Reading the news on your phone should be a glorious experience. You have all the stories in the world to choose from, right there in your pocket. All you need is a good way to figure out which of them are most worth your time.
Good luck with that.
Many have tried over the years to build the ideal news-reading app—one that magically brings together all the stories you’d choose yourself if you had infinite time to browse the Web every day. Some have turned to algorithms, others to human editors. There have been admirable efforts like Flipboard and Prismatic. And there have been less inspiring ones, like Apple’s charmingly anachronistic Newsstand app (R.I.P.). The ones that have come the closest to succeeding are those that set out to do something else entirely. Facebook and Twitter began as ways to keep in touch with friends. Yet they’ve become the de facto newsstands of the mobile age, with Facebook in particular dominating the flow of traffic to news sites across the Web.
Addictive as they are, Facebook and Twitter weren’t built for this role, and it shows: Spend a while perusing your feed, and you’re more likely to come away amused, disgusted, outraged, or annoyed than enlightened or informed. It would seem, then, that the field remains remarkably open, after all these years, for someone to come along and build the clutter-free Netflix-of-news that we’ve all been waiting for. In the process, maybe it could rescue serious, original journalism from the hellish click-chase that social media has wrought upon the media.
If anyone can do it, perhaps Apple can.
Its Apple News app, launched Wednesday with the release of its new mobile operating system, is the latest and perhaps best-funded attempt to deliver readers a selection of stories from around the mediasphere that’s tailored to their specific interests and reading habits. (Disclosure: Slate is one of about 50 publications that signed on with Apple to place their content on Apple News prior to its launch. Many more are lined up to join them in the months to come.)
Like Apple Music before it, Apple News enjoys a distinct advantage over third-party competitors by virtue of coming preinstalled on every iPhone and iPad running iOS 9. Also like Apple Music before it, the app immediately impresses with its appealing interface—and gradually disappoints with its as-yet-unrealized potential.
Open the app, and you’ll be prompted to select some publications and topics you’re interested in. Oblige, and Apple will reward you by populating the default “For You” tab in the app with a cleanly arranged stream of recent stories culled from the outlets you’ve specified.
There are four other tabs—“Favorites,” “Explore,” “Search,” and “Saved”—but “For You” is the default, and its tailored recommendations form the core of the app’s offerings. So if Apple News is to succeed where everyone else has failed, those recommendations had better be good.
They’re not. Not at first, anyway.
My first batch of personalized picks began promisingly enough, with an attractively illustrated essay by New York’s Jonathan Chait on how the world is finally getting serious about climate change. Presented in a larger format than the other stories in my stream, it appears to have been Apple News’ version of a cover story.
That’s a good idea, if so. An enlarged lead story provides the instant focal point that is a staple of Web and print design but largely absent from social media feeds. But I’ve yet to see another story presented in this format on Apple News, which suggests that this feature may be at least partly a product of human rather than automated curation. Apple has said that both factor into its recommendations, but declined to specify their respective roles.
What came after that lead story was less impressive. Next in my feed was an ESPN news brief about an Atlanta Falcons player getting fined $23,000. For what, I couldn’t say, as the story description cut off and I didn’t take the time to click. I like sports and had selected ESPN among the publishers I’m interested in, but I have no particular interest in the Falcons. Even if I did, this story didn’t feel like the sort of big news or juicy tidbit that a good ranking algorithm would put second in anyone’s feed. Also ranking highly in my Apple News feed, but quite low on my personal priority list, were:
- a Vox review of a new NBC show I’d never heard of
- a Slate post on the Kardashians and #thighbrow
- an Economist infographic on emerging-market vulnerabilities, which didn’t render properly in the app
All of these stories came from publications I’d selected in the Apple News onboarding process. Yet I can’t imagine any of them appearing in my Facebook feed, even though I’ve also followed those outlets on Facebook. It isn’t just that the subject matter doesn’t align with my interests. It’s that they aren’t stories that made a big splash, either on social media or their own publications’ home pages. Yet here they were, among Apple News’ very top recommendations for me. How Apple chose them, I have very little idea. The mechanisms that drive its algorithm are opaque.
It would be a mistake, of course, to write off Apple News based on these first impressions. As I’ve explained in the past, recommendation algorithms rarely work well when you first use them. Rather, they’re designed to throw a bunch of stuff at you and see what sticks. In Apple News’ case, that means looking at which stories you actually read, and which of those you bookmark or mark as a favorite. They then use that feedback, from both you and their broader user base, to continually refine their picks.
