Tool, or trouble? Facial recognition might be driving some sources away from the news

At first, Brittany Cantarella had no idea the man she accidentally swiped with her Chevrolet was named Lord Jesus Christ. But within two days, the minor traffic incident had gone viral. Reporters snatched the then 20-year-old’s Facebook profile picture and left messages on her grandmother’s answering machine. “It’s the girl that hit Jesus!” a man in Stop & Shop yelled.

“I wanted to hide, I wanted to run, I wanted to go far away,” Cantarella said.

Two months later, she was willing to talk to me about the accident at a coffee shop in western Massachusetts. She was resolute, though, that I not take her picture or shoot video. That’s because Cantarella’s experience with viral fame made her wary of having her image wedded to a traffic accident that would never go away online.

This small anecdote is part of a new media conundrum dogging the relationship between visual journalists and their subjects: most people happily publish their own picture online, but a growing number of them are becoming wary of having their image captured by visual journalists.

With facial recognition software becoming commercially available in the past few years, new technologies could further reshuffle the relationship between a subject and a visual journalist.

Ed Kashi is a renowned photojournalist who has spent the past 30 years shooting for National Geographic, the VII Photo Agency and dozens of other outlets. And, he told me in an email interview, he’s noticed individuals and organizations becoming more reluctant to allow visual access.

“There is more wariness and a desire to have more control over access and what you are allowed to show,” he said. “In some cases and with certain subjects, this new paradigm presents a dilemma and can halt worthy work.”

On balance, Kashi sees the change as positive. “Photojournalists are more accountable,” he said, since the people in the pictures can watchdog for accuracy whether they’re in New York City, Nigeria or the West Bank.

Pulitzer-Prize winning photojournalist Martha Rial, whose career spans over 20 years, agrees that photojournalists have a higher hurdle to get started on projects. “People are aware of how the 24-hour-news cycle has changed the perception of everything,” she told me.

“There’s no denying that it’s getting harder to convince people to allow photographers into their lives in a meaningful, substantive way,” said Jason Cohn, a Pittsburgh-based photographer and videographer, “and there’s no denying it’s for good reason on their part.”

Aside from his work as a photojournalist for outlets like Reuters, Cohn has been a member of his hometown city council since 2005. As a public official, he has become “really wary about photos taken of me, because you never know when a photo will be twisted or turned to be used against you out of context years down the road.”

With a few exceptions like spot news, visual journalists depend on their subjects’ consent. For a subject, that often means ceding control of your own image to a stranger.

Patience, respect and tenacity are the traits that photojournalists are taught to convince a waffling subject to appear in a story. Superlative photojournalists are renown precisely because they can find subjects who allow them to tell visual stories, regardless of obstacles.

How, then, does this wariness affect visual journalism? The problem arises when patience and time are not options visual journalists. That may be because they’re overworked daily journalists who don’t have time to talk their way into a storytelling picture. Or it might be because they are citizen journalists or students without the experience to explain the importance of their assignment. All might be too willing to take the first “no” as the final answer. That stops worthy visual coverage.

New technologies barreling into consumer products have the potential to further sandpaper the relationship between visual journalists and their subjects. That technology is facial recognition technology.

Over the past few years, versions of the technology have moved from law enforcement and big businesses to consumer uses, notably in Facebook and Google Picasa, according to Alessandro Acquisti, an associate professor of information systems and public policy at the Heinz College at Carnegie Mellon University. He made headlines last summer with research showing he could find the name of 31 percent of CMU students who stopped at his research table simply by using facial recognition software and a database of public images from Facebook.

“You will never find me,” one student boasted before the experiment, Acquisti told me. The research team quickly found him thanks to a picture a friend posted.

“The subjects we identified, they were quite surprised,” said Acquisti.

Right now the kind of facial-recognition software Acquisti used for his research needs a frontal, well-lit shot to return a match.

“I’m pretty confident that [the technology] will get better and better over time. Whether it will ever meet or surpass human ability, it’s a difficult question,” says Acquisti.

