Wordyard

This month I have watched with fascination the unfolding of what future historians may dub the Great Reverse Engineering of the Facebook algorithm.

It is a noble and important effort! I applaud those who have labored valiantly in its trenches. And I also have to say: It's doomed, hopeless, a dead-end street.

Here's what's been going on: People are playing public games with Facebook in an effort to get a finer-grained understanding of how, exactly, the newsfeed algorithm decides to hoist or bury individual postings.

To wit: Caleb Garling wrote in the Atlantic about his effort to trick Facebook into displaying a post more widely by sprinkling it with ringer language. "Hey everyone, big news!!," he wrote. "I've accepted a position trying to make Facebook believe this is an important post about my life!" Sure enough, the post went into heavy rotation.

Garling's post was content-free — an elaborate self-referential stunt. Next, media scholar Jay Rosen started applying the same technique as a tactic for boosting the visibility of his posts on the future of journalism: "Big news in my family! (You can ignore that. Just messin' with the FB algorithm so maybe you will see this.) I have a new post up at my blog, PressThink." It seemed to work pretty well.

Meanwhile, Wired's Mat Honan conducted an experiment to see what would happen if he "liked" everything he encountered on Facebook for 48 hours. He fell into a rabbit-hole of alienation, and discovered that there is no end to the like — it is an infinite loop.

Elan "Schmutzie" Morgan took precisely the opposite tack, forswearing all use of the "like" button. Morgan found that removing the "like" option from her palette forced her to connect more substantively by writing actual comments on posts if she wanted people to know what she thought:

It seems that the Like function had me trapped in a universe where the environment was dictated by a knee-jerk ad-bot… Now that I am commenting more on Facebook and not clicking Like on anything at all, my feed has relaxed and become more conversational. It’s like all the shouty attention-getters were ushered out of the room as soon as I stopped incidentally asking for those kinds of updates by using the Like function.

One inspiration for all these experiments is the long-term success that outfits like Buzzfeed and Upworthy have found in plumbing the mechanics of virality. Another, I'm guessing, was an event at Harvard's Berkman Center last month, at which scholars talked about the results of a "collaborative audit" of the Facebook newsfeed algorithm.

Their findings were fascinating, but the single most important result was also the simplest: a majority of the people in the study didn't have a clue that Facebook filtered their newsfeeds at all. That should give the rest of us pause: While we struggle to fathom the nuances of Facebook's post filters, it seems likely that a vast number of users don't even understand that their feeds are shaped to begin with. So there’s one beneficial side-effect of these various experiments in fooling Facebook’s machine: they help make people aware of the machine’s presence.

But the biggest problem with the reverse-engineering project is that we are not studying some natural phenomenon or physical product. The newsfeed algorithm is malleable software that's mutating all the time. The harder we game it, the faster its operators will change it.

An algorithm’s flexibility is one of its great strengths. Facebook’s changes for all sorts of reasons — including the well-intentioned efforts of Facebook developers to improve users' experiences, the competitive demands of the social-media marketplace, and the specific needs of Facebook-the-corporation to satisfy shareholders.

But the algorithm also moves in specific reaction to just the sorts of reverse-engineering projects I've compiled here. Any edge you can build by faking Facebook out isn't going to last long. A decade's worth of SEO-expert experience with Google bears this out. It's a game of Whac-a-mole that's rigged in favor of the platform owners, who have a direct line into the code that the rest of us are just speculating about.

Rosen, like many other journalists, expresses a preference for Twitter's structure, which by default shows you all the updates of every user you've chosen to follow. That makes it more transparent and gives users more direct control over their informational diet. There’s less guesswork involved, and that gives it far more value for sharing news.

But you can't count on it to stay that way. Twitter looks likely to evolve in Facebook's direction — it has shareholders to satisfy, too.

In any case, the longer we play this cat-and-mouse game with the social network operators, the clearer we can see that we are the mice. The experiments we perform from the insides of our Facebook compartments will grow increasingly desperate — like the ruckuses prisoners create as they try to capture or divert the attention of their jailer. But they'll never give us answers we can rely on.