Indeed, Apple News’ suggestions have already begun to improve a bit in the 36-plus hours I’ve been using the app. It quickly noticed that I’m likely to read stories about self-driving cars and fed me a bunch more of those. It also noticed that I read a lot from Slate (surprise!) and began bombarding me with the work of my colleagues. It has not yet figured out, however, which sorts of Slate stories I’m mostly likely to be interested in. At this point I’d fare far better simply by opening the Slate app or website, where the top stories have been hand-picked by our sentient human editors. But I wouldn’t be shocked if Apple’s algorithms surpass them soon (sorry, editors!).
Most encouragingly, a few of Apple News’ early picks turned out to be fascinating stories that I would have been unlikely to encounter in my daily rounds of Facebook, Twitter, Techmeme, Reddit, Hacker News, and the New York Times. An NPR feature on the outlook of ordinary residents in the Chinese city of Luliang, a “boomtown gone bust,” taught me about a place and a topic that I wouldn’t have sought out on my own. It also offered a brand of unsensational, on-the-ground reporting and nuanced analysis that is sorely underrepresented on Facebook, Reddit, and the like.
In contrast, I would very likely have seen on Twitter or Prismatic at least one story about an impending collision between supermassive black holes in a distant galaxy. Yet odds are the one I stumbled across would not have been half as thoroughly reported or deftly explained as Dennis Overbye’s stellar New York Times article, which carried the modest headline, “More Evidence for Coming Black Hole Collision.” That’s because, on social media, clickable headlines often trump substantive journalism. (Sometimes, of course, they can go hand in hand, but the New York Times and other old-media stalwarts have yet to master that art.)
By giving weight to a story’s source, as opposed to its “shareability,” Apple News has an opportunity to serve as a sort of corrective to the viral news machine that has lately threatened to reduce online journalism to a manipulative headline-writing contest. It is not alone in this endeavor: Snapchat’s Discover tab and Facebook’s ballyhooed Instant Articles are taking a similar tack, working directly with prestigious publishers to showcase some of their best work in a format that looks pretty and loads instantly within mobile apps that are already widely used. Unlike Facebook’s News Feed algorithms, that’s not a process that can be easily gamed.
These projects, which fall under the jargon-y umbrella of “distributed content,” have provoked handwringing within the media as to the wisdom of publishers allowing tech and social media companies to host their content (and, in most cases, take a cut of their ad revenues). I’ve been among the handwringers, although I’m now more optimistic than most. But at a time when the rise of ad-blockers threatens to puncture online publishers’ business models, distributed content is beginning to look like a lesser evil.
But it isn’t enough for publishers to like Apple News, Snapchat Discover, and Facebook Instant Articles. Readers have to find them compelling too. And the early returns have been mixed. BuzzFeed, for one, is finding Snapchat Discover to be a boon, with more than 20 percent of all its traffic now taking place within the Snapchat app. But after the initial hype, Facebook’s Instant Articles have gone quiet, for reasons that the company has not yet made clear.
In order for Apple News to succeed, it will need to be much more than a fancified RSS feed of stories from each user’s favorite sites. For all its flaws as a news-reading utility, there is a reason people keep opening Facebook’s app every day on their phones and tablets. The social network has devoted enormous effort over the years to figuring out exactly which types of data its algorithm needs, and what weights it ought to assign them, in order to show people the posts they’re most interested in. And that has turned out, in many cases, to be quite different from the sort of content they think they’re interested in.
Apple News holds immediate appeal for newshounds because it promises to restore some human control and judgment to the curation process. In another contrast to Facebook, Apple promises that its app will respect users’ privacy by refining its recommendations based only on their behavior within Apple News, rather than tracking their habits and preferences willy-nilly.
That protective silo is a strong selling point—in theory. But it could also be Apple News’ biggest drawback.
As much as we clamor for a calmer, sleeker news reading experience, it’s not clear that’s what most people really want. That is, they may actually prefer to get their news sprinkled across a social feed, so that it comes packaged with entertainment, interaction, and a soupcon of persona drama. It’s easy to forget that print newspapers, stodgy as their reputation has become, were pioneers in this regard, nestling their investigative journalism alongside gossip columns and funny pages. If Apple news fails to reshape the industry, as similar apps have failed before it, it might not be a failure of execution. It might be that the concept of the ideal news app is fundamentally flawed.
These questions aren’t even worth pondering, of course, if Apple’s product doesn’t quickly get better at learning what people actually want, as opposed to what they say they want. Apple might soon find that there’s a good reason Facebook pays so much attention to clicks, likes, and shares, and why it has gotten so greedy when it comes to data on users’ behavior. It may be that you simply can’t build a compelling recommendation algorithm without it.