If the consumer technology does become more powerful, it could have a significant effect on the subjects of news pictures. If a brutal actor like the Syrian government could find the identities of every protester, would it be ethical to take or publish a picture from a demonstration?

Sites like The Chive already feature galleries where they take pictures, sometimes from news or sports events, of women. Would those women want to be casually identified by anyone online? Those and other scenarios create an undeniable logic to not appearing in any news pictures.

Acquisti’s scientific research is more rigorous than the anecdotal wariness some visual journalists see. But “the behavioral economics of privacy,”
as Acquisti calls his research focus, portends a future where the subjects of visual journalism have new incentives to appear or not appear in the media.

“The joke is don’t put anything online that you would not like to have on the front page of the NY Times 10 years from now, because chances are that if you become an important person or you are about to be considered for an important position, that information will resurface.”

Should that message sink in, visual journalists may find themselves trying to fit in to a different equation.

An ad buyer's SEO advice for online news publishers

The best S.E.O. strategy,” I told a group of UMass journalism and marketing students, fidgety at the smell of the free pizza cooling in the next room, “is to produce strong, original work.”

I’m paraphrasing, because that line wasn’t in my notes. But I liked the sound of it. It’s a comforting thought for a journalism professor that good journalism will find a large audience. Yes, it’s vital to understand the basic principles of writing pithy <title> tags, wise linking practices, and analyzing audience behavior. Those are parts of the “Plain English Optimization” the Online Journalism Review rightly promotes, and I teach those skills in several of my classes.

But isn’t strong, original journalism paramount? Or was I being naive during my S.E.O. talk, wiping pizza sauce off my chin in my ivory tower as working journalists were getting crushed by the weight of Charlie Sheen keyword trends? I decided to ask a person who would be clear-eyed about how journalists can best approach S.E.O.: an ad buyer.

Chris Lorenzoni has been planning, designing, and buying ad campaigns for various New York City ad agencies since 1998. He has bought ad space in the New York Times, Yahoo News, the Wall Street Journal, and the Washington Post, and he’s spent his career on technology’s leading edge, most recently working with mobile and tablet advertising.

“Mass marketing as we grew up learning it is pretty dead,” Lorenzoni said. “Even if it’s for the same car, you can’t run the same ad for a 35-year-old guy and a 21-year-old girl. You have to carve out and look for different audiences.”

The need to identify those audiences analytically was a profound cultural change for the ad industry.

“Nowadays when you look at people that do the initial research about sites, it looks like they’re stock traders. People are logging in and matching audiences against another third-party verification. There’s a lot more science and math in media buying.”

What that science and math means for online journalism outlets is that ad buyers can chart detailed demographics of site users over time, and they use that targeting extensively. A WSJ series demonstrated how pervasive and sophisticated analytical tracking can be.

From an advertising point of view, then, it often makes more sense to target several smaller audiences with interests that correspond to the product than one mass audience with more diverse tastes.

It also means that if your content stinks and and your S.E.O. strategy attracts viewers who quickly bounce off the page with a bored shrug, that’s an obvious fact to advertisers.

The truth for most journalism outlets is that understanding your niche – whether it’s geographic, demographic, tonal, or otherwise – and building up a large and enthusiastic community within its limits is the real value of knowing S.E.O. Most of us are not the New York Times, with a mission to serve a mass audience. When it comes to the content itself, the blogger Matt Yglesias suggests that “people in the ‘writing about important things’ business need to roll up our sleeves and try harder to make our output compelling to people.”

A mass audience certainly won’t hurt, but it may not be the smartest place to put your energy in a world of finite resources.

Online advertising, which is often run though middleman networks, also ups the stakes on an old problem: the very events that drive massive news consumption are often not events advertisers want to be associated with.

“That big bump of traffic can be toxic. We don’t want to be adjacent to war or dead bodies,” Lorenzoni said. “Honestly, the most valuable stuff they [media outlets] end up doing are more of their special sections, the things they can promote beforehand.”

