The Car-Based Truck

A few weeks ago, Steve Troughton-Smith and Guilherme Rambo found a software update that contained the code for Apple’s upcoming iMac Pro. Confirming some earlier rumors, it appears that the iMac will have both an Intel processor, as previously announced, and an A10 — the Apple-designed processor powering the iPhone 7. Inevitably, someone will soon bring up the famous quote from Steve Jobs about how traditional desktop computing is a truck, and iOS devices are the cars that everyone will use from now on. In this case, it seems even more appropriate than most. 

In 1979, AMC introduced the Eagle. It was ungainly, slow, and ugly — yet was a popular car for most of its eight-year production run. Essentially, it was a compact car with four wheel drive and a 3” lift — ordinarily not enough to cover up the underlying flaws inevitable in a car built cheaply by a dying company. But the Eagle itself started a revolution in transportation — it was the first car-based truck.

I can guarantee nobody ever used their Eagle like this

I can guarantee nobody ever used their Eagle like this

Until the Eagle’s introduction, the only way to get a family car that could also handle heavy snows, dirt tracks, and withstand the rigors of even moderate abuse was a heavy, slow, noisy, and crude truck-based ‘wagon.’ Suddenly, it was possible to take your family car to the grocery store, and make it home even when the weather had turned completely sour. This was a revelation at a time when trucks (and their family-wagon derivatives) had barely advanced in the previous thirty years — one of the headlining features of the 1973 Chevrolet Suburban was a fourth door. By taking advantage of the advances that had happened in the car sector, the Eagle provided creature comforts with an acceptable compromise to its off-road abilities. 

A few years later, Jeep (then a sister brand to AMC) introduced the XJ Cherokee which — rather than being based on a previous car body — was purpose-built as a car-truck hybrid. It was built for over 30 years, selling more than 2.1 million units, and created an entire segment of the automotive market that hadn’t existed. Today, a plurality of vehicles sold are crossovers, the name that was coined to represent these blends of heavy-duty and approachable transportation. 

The iMac Pro, operating under this somewhat-tortured analogy, is the AMC Eagle. It’s the first Mac to be car-first, run by the A-series processor which then boots the legacy Intel hardware. For now, it will (like the Eagle) only provide a few benefits to security. Eventually, this will become the way all mainstream computers are built, providing the power of a truck with the ease of use and simplicity of maintenance of a car.  

This iMac, along the Touch Bar-equipped MacBook Pro, may not be a lasting design. Like the Eagle, they might turn out to be a mere hint of the future; with time, they’ll fade into history as the awkward predecessors of a new kind of computer. Like the automotive industry, the line between light-duty, everyday tools and powerful special-purpose machines will begin to blur. Today, you can buy an eighty-thousand-dollar Ford SuperDuty pickup truck with more features than the most luxurious cars from the’90s. Similarly, desktop computing will begin to take on some of the characteristics of mobile devices until it is difficult to tell where one ends and the other begins.

Some Personal News

I'm thrilled to share that I've accepted a position with Oscar Health here in New York, working on their internal product team. Here's to helping build the future of health insurance (sorry, Mom & Dad!) 


Hi Oscar!

Mixing at Scale

Yesterday, Xiaomi announced the Mi Mix. MKBHD has, as usual, an excellent hands-on summary of the news: 

A brief bit of background on Xiaomi, if you haven't heard of them, is in order. It's a relatively young Chinese electronics manufacturer, which has risen rapidly in prominence because they sell their phones at essentially zero margin (this will be important later). They also control pricing by only building phones when there is demand, meaning there is rarely enough supply of their devices — many models are often sold out for months at a time (ditto). They've even poached some of Google's most senior Android product folks in order to build Android-based devices with a level of engineering & attention to detail not unlike Apple's. With a strong team, lots of venture capital backing, and more than a little bit of grit, they've become the dominant manufacturer in their home market, and are rapidly expanding into developing markets around the globe. It's not exactly surprising that they're ready to show off their engineering chops a bit, with a splashy, envelope-pushing device.

