headed, Apple looks like the odd man out. Apple is dominant in smartphones and tablets as a manufacturer, but iOS as a whole does not have the breadth of the platforms that Amazon and Google are building. And although it's one of the world's most valuable companies at any given time, (to my untrained economic eye) Apple's success seems intrinsically tied to the success of the iPhone. Selling hardware is just a side business for Amazon and Google -- heck, Apple even pays them to store iCloud data.

But in the age of data privacy and cybersecurity pandemonium -- two areas where Apple has a good track record and appears to be doubling down -- Apple also has an opportunity to make the iPhone and, by proxy, the broader iOS ecosystem, even more valuable. I obviously have no idea what Giannandrea's or Tim Cook's visions are, but it's appealing to envision a world where not only is most AI processing done on the device, but where most of the knowledge lives locally, as well. Imagine a smart home that doesn't go dumb when the wifi drops, or where every device doesn't even need an internet connection. The iPhone is the locked-down the brain that pushes updates and phones the cloud when necessary.

The actual AI in all of this -- the computer vision, speech recognition, NLP, etc. -- is the easy part. But if I were Apple and realized I couldn't hang in the cloud, I would be looking for a way to cut it out of the picture.

Read and share this issue online here.

AI and machine learning



Cloud and infrastructure



Sponsor: Replicated

Data and analytics