Why Apple had to change its maps

They may not be great, but they were inevitable.

It's been just under a week since Apple released iOS 6 for download, which contained an uncharacteristically unpolished new version of the built-in Maps app. So why did they do it?

They needed to switch provider

Prior to iOS 6, the information in the Maps app had been provided by Google. It's hard to remember now, but way back in 2007, Apple and Google were best of friends. Google's CEO Eric Schmidt sat on the Apple board of directors, and the two companies operated in completely different spheres. As a result, it made sense to ship the original iPhone with a ton of Google's products built-in.

As well as Maps, there was the YouTube app; Google was the default and only search engine in Safari; and it was the only email provider which was built-in to Mail. Gradually, as the two companies have come into conflict, largely from Google's attempt to enter the mobile sphere with their Android OS, Apple has dialled down these commitments, and iOS 6 is the largest step away yet, with the removal of the YouTube app and the change to the Maps.

But it's not just the case that Apple doesn't want a competitor on their phones. Google played its part in forcing Apple's hand.

Even by 2011, the launch of iOS 5, it was clear that Apple's implementation of Google Maps was falling behind the cutting edge. On Android phones, Google had implemented two key features which it had declined to make available to Apple: turn-by-turn driving directions, and vector-based map tiles. The advantage of the former is self-evident, particularly in car-obsessed America. The latter, while more of a technical change, allows the maps to use significantly less bandwidth, as well as making zooming clearer and smoother.

Neither of these features were in the original contract, signed for the launch of the iPhone, nor, it seems, were they included in the renewed contract, which was signed around May 2011. If Apple wanted them, they had to renegotiate with Google – and the terms the search giant wanted probably weren't ones Apple would accede to lightly.

The Wall Street Journal reveals some of the demands of both sides (£):

Apple executives also wanted to include Google's turn-by-turn-navigation service in the iPhone—a feature popular with Android users because it lets people treat their phones as in-car GPS devices. Google wouldn't allow it, according to people on both sides. One of these people said Google viewed Apple's terms as unfair.

Google executives, meantime, also bristled at Apple's refusal to add features that would help Google. For instance, Google wanted to emphasize its brand name more prominently within the maps app. It also wanted Apple to enable its service designed to find friends nearby, dubbed Latitude, which Apple refrained from doing, said people on both sides.

The inclusion of Latitude is less of a "feature" than it may seem. Although the friend-finding service is moderately popular – Apple has launched its own version, called "Find My Friends" – it also exists as a handy way for Google to harvest location data.

Anyone who knows much about Apple knows that those terms are not the sort that the company usually accedes to. Its aesthetic is minimalist, and its protection of customer data is legendary – its refusal to give subscriber records to magazine publishers, for example, is the reason why the FT cancelled its iOS app in May this year.

And even if Apple had agreed to those terms, the trajectory they and Google are on would only delay the switch temporarily. Tying a key service to your most important competitor is not the makings of a good business. If Google demanded more prominent branding this time, who's to say they wouldn't demand the right to serve ads next time? Or require a Google login to use advanced features?

The real question isn't whether it made sense for Apple to switch providers, but whether switching was a move which made sense for the consumers, or just for Apple themselves. Have they, as Anil Dash put it, "put their own priorities for corporate strategy ahead of user experience"?

That's debatable. Certainly, the benefits of an incorrect map are minimal. But the downsides to the change are short term – nobody seriously expects the maps to stay this bad for very long at all – while the upsides are permanent. Users do benefit from having their privacy protected, and from having a Map app which uses the whole screen to display maps, rather than reserving one corner for a Google logo. And having control of the app back in Apple's hands presumably means that users won't see another six years with no new features, either.

They needed to switch provider now

But if Apple did need to switch from Google's data, why do it now, when their own data is so clearly incomplete? After all, Apple had over a year left on their contract – they could have spent at least that much time improving their service.

The question takes on further import when we find out that the reason why Google hasn't got its own replacement maps app ready to go is that they the timing of the announcement apparently took them by surprise – the New York Times reports that they were expecting Apple to wait until their contract ran out before replacing them.

They certainly knew, long before it was confirmed in June this year, that Apple was developing their own maps. The company started making acquisitions in July 2009, and has made more since then.