All the reverse engineering in the world won't solve our deeper problem as users and builders of digital networks. The more we depend on such networks for our information, our social connections, our government and our entertainment, the more vital it becomes for their workings to be transparent, fair, and organized for the public good — and the less willing we should be to subject ourselves to the vagaries and whims of fickle companies. (This week, for instance, Twitter — which has touted its free-speech credentials — decided to censor controversial images of a reporter’s beheading. The images are loathsome; but do we really want Twitter-the-company to make these calls for us?)

Services like Facebook and Google don’t share the details of their algorithms because that code is their “secret sauce.” We’ve all heard that cliche. Secret sauce can be tasty. But if it's a big part of your diet, sooner or later, you're going to need to want to get a full breakdown of the ingredients.

Once upon a time, "retweeting" was something you did by copying the full text of someone else's Tweet and adding "RT @username:" to it. If there was space, you could add your own comment in front.

Then Twitter incorporated retweeting into its software — you just pressed a button to share someone else's message with everyone who follows you. Retweets became a lot easier; and overnight, the old-school technique became known as a "manual" retweet. For a while, old-school users (like me) stuck with it, either out of habit, or as a badge of ur-Twitter cool, or because we actually relished the opportunity to add our two cents. Meanwhile, more modern social-media mavens began complaining that it was a bad practice — it messed up their analytics. Some indignantly told us that manual retweets were actually evil self-promotion, because they go out under your own name rather than that of the author of the original tweet.

Today, those of us who still use it occasionally are like drivers who still like a shift and a clutch. Don't we know the technology has moved on?

Now Twitter is tinkering again! The latest change is that the "favorite" — a tool most users rely on either to bookmark links they want to return to or to send a little head-nod of acknowledgment out to the tweet's creator — is being put to use in a new way. Twitter is experimenting with showing you tweets from users you do not follow if those tweets are favorited by lots of other users (presumably, a lot of other users who you follow).

Like most of Twitter's steady evolution since its debut in 2006, this change pushes the service away from early-adopter enthusiasts and toward a bigger crowd. Twitter already has hundreds of millions of accounts and probably tens of millions of actual real people using it. But that's not enough to support Wall Street's expectations; post-IPO Twitter has billions in its eyes.

Quartz's Dan Frommer thinks that's all OK: "Twitter needs to keep growing," he writes, and "if additions like these…could make Twitter useful to billions of potential users, it will be worth rewriting Twitter’s basic rules."

Over at The Next Web, Jon Russell is less enthusiastic, calling the changes "confusing and seemingly unnecessary." When will Twitter no longer be Twitter?, asks Robinson Meyer in the Atlantic.

Russell argues that, even though you could always look up someone's list of Twitter faves, "favoriting is inherently a private action" — you were saving something for yourself or winking at some specific other person, not broadcasting.

Guess what? Twitter doesn't care. It's doing the same thing that Web platform-builders have been doing since the early days of Web 2.0, when Flickr and Delicious set "public" as the default for bookmarks and photo posts. They did so because there was so much interesting stuff you could surface that way.

Facebook moved that whole dynamic in a different direction by figuring out that if you trained people to share the details of their lives, and then kept changing the rules in ways that made those details more and more public, you could mine a network of billions of people for billions of dollars.

Back in 2007 Jason Kottke observed that what many Web companies do is "take something that everyone does with their friends and make it public and permanent." A decade ago, this technique seemed to be a method of expanding creative horizons, broadening the possibilities of online sharing, and enabling exciting new data mashups. Then we began to see its darker side, as financial incentives drove services to get aggressive about flipping the switch from "private" to "public."

Twitter may find that its "favorites" experiment works well. Maybe it "increases engagement" or improves the experience for casual users. It is also reminding us, as Facebook's mood experiment did, who is in charge, and what their motivations are. They control the vertical; they control the horizontal.

“I have come to believe that advertising is the original sin of the Web,” Ethan Zuckerman declared last week, in a remarkable essay posted at the Atlantic.

