Clickbait, the Attention Economy, and the End of Journalism

The article Google and Facebook don’t want you to read!!!

When you think of the giants of Silicon Valley, you might typically think of the names Google and Facebook. Though they market themselves as a search engine and a social network, respectively, they are actually advertising engines. And they are slowly eroding away our ability to reason.

Google uses search to help draw your attention to things — things that you might search for on their website. Facebook lets you stay connected to people you know and things you like — so that they might better serve you targeted advertisements. This concentrated gathering of peoples attention have directly given page views, clicks, and Facebook likes a monetary value.


But of course, this isn’t a new thing. This is how most publications have worked throughout history. Newspapers, billboards, television have all been profitable and influential using this model of using views to exchange for money. What’s new is that the democratization of the internet allows everyone to be included in the conversation.

Without arguing cost/benefits of democratizing the internet, the reality is that many people have become famous and profitable using social media. Views, clicks, and likes now have monetary value to significantly more individuals. In addition, having lots of likes or retweets demonstrates social value (for example, a photo of a girl in a bikini with 10 likes likely has less social value than one with 10 million) and credibility for an individual. The type of credibility or status once only available to celebrities and publishers.

Today, individuals, news organizations, or anyone with either the time or money to do so are all competing for the same amount of attention. All that matters is page ranking and site metrics. It’s a free-for-all for human attention — a limited resource with plenty of value now attached to it.

Internet multi-millionaire, pewdiepie

Five thousand status updates are posted on Facebook every second. That is an incredible amount of information to process for any individual. And so, we all must constantly optimize to become more and more relevant — because relevance equals attention, equals hits, equals money.

As individuals, to raise our social (and/or economic) value, we are now all posting things that give others acute emotional responses, because it draws attention, and attention is a shortcut to value. And as capitalists, we adhere to the belief that creating value makes one even more valuable.

However, over time, these same things give us less of an emotional response due to repeated exposure — at which point, we must