Owing to the way Apple ships software updates, the map switchover could only have come now.

The deal, according to John Gruber, expires in the first half of 2013 – crucially, long before iOS 7 is expected. And as he writes:

An all-new maps back-end is the sort of feature that Apple would only want to ship in a major new OS release. Technically, they could roll such a thing out in a 6.1 or 6.2 update, but major changes — and I think everybody can agree this has been a major change, for users and app developers alike — should be delivered only in major new OS updates.

If Apple wanted to replace Google – which they did, and which Google knew they did – they had to do it now, run the risk of having to rush a major release, or confuse users even further by putting a major software change into a "minor" release.

They needed to release first, improve second

And the thing is, holding the update back wouldn't have actually helped. The problem with the iOS 6 Maps isn't a lack of polish. On the contrary, the experience is actually already more "Apple-like" than the old Google-based maps were, thanks to significantly better-looking map tiles, a more minimalist UI, and the very well-designed turn-by-turn directions.

The problem is a lack of data. And that just isn't something you can get with a few hundred testers based in Cupertino. Yeah, you can tell looking at the maps that your favourite coffee shop is on the wrong side of the street – but until you tell them, Apple has no easy way of knowing that they've got it wrong.

The Atlantic highlighted earlier this month just how hard it is to build good maps, and it's a process of eternal refinement. You don't just release a perfect map. You iterate, iterate, iterate, and hopefully you eventually get a map which is correct before the world itself changes to make it obsolete again.

iOS 6 Maps. Photograph: http://theamazingios6maps.tumblr.com

Alex Hern is a technology reporter for the Guardian. He was formerly staff writer at the New Statesman. You should follow Alex on Twitter.

MAHMOUD RASLAN
Show Hide image

“I was frozen to the spot”: the psychological dangers of autoplay video

Videos that play automatically are now ubiquitous across social media, but the format leaves many viewers vulnerable to harm and distress.

Have you ever seen a dog being skinned alive? Witnessed a child, whimpering for his mother, getting beheaded? Have you watched a man, pinned down by two police offers, being shot multiple times in the chest and back?

A few years ago, if you answered “yes” to these questions, you might have been considered deranged. Possibly, you would have been on a list somewhere, being monitored for seeking out disturbing and illicit videos online. Now, you’re more than likely just a victim of social media’s ubiquitous autoplay function.

No one likes autoplay. Back in the Nineties, homepages often came with their own jaunty background tune that would automatically play, but it didn’t take long for this annoying and invasive practice to die out. Nowadays, when you click on a website plastered with noisy adverts and clips, you immediately click off it. But although users frequently bemoan them, autoplay videos remain a huge business model for social media sites such as Twitter, Facebook, and Tumblr.

That’s fine, of course, when the autoplaying video in question is a bird’s-eye view tutorial on how to make nacho-chicken-pizza-fries (though even then, the videos might be gobbling up your data allowance without your consent). The problem arises when disturbing content is posted by users, and even media outlets, without any warnings or disclaimers.

“There are many incidents where the autoplay feature has affected me negatively,” says Sarah Turi, a 19-year-old college student from Boston, USA. Turi suffers from anxiety, and says that anything related to horror or gore can keep her awake for multiple nights on end. She has previously experienced anxiety attacks after viewing autoplaying horror movie trailers.

“Recently though, many of the videos that begin playing have to do with police brutality or terrorism or riots,” she says. “There was one incident where someone had shared a video of an execution. The video started playing, and before I could scroll away, I watched a man get beheaded by a terrorist organisation. It left me pretty shaken to say the least. I wasn't crying, but I was frozen to the spot. Even just thinking about it now leaves me feeling somewhat ill.”

Dr Dawn Branley, a health and social psychologist specialising in the risks and benefits of internet and technology use, tells me that autoplay videos on social media raise a variety of ethical concerns.

“Social media is more personal in nature compared to news channels and it is also often idly browsed with little conscious effort or concentration, and, as such, users may not be mentally prepared for the sudden appearance of a distressing video,” she says. “Suddenly witnessing a beheading, rape or graphic animal cruelty whilst scrolling through photos of your friends and family, holiday snaps, baby videos and wedding announcements may provide a real shock to the viewer.”

