AI

Apple finally makes an intelligent foray into AI

Apple is cutting hundreds of workers from its self-driving car projects, according to a report from CNBC.

The news comes at an inflection point for the Silicon Valley company. iPhone sales are slowing, and cars are clearly not going to be the next big thing to come out of Cupertino.

However, the autonomous car reshuffle is a chance for Apple to become a real player in artificial intelligence.

That is good news for the company. And it’s huge news for shareholders who have been worried about its otherwise depressing lack of innovation.

Apple’s foray into self-driving cars always seemed out-of-character. It was a “me too” move for a company with a long history of market leadership.

Some will argue that Apple is rarely first to anything really big and new in tech. But fans and foes alike tend to agree that the company is good at reimagining the important bits and pieces that show the way forward.

Silky-smooth touch interfaces with haptic feedback. Fingerprint sensors to make smartphones more secure. Putting an EKG heart rate system inside the Apple Watch. Features like these helped justify the iPhone’s growing price tag. Plus, they are beyond cool.

Related post: Heartfelt Watch 4 ushers in new time for Apple

Apple started its autonomous car project five years ago. In 2014, CEO Tim Cook assigned a former Ford engineer to explore the possibility of building an electric, autonomous vehicle.

Project Titan, The Wall Street Journal reported, was a 1,000-person team holed up in a location a few miles away from Apple headquarters. By 2018, a court filing revealed the project had swollen to 5,000 people.

I get it. Cutting staff at any Apple project is a dog-bites-man story. It is ordinary and feeds into the narrative that the company is flailing.

However, investors should see the move for what it is …

Apple is not killing its AV program. Two hundred engineers are being moved to other projects where the focus is machine learning and other initiatives.

In April 2018, Apple hired John Giannandrea away from Google. During his eight years with the search giant, Giannandrea pushed AI integration into every Google product. Eventually he took over the machine intelligence, research and search teams.

In Cupertino, his star has risen dramatically. He was quickly promoted to the executive board. His portfolio now includes AI, machine learning and Siri, Apple’s digital assistant.

Siri is perfectly emblematic of AI development at Apple. Despite getting to the market first, Google and Amazon quickly leapfrogged Apple’s efforts. Building better algorithms requires lots of data, and the computers to make sense of the information. And Apple lacked the edge of its competitors.

Google Search, Maps, YouTube, Amazon.com, Echo, Fire TV and other properties provide a wealth of useful information. Google Cloud and Amazon Web Services are among the most powerful cloud-computing networks in the world.

However, Apple’s business model depends on selling hardware. Managers neglected AI software development. They allowed it to become an unfocused, fragmented mess.

The ascendency of Giannandrea means the era of neglecting AI is over. Shifting hundreds of engineers his way means Apple is getting back into the game.

The timing could not be better …

During the past decade, the iPhone has become a truly impressive computer. Proprietary microchips are capable of processing private information on the device. With a network of 1 billion devices in the field, and more focused machine learning running in the cloud, the company is setting on a fledgling, but gigantic, edge-computing network.

Cook understands this advantage and he is ready to exploit it. He announced December 2018 that Apple is spending $10 billion during the next five years to build data centers in the U.S. He is also blasting competitors with different business models.

There is a good measure of hypocrisy involved. While Cook criticizes competitors for what he deems surveillance, Apple continues to play nice with China, an actual surveillance state.

Related post: Apple’s pious privacy pledges ring hollow

Investors should focus on the bigger picture. Apple is trying to define the future of networks.

We are currently in a world where most of the computing occurs on-the-fly, at the edge of the network. In this context, iPhone development during the last 10 years makes perfect sense.

And Project Titan works, too. A self-driving car, at least the computing hardware and software, is a perfect edge-computing application.

Apple shares were rocked at the beginning of the year following an uncharacteristic sales warning. However, at 11.7x forward earnings, 2.6x sales and 15x free cash flow, the stock is cheap by historic standards.

Investors should use weakness to accumulate longer-term positions.

Best wishes,
Jon D. Markman

Leave a Reply

Your email address will not be published. Required fields are marked *

Comments 1

Bob Schubring January 28, 2019

What only Ford so far has had the guts to come out and admit, is that the way highways handle information, just plain sucks. Cars, whether human-driven or robot-driven, blunder autistically down a path, ignoring all communication and data-sharing options, until the path becomes blocked. Then they sit there, stupidly waiting for the path to reopen. Particularly ironic, given the millions being spent on machine vision so that robot drivers can see where they are,

Nobody communicates, in an effort to route traffic via the quickest path. Even in a world agog over 5G cellphones that enable humans to show each other live streaming video, one would think somebody would consider that a line of robot drivers ought to send back a picture from the robot at the front of the line, revealing just why the hell traffic has stopped moving. Alas, the great minds of Sand Hill Road missed that one entirely.

A recent Ford patent, issued to MacNeille et al (http://www.freepatentsonline.com/y2018/0082590.html) discloses how robot-to-robot communication could bring about teamwork that diminishes traffic congestion, and a cryptocurrency-based auction system is proposed, that enables slow drivers to be paid to move out of the way. The whole thing resembles an operation Ford began with Conrail in the late 1970s, called “Tracing, Diversion & Reconsignment”, which enabled the railroad to act as a rolling warehouse for auto parts enroute from factory to factory…cargo that wasn’t needed until 9 hours from now, got diverted onto a slower train and arrived an hour early, instead of 7 hours early. On highways, if high-priority loads got through on time, while cruising Ubers waiting for a fare to pop up, cruised where they didn’t block traffic, it would be much easier to get through urban traffic.

Getting autos to be less autistic and more interactive with their surroundings, is a key insight to the entire issue of traffic flow. What began at Tesla Motors as a way for parked cars to drive themselves to a recharging station and then drive back to meet their owners, took a dogleg through the Uber company’s fundraising machine, and finally emerged as another Better Idea from Ford, that actually fixes a real-world problem: Congested cities.

Apple Computer has a nasty habit of never being first to market with a new product, but first to market with a great product. They didn’t launch the first Atari 8-bit machine with a 6502 Processor. They didn’t launch the IBM PC. They didn’t launch the Sony Walkman. They didn’t launch the BlackBerry Smartphone. But they have a history of learning from what frustrates users about new technology and then building elegant solutions that consumers love.

Can AI resolve the next generation of traffic problems, after robot drivers learn to communicate and hackers start figuring out how to hack the system and make pests of themselves? Probably. And probably we need Highway 2.0 before Apple can figure out how to create Highway 3.0.

Reply