The fallen state of our Internet is a direct, if unintentional, consequence of choosing advertising as the default model to support online content and services. Through successive rounds of innovation and investor storytime, we’ve trained Internet users to expect that everything they say and do online will be aggregated into profiles (which they cannot review, challenge, or change) that shape both what ads and what content they see.

Regular readers here know that I largely agree with this view. It’s possible to imagine a different world in which advertising actually serves us as individuals: This is the work Doc Searls has been doing in defining and building VRM (“Vendor Relationship Management,” which turns the corporate world’s “Customer Relationship Management” on its head). it is also possible to point to ads that are respectful and effective (I admire the way The Deck works, for instance).

But mostly, advertising online today is a plague. I don’t believe advertising itself is inherently evil — so Zuckerman’s “original sin” formulation might ring the wrong bells. But it warps the good things about the Web. It stands between us as we encounter one another online. It creates incentives for publishers to make things less usable and less useful, and to blur distinctions that should be crystal clear (as with the new “native advertising”/advertorial fad). So far, it has kept us from fulfilling the potential of the medium we are building.

I argue that nearly every characteristic of the digital medium that critics abhor — its pileup of distractions, its invasions of privacy, its corrosions of trust — turns out not to be an inherent characteristic of the technology at all. These modern media maladies are byproducts of bad business models.

In a retort to Zuckerman, Jeff Jarvis says, “it’s not advertising itself that’s the problem, it’s that we’re doing advertising online wrong.” Maybe so. But this sounds a little like the insistence by political purists (old-school leftists, or present-day libertarians) that their particular system is The Answer, despite repeated failures to implement it in the real world, because “nobody’s done it right yet.”

I’ve been doing this Web thing for 20 years now, and in all that time, no one has solved the basic problems with online advertising that existed from the start. If someone does, I will cheer along with everyone else. But the calls to “give it time” because “we’re just at the beginning” have started to sound hollow. In the meantime, I will explore other options, and I encourage you too, as well.

If you keep noticing that every implementation of an idea sucks, it might be that you just haven’t found the right one. But it could be that the idea itself is simply wrong.

Jeff Pooley’s essay “The Consuming Self: From Flappers to Facebook” introduced me to the concept of “the authenticity bind.”

The pervasiveness of what might be called “calculated authenticity” leads…to rejectionist forms of authenticity — real authenticity, untainted by the professional smile and the glad hand. These flights to deeper kinds of authenticity are, however, marketed in turn — returned, that is, to the promotional fold.

I wrote about this dilemma earlier this summer. Since then I’ve engaged in a sporadic but fruitful email interview with Pooley, who is associate professor of media & communication at Muhlenberg College, in Allentown, Pennsylvania. He’s got a book in the works on the “calculated authenticity” theme, which is exciting. (And another book on its way before that, about the media theorist James W. Carey.)

I love Pooley’s analysis but recoil from some of its implications. I wanted to press him on some of the questions I had about where the authenticity bind leaves us, practically speaking. And I didn’t want to wait for his book.

Here’s our exchange.

ROSENBERG: I want to start by asking about how you arrived at the concept of the authenticity bind. Authenticity is clearly a complex term, and one whose intellectual-historical pedigree you trace in your essay. But the “bind” part was new to me. Where does it come from?

POOLEY: The main source for this idea comes from my sense that markets are fundamentally dynamic and adaptive. Marx was right about this at least: there’s a manic, restless energy built into capitalism. Schumpeter celebrated this dynamism as the “perennial gale of creative destruction”—which is celebrated in turn by peddlers of Silicon Valley-style disruption porn. (There’s the irony that, in American politics, defenders of tradition make common cause with prophets of creative destruction; it’s their tradition that the market melts into air.)