Lorenzoni’s advice to journalism outlets is to spend their technical energies developing “platform agnostic” online presences – meaning that your content is “readable, accessible, and trackable from multiple sources” like iPads and mobile phones. Right now, for example, advertisers are paying a premium on iPad content, he said.

Social networking is also a vital actor in building the kind of engaged audiences advertisers covet, said Lorenzoni. Having your content appear high in the SERP is certainly worth pursuing, he said, but it’s a shrinking part of building a successful audience.

Which brings us back to the original question: is good journalism a good S.E.O. strategy?

“When you go back and you look at what kind of stuff look at and pass around and share, it becomes the well-produced videos, the well-written articles,” said Lorenzoni. “That’s the kind of stuff that brands really want to get behind.

What the 'Ground Zero mosque' flap says about the state of journalism

The Ground Zero mosque does not exist.

There is, of course, the planned Park51 Muslim community center and mosque, which local authorities approved for construction on Park Place in lower Manhattan about two blocks, or about 600 feet, from Ground Zero.

And there is also, of course, a myth – the latest outrage brand- of a “Ground Zero mosque.” Headlines from dozens of outlets have trumpeted that three-word shorthand, tempered at best by the flimsy embrace of quotation marks. Yet the phrase “Ground Zero mosque” violates the most basic tenets of journalism: be truthful and be accurate.

So what’s false? Simple: the mosque in question will not be built at Ground Zero. To conflate the lingering psychological toll of the destroyed World Trade Center with a building 600 feet away is as absurd as calling the Lace Gentlemen’s Club on 7th Avenue in Manhattan the “Fox News Strip Club” by virtue of its two-block proximity to Fox’s headquarters.

Speaking to Michael Calderone at Yahoo News, AP New York assistant chief Chad Roedemeier said that the slug on the story has always been “Ground Zero mosque,” and that phrase has often appeared in headlines. But he said the wire service has always said the mosque was “near” ground zero in stories. (I used to work as a freelance photographer with the AP in New York City.)

That distinction isn’t good enough in an age of six-word iPhone headlines, warp speed online skimming, and well-financed PR and political hucksters trying to smoke-bomb plain languge. Whether it’s birthers, Breitbart, or BP, there will always be cynical and reductive operators trying to exploit the uninformed in the age of too much information. The question is why responsible news media doesn’t fight as aggressively to reframe stories with the facts.

Our brains, like search engines, gauge information on a hierarchy, prioritizing headlines and the active nouns and verbs they employ. Copy further down in the story or watery qualifiers like “near” or “so-called” don’t stick in our brains as much, nor do they help a website climb the SEO ladder.

Yes, we all skim. And no, it’s not the job of a journalist to stand over the consumer’s shoulder to make sure he or she reads to the end. Nonetheless, a basic standard of factual accuracy like “tell the truth about a location” should be self-evident, or else media outlets are no better then performance art.

But let’s go even further: responsible media outlets needs to stiffen their ethical spine and take a more active role in ignoring or correcting manipulative catchphrases. A story whipped up by a group called ‘Stop Islamization of America,’ like blocking the construction of Park51, should not have an even playing field with a more sober truth because of false or tricky language. Yet right now, the language manipulators are on a winning streak.

Last year’s “death panels” (there was no such thing) fit a headline far more easily than Atul Gawande’s nuanced and moving 12,000 words on end of life care. We’re now hearing about “terror babies” despite the fact that they don’t exist either. There are dozens of other examples.

Media ethics today fight a fierce riptide. Online media’s economic success depends on page views and click-through rates that in turn depend on the most blunt, emotionally engaging subjects. That’s a powerful incentive to create controversy.

“Not all words have equal meaning,” Frank Luntz, who coined dozens of phrases such as “death tax,” told the New York Times Magazine in May 2009. The PR executives, activists and politicians recognize that fact, and will continue to exploit it until news media outlets robustly resist. Let’s hope that day comes soon.