Their newest flagship (though billed as a "concept", it replaces the one-month-old Mi 5s Plus) is the aforementioned Mi Mix — a phone which is simply incredible, and looks straight out of a near-future sci-fi film. The display covers almost the entirety of the phone's face, and the chassis is made out of a highly polished ceramic. It's hard to believe this phone can even be made in 2016, much less sell for just over $500 (U.S. pricing has not been announced yet). Since it's likely being sold break-even, or even a slight loss, we know it costs close to three times as much as an iPhone does to make, which is incredible. After factoring in engineering & marketing costs, this phone will certainly cost Xiaomi a fair amount of money. Then again, as a venture-backed growth-stage startup, they can afford to take the hit. It's likely that it won't even cost them that much because of a few words in the press briefing — "limited availability."

When Apple says that there's limited quantities of the iPhone, it's typically a sign that they are going to sell every phone they can make, and then some. With a massive global logistics operation at full tilt, just over 13 million iPhone 6s were sold over the first weekend of availability, despite selling out nearly everywhere (Apple did not release first-weekend sales numbers for the iPhone 7). Each year, over 200 million iPhones are manufactured and sold — a daunting task unprecedented in human history. Nothing of equivalent complexity has ever been built in numbers anywhere close. For some excellent writing about the challenges of making the iPhone at scale, check out Greg Koenig's excellent, thoroughly-researched blog posts on how Apple manufactures at scale — in particular, his post on ceramic iPhones inspired this one. The difficulty of pulling this off starts to explain why my iPhone 7 doesn't have an incredible edge-to-edge display, or a slick hidden speaker & proximity sensor, or any number of the amazing design touches in the Mi Mix. 

Xiaomi's history of building their products in extremely small batches, and allowing them to be 'sold out' on their web store for periods of several consecutive months offers clues to how they've managed to pull this off. It's actually fairly simple, in that they haven't, and they've pretty much told the world that fact. It's repeatedly described as a "concept" with "limited availability" in their statements to the press, and they clearly don't intend on selling anywhere near the numbers that their mainstream devices do. So far, they've taken just over 800,000 reservations according to the Mix' product page, and I'd be willing to bet many of those customers won't actually be able to convert a reservation to a product order.

The Mi Mix is a feat of engineering, that is not in doubt. As a product, it's a triumph, and if I were a Hollywood prop master I'd be struggling mightily to get my hands on a few for any sci-fi project in the works. But it's certainly not an example of what the iPhone 7 should be. It's doubtful that the iPhone 7 even could look like this, at least not if anyone wanted to actually buy one and use it. The Mi Mix is a peek into the future of the smartphone, a glimpse at what every manufacturer — yes, even Apple — has sitting under a black cloth in their design & prototyping shop. But it isn't more than that. I'll likely never see one in person (but if you're willing to let me play with yours, I'll buy you a beer). A few years from now, it'll be a curiosity because it pushed towards the inevitable all-screen future. We've been working towards it for some time. But hey, at least Xiaomi is first with a bezel-free phone, right?

The tweet that inspired this post:

Why Can't I Just Watch Baseball?

Last night I was frustrated. Not by the Cubs, who are making a historic playoff run — and who I'm hoping will win their first title since the fall of the Ottoman Empire — but by my inability to watch it. I tried three different, entirely legal, ways to watch the Cubs-Dodgers game and was stymied each time. You'd reasonably assume that postseason baseball is widely viewed enough that it'd be available over-the-air in the largest media market, but no dice. Due to a complicated and ridiculous TV deal, Fox owns these games, but chooses to air them on their new sports channel, Fox Sports 1 — thus forcing cable providers to include it in basic cable packages. As a card-carrying millennial, I don't have cable, preferring the extra $64.95 per month to spend on 🎮 and 🍺. It turns out that the other two ways I have of watching baseball also were blocked. Thanks to Jon Legere's insanity and an old T-Mobile iPad, I have a subscription to except — surprise! — that's blacked out until the World Series. And then we come to the message above. What Fox is saying (via TiVo and Verizon FiOS) is that even though I've paid for this channel, and want to watch it live, with commercials, I'm not allowed to because... something.