My view is that this carnal, rudderless energy doesn’t just empty out Ohioan steel towns, but also courses through the cultural industries. Popular culture in a consumer society is a lot of things, not just a reflection of market pressures. But the same frantic search for new markets and higher profits is gale-like here too. Media makers are gripped by the relentless pursuit of new resonances — new bundles of consumer attention. That attention is sold to advertisers, who are also busy scavenging the flotsam of resonant meaning floating about.

To get to your “bind” question: the theme of authenticity is an especially attractive hook to sell stuff. As I describe in the article, the authenticity ideal (though of recent origin) has a strong hold on us. We want to think of ourselves as singular, expressive beings. Hence the resonance for advertisers. But the tie-in with the ad pitch discredits the very language of self-fulfillment. So we go looking for a deeper kind of authentic living, resistant to marketing.

The problem is that there is no market-free space, at least not for long. The Dionysian core of a market-driven culture — with its gleeful agnosticism — makes the dog-collar-and-safety-pin aesthetic attractive to the ad guys. In whatever new version it appears. That’s the bind: there’s nothing prude or restrained about them, as Thomas Frank has brilliantly shown. There’s no way to out-rebel them; they’ll party with you at Burning Man. And you’ll realize that they’re using you to sell stuff. Yet again.

ROSENBERG: So I guess the first set of choices this presents us with is: you can just keep trying to dig a new level down toward the authentic — knowing there’s no destination, no core, at bottom, but hoping to stay one step ahead of the marketing machine at least some of the time. Or you can turn around and say, OK, if the “real me” is going to be used to sell stuff, I might as well get the profit myself — which leads you into the realm of what you call “calculated authenticity,” the conscious choice to aim for the Real Me because you know it’s going to pay off in some way.

I keep trying to think of other creative responses but coming up dry. Is that all there is?

POOLEY: This is a big question. I should probably back up and say that I do not view the market and its cunning adaptations as the main story. The union of — and tension between — self-promotion and self-fulfillment is not a product of Madison Avenue.

Leaning a lot on the philosopher Charles Taylor, I argue that “calculated authenticity” has earlier roots in a pair of moral ideals that have taken hold in the recent history of the West. Taylor, in his Sources of the Self, tells the backstory to the sense that we have individual selves in the first place. In telling this story, he traces the emergence over time of certain moral ideals that we, in the modern West, have inherited.

One is the sense that our self is something that we possess and control — an object that we can manage and work on. Another, more recent ideal is authenticity: the sense that we all have inner depths unique to each one of us, which we should explore and express. What I am saying, following Taylor, is that there was already a tension, long before the billboards and the NYSE: On the one hand we feel like the owners of our selves — that our lives are projects to work on — but also, on the other, that we ought to find and express our true selves. I realize this all sounds pretentious and way too sweeping, but I think Taylor is basically right.

So, back to “calculated authenticity”: the contradiction is already there, but is picked up by the self-help industry, pop psychology and advertising. The message is that the best way to work on yourself is to consciously cultivate an authentic persona — to use authenticity as a means to the end of self-promotion.

We encounter calculated authenticity all the time in our daily lives: friends talking abut their “brands,” labored-over insouciance in Facebook statuses, the professional smile at Cold Stone Creamery, etc. And it’s nauseating and blatantly inauthentic. So we go looking for the real thing — maybe it’s artisanal cheese or long beach walks, or even irony-bathed Oscars viewing — all of which (as I talked about above) is then processed, repackaged, and sold back to us. Rinse and repeat.

No, I do not think that a resigned embrace of calculated authenticity is the way out. That’s too depressing. I think it’s still worth striving for authenticity in personal terms — for not being a self-promotional ass, especially if that douchiness is disguised by sustained eye contact and the gladhand. I also believe that we should call bullshit, over and over again, whenever authenticity is used for shilling. Even when — especially when — the bullshit-calling itself becomes a slogan.

This isn’t because there’s anything pure or timeless about the authenticity ideal. The sense that we should be true to ourselves in this demanding way is a product of (fairly recent) history. Strictly speaking — and as academics never tire of pointing out — there’s no such thing as “real” authenticity, nor is the ideal written into the human soul. But just because authenticity has a specific history — is not natural nor timeless — does not mean that we should dismiss its claims on us. It’s part of who we are, and we can use its high standards to call out the debased form the ideal has taken on.

