At this point, it's hard not to notice the Apple Vision Pro, a brand new AR and VR device Apple announced at its developer conference.
This post does not discuss:
- Whether Apple is still an innovative company.
- Is "spatial computing" a thing (it's not, it's more of a marketing gimmick than a technical feature).
- What useful use cases will this device have (not really gaming, not really education, professionally designed, but more general use)?
- Who would pay $3,500 for a device like this?
But I'll comment on how the Apple Vision Pro is (or isn't) relevant (or not) to the big wave of artificial intelligence we're in.
No artificial intelligence at all?
I'm not the first to notice that there was little artificial intelligence in Apple's Vision Pro announcement, which is pretty obvious given the prominence of artificial intelligence in big tech companies today.
In my opinion, artificial intelligence is not absent, but implied. Craig Federighi didn't talk about the AI, but how little details like autocorrect would be overhauled.
You see, traditional autocorrect is done using statistical methods: when you don't type the word with the highest probability, it changes what you write, often with hilarious results.
But the new autocorrection will be done with machine learning that takes into account the context of what you've written so far, meaning corrections will hopefully better match user intent.
Hold on! Isn't machine learning also statistical?
Well, yes, but standard statistical methods consider the first few words, while machine learning considers, for example, what are you talking about. So, for example, if you write,