Machine learning is everywhere these days. It’s a popular buzzword for good reason. Every major tech company (AndPlus included!) is building complex software to solve everyday problems using artificial intelligence and machine learning.
At WWDC on June 5th 2017, Apple introduced Core ML, a brand new machine learning API for all iOS developers to bring to their apps. The primary use of Core ML is for mobile devices to learn and execute common tasks as quickly as possible. Your iPhone, iPad, Apple Watch or iPod Touch (does anyone have these anymore???) can and will benefit from this new tech starting this Fall when iOS 11 launches. Apple has already used machine vision to index the Photos app and even face detection within the camera app. They’ve used machine learning to have Siri provide you with a heads-up on traffic before you head to your next meeting and text analysis for keyboards.
Core ML will support all the modern machine learning libraries and tools you’d expect. This will allow engineers to build out complex neural networks, vector machines, tree ensembles as well as linear models. It will also automatically convert Caffe, Keras, XGBoost and several SciKit Learn model types.
We knew this move was coming because late last year Apple released a paper that focuses on artificial intelligence. The research was written by six of Apple's AI experts on how to effectively train AI models and focuses on two primary uses, where a user is looking and hand gestures such as waving. As the company grows more and more prominent in the AI industry software developers will keep a close eye and create their own takes on what a machine learning application should be.
One of the most Apple-esque parts of all this news is their focus on security. While competitors in the AI market are doing all they can to hoard user data for potentially malevolent use, Apple is reassuring all users that their data is inaccessible to anyone but the end user.
We've been hard at work building our own machine learning software here at AndPlus and are beyond excited to get our iOS team started on a Core ML application this summer, is there anything you'd like to see? Let us know!