Branley says that, in her line of work, she has spoken to users who have experienced distress at photos of abuse and violence on social media, and speculates that autoplay video could only exacerbate this problem. She also notes that they can trigger vulnerable users, for example, people who suffer from eating disorders or PTSD.

Even those without pre-existing conditions can be negatively affected, however, as anyone who has seen disturbing footage before knows how it can pop into your head intrusively at any time and refuse to budge, remaining plastered to the edges of your skull. Even trolls are aware of this, as some tweet distressing footage at people, aware that it will autoplay.

In January 2015, Facebook responded to these issues by adding warnings to videos users flagged as graphic, meaning the footage wouldn’t autoplay and was preceded by a warning message. Viewers under 18 would be shielded from seeing violent content on their feeds. Yet just over seven months later, in August, autoplay meant thousands inadvertently watched footage of the shooting of TV journalists Alison Parker and Adam Ward.

Remember when I said no one likes autoplay? That’s not strictly true. You have three seconds to scroll away from an autoplaying video before Facebook counts it as a view. In a world where Facebook, and the users of it, are desperate to tally up as many views as possible, autoplay is considered a smart business model.

“Autoplay video originated as a way to capture viewers’ attention and prevent them from ignoring or scrolling past website content,” says Branley. “The autoplaying nature of a video is likely to capture the viewers’ attention and may potentially be harder to resist watching – compared to a static image and text.”

For those profiting, it seems not to matter that some people who can’t look away are viewers like Turi, frozen on the spot by distress.

Because of how profitable autoplay is, then, many news outlets continue to upload sensitive footage that might better be suited on their website – a consensual click away. They might add their own pre-roll warnings, but Branley notes that these are easy to miss if the video is autoplaying. If you were online yesterday, you might have seen this in action, as footage of a boy – or rather the boy – in an ambulance, distressed and bloodied, autoplayed repeatedly across social media.

News outlets have been called out for this before, and have admitted their mistakes. In August 2015, New York Times social media editor Cynthia Collins told The Media Briefing that she wishes she’d added a warning to a video of men being shot and killed at sea. After backlash from their audience, she said:

“If we could do it all again . . . there would have been a discussion about whether or not we should upload the video at all. But if we decided to upload the video I absolutely would have added a warning.”

The video ended up on the website, and viewers had to click through a handful of warnings before watching it. News footage has always had the potential to alarm and distress, but at least in the past viewers had a choice about whether they watched it. Although many news outlets have guidelines on graphic content (such as, for example, the famous breakfast test), these haven’t always been updated for social media.

It’s important that users are made aware of potential solutions to this problem,” says Branley, noting that Facebook and Twitter include options in their settings to turn off autoplay, and your browser or phone may also allow you to disable all autoplay. “However, that does not detract from the moral obligation that internet platforms should consider when introducing autoplay.

“I would suggest that an ‘opt-in’ approach (where users are required to switch on the autoplay setting if they wish to enable this function) would be much more appropriate than the current ‘opt-out’ approach, which requires users to find the settings to switch off autoplay if they do not wish to use it.”  

This seems like the simplest and fairest answer. It’s hard to argue that distressing videos shouldn’t be posted on Facebook – last month, the footage of Philando Castile’s shooting dramatically shed light on police brutality – but it seems only natural that viewers should have a choice about what they watch.

“It is possible that autoplay videos could be used to raise awareness of sensitive topics and/or to grab users' attention for positive reasons like charity campaigns,” says Branley. “However, it is a fine line between raising awareness and distressing the viewer and what one viewer finds acceptable, another may find highly distressing. Therefore, care and consideration is required.”

Right now, both care and consideration are lacking. In its current iteration, autoplay is like Anthony Burgess’ Ludovico technique – pinning our eyes open and forcing us to watch violence and horror without our consent. There are things I know I never want to watch – the curb stomp in American History X, an Armenian weightlifter dislocating their elbow during the Olympics – that could be sprung upon me at any time. Why? Because someone, somewhere, profits.

“I don't think autoplay is necessary in Facebook,” says Turi. “I think that people should decide for themselves whether or not they want to watch something. And yes, I think that it should be banned.”

Amelia Tait is a technology and digital culture writer at the New Statesman.