ROSENBERG: It sounds like this tension you’re describing is one between treating the self as an asset and seeing it as an unexplored territory of sorts: something that we have and control and deploy and revise at will, or something that we must go on a quest to find and reveal and release. That feels to me like a reasonably subtle differential — but it’s almost as if the force of the market sticks a big pry-bar into this little fissure and cracks it wide.

I mean, plainly there’s no path back to some sort of state of innocence here, no way to press reset and lose the knowledge that personal authenticity has market value. On my previous post about the authenticity bind, Mitch Skinner posted this comment:

“Recognizing that authenticity has value doesn’t require you to glad-hand or to calculate your smile. Being aware of it doesn’t make it fake. That awareness just makes it a choice. But that choice can still be fully authentic.”

Do you think there’s any way to hang on to one’s sense of being authentic while simultaneously understanding that this quality is going to serve a practical, self-interested end? What could that look like?

Calling bullshit on crass authenticity plays and the marketing cycle is always useful, for sure. Is there anything we can point to on the other side of the ledger?

POOLEY: Can you strive to be authentic knowing that it will gain you something? The question reminded me of what philosophers (drawing on Catholic thinking) call the “doctrine of double effect.” The idea is that it’s not wrong to do something harmful if it was a side effect of trying to do something good. The classic case is killing someone to save your life. The distinction is between intending and foreseeing: Yes, you realize that the knife thrust may kill your attacker, but your intent is to spare yourself.

I realize that the example is ridiculous — we’re talking about glad-handing, after all. But I still think the idea is relevant. You may know that, in general, acting in an authentic way is likely to benefit you. It’s a foreseen side effect — the fact that you may win friends and influence people. But your intention is to be authentic.

That’s one scenario, and in my view you’re fine. You’re a long way from soulless hypocrisy. But it’s different when you set out to use authenticity as a means to get something from someone. Take Dale Carnegie’s advice as an example: “Become genuinely interested in other people,” one of “Six Ways to Make People Like You.” Extracting admiration from others, for Carnegie, isn’t merely something you foresee. It’s the whole point.

The analogy to morality and the “double effect” is not a perfect fit. In the case of killing an attacker in self-defense, the (permissible) bad action is the killing. Homicide, in almost any other context, is wrong. With Carnegie’s advice, however, the problem isn’t the instrumental motive per se; we use people for our ends (like purchasing coffee) all the time. Buying coffee isn’t murder. The problem with the Carnegie tip is that it is self-refuting. When you set out to get something (like the admiration of others) by pretending to be authentic, you’ve already abandoned the ideal. And pretending like this, with self-serving ends in mind, has bad consequences for everyone: We all come to distrust sustained eye contact and attentive listening. Maybe, we think, they’ve been trained at the Dale Carnegie Institute.

So I would stress the distinction between foreseen side effects and intention. In good faith we can attempt to act in an authentic way, even though we think that we may profit somehow. But not if the profit is the point.

In his great little one-man show, Feisty Old Jew, my old friend Charlie Varon makes copious use of the word “vigorish” or “vig” — a Yiddish word so outlandishly pungent that half the audience regularly assumes he made it up. (He actually polls the house, impromptu, each night.)

Me, I knew it was a real word. I knew that it refers to the percentage of a bet a bookie takes for himself — or, more broadly, any commission on a transaction. I knew this because “vig” had cropped up in a prominent way in the annals of the ’90s Internet boom. When I first heard it then, I couldn’t quite shake the suspicion that the man who had used it, Nathan Myhrvold of Microsoft, must have made it up, so I looked it up.

At Microsoft in the ’90s, Myhrvold was known for writing voluminous in-house memos dedicated to crystal-balling the tech future. In a much-circulated 1993 memo titled “Roadkill on the Information Highway” (RTF file) he introduced the term “vigorish” to readers. But he really put it to work in a 1994 followup — a memo that tried to make up for the previous one’s inexplicable failure to account for, or even mention, the Internet, which was not the form Myhrvold or most everyone else at Microsoft had expected the “information highway” to take.