We should look to the NFL for a cautionary tale. Despite a (rather impressive) streaming deal with Twitter that has revitalized a flagging Thursday-night product, ratings are down 11% year-over-year. As commentators and bloggers scramble to find excuses for an unprecedented ratings drop — suspensions, injuries, even Hillary Clinton — the simple fact is that it has become more difficult to watch major American sports leagues than most other forms of entertainment. In an era when anyone can plop down on their couch and binge on Netflix' ($9.99/month) or HBO's ($14.99/month) high-quality content whenever they want, in HBO's case often before an episode airs on linear TV, why should they sit through endless commercials to watch an entertainment product that isn't better? During a broadcast earlier this year, a sentence was uttered by a broadcaster which left me speechless: "due to NFL rules, we're not going to be able to show you the end of that game. Instead, here's [our anchors] with the [team] postgame show." Even the digital products put out by the leagues is mediocre — I'm not going to pay double the cost for baseball ($25/month) or quadruple the cost for football ($40/month) to run into onerous blackout restrictions, time delays ($40/month doesn't even buy you live-streaming, only time-shifted viewing later), and general disrespect for the fans. Yes, teams are making money hand-over-fist, but this won't last. 

Respect for your customers, fans, users, members — whatever you want to call the consumers of your product — is essential to building a sustainable, long-lasting business. We're long past the era where football was the only entertainment available on Sunday afternoons, or everyone would tune in to watch a sports championship; The fans that do tune in are clearly more dedicated, and more interested, than most of the prior (much larger) audience. Yet this loyalty is met with indifference by the industries that are selling against the fans' eyeballs. Making the content available to paying customers wherever and whenever they want isn't benevolence, it's table stakes in the digital era. Without a significant rethinking, ratings will continue to decline as people give their attention to content that appreciates their loyalty. Meanwhile, the NFL recently renewed a deal that prevents them from offering live streams of most of their games, meaning they cannot even start competing in the streaming space until 2022.

We Asked For This

Each September, Apple releases a new version of iOS to a chorus of cheers from the average user, and a series of frustrated blog posts by the more technically-inclined. Every crash, sync failure, or edge case is held aloft as evidence of Apple's slow decline into irrelevancy. Jokes reference Marco Arment's (since-retracted) post on the increase in weird bugs in Apple's software. Screenshots abound of weird behavior. Sarcastic tweets are retweeted, quote tweeted, debated, and debunked. Flame wars spread across the internet, as people either argue that Apple and its products are in a decline akin to the Roman Empire after the sack of Rome or vehemently deny there's a problem like Neville Chamberlain in 1938.

Those who argue that there are far more shipped bugs in Apple's (or anyone else's) software aren't wrong. It's an inevitable result of the tremendous increases in what is possible on iOS over the past decade. 10 years ago, iPhone OS 1.1 had just arrived, bringing with it the iTunes Wi-Fi Music Store. Today, we're awaiting the launch of iOS 10.1, which will let the iPhone use a 3D map of the world to simulate the effect of a much bigger lens. With increased software complexity comes an increased potential for bugs, and makes each user's experience more unique. With no two people using their phone in exactly the same way, Apple's now building an OS for one billion different edge cases. It's inevitable that each one of that billion users will hit some untested, undocumented behavior that just doesn't work right.

It's not Apple's fault that iOS has gotten this elaborate, it's that we wanted it to be. From the very early days, we — the tech-y users, the developers, the fanboys — wanted more. Even features like copy and paste were added at the behest of users, leading slowly, inexorably to the intricacy found in iOS 10. Add to this the pressure caused by the millions of non-tech-obsessed folks who use iOS devices as their only computer, and it's a recipe not for disaster, but for imperfection. This isn't necessarily a problem, even, as iOS has become a full-fledged computing platform in its own right, with all the great and terrible things that label implies. So remember, the next time some widget acts wonky, or a notification behaves oddly — we asked for this.