One of the reasons the Internet took Microsoft and so much of the tech industry by surprise was that its government-and-university roots and its open-computing culture made it seem like a singularly inhospitable place to do business. Microsoft was halfway through building its own Microsoft Network as a competitor to Compuserve, Prodigy, and AOL when the Web browser known first as Mosaic and then as Netscape started getting popular in 1994. Sure, the Internet was opening up to commerce thanks to recent regulatory changes; but why would you want to set up shop there?

As Ken Auletta wrote in a May 12, 1997 profile of Myhrvold for the New Yorker:

It also infuriated Myhrvold and Microsoft that the Internet was free. They saw it as a flower-child culture that disdained profits and copyrights — and Microsoft… “Nobody gets a vig on content on the Internet today,” [Myhrvold] wrote. “The question is whether this will remain true.”

We all know what happened after that: Microsoft gathered its forces and smote Netscape, preserving at least another decade of massive profits and sparking a federal antitrust suit. Meanwhile, the Internet became a pretty good place for at least some people to make a lot of money.

I revisit this tale of yore partly because the word “vigorish” is just such a treasure (as is Charlie’s show — go see it if you get the chance!), and partly for two more substantive reasons: It puts today’s news of Microsoft layoffs in perspective, and it reminds us of how heavily today’s Web business depends on “vigs.”

Sic transit gloria Microsoft

Today Microsoft announced roughly 18,000 layoffs. That’s a lot of jobs to vanish, even if more than half of them are from the company’s recent acquisition of Nokia’s phone business.

Microsoft isn’t going away; heck, IBM is still around and making plenty of money. But Microsoft isn’t exactly leading the industry the way it did when Myhrvold was writing those memos (these days, he’s busy buying up patents and publishing $600 cookbooks). It has been displaced as thoroughly as it displaced its mainframe and mini-computer predecessors.

Myhrvold’s dream of Microsoft interposing itself as the Internet’s middleman, taking a vig from every transaction, never came true, thank goodness. Microsoft already took another kind of vig from us in the form of Windows and Office licenses and upgrades, but those are less and less central to our work lives, and almost irrelevant to our casual/personal digital lives.

It was hard to imagine such an outcome in the late ’90s. It is similarly hard today to imagine a relatively near-term scenario in which any of Google, Facebook, or Apple have faded into near-irrelevance in shaping the future. Rest assured: it will happen.

Today, you can’t click without tripping on a vig

The Web remains relatively vig-free today: if you set up shop there, you need to pay for hosting, but there aren’t a lot of people squeezing a percentage from you. (There are credit-card transaction fees, but they exist offline as well.)

But as we move into the world of mobile and apps, in which private vendors maintain tighter holds on app and content distribution, we’re suddenly back in the land of the vigorish. Apple’s app store takes a big fat cut, as do most other app stores. Ditto for content marketplaces like the iTunes store and the Kindle store. Apple makes sure to get a cut of in-app purchases, too. As Mike Cane wrote in a 2011 post, the day Apple’s app store started insisting on its cut of in-app purchases was “the day Apple became Nathan Myhrvold.”

These are obvious vigs — but there are other kinds. Facebook and Twitter both take a slice of us, too, not from our transactions but from our social lives and our attention. Think of this as an emotional vigorish.

The “disintermediation” so many predicted back when Myhrvold dreamed of vigs was real enough. But today we are facing a new infestation of middlemen, in previously not-intermediated realms like finding a parking space or booking a restaurant reservation.

If you don’t like that — I don’t — it can be a frustrating moment in digital time. But hold tight: This sort of market advantage is usually fleeting.

It’s not that some competitor will come along and beat the app store at its own game. Instead, our needs and habits will change, and just as we find that we can get along OK without Microsoft Office, we will wake up one day and realize that we haven’t spent anything at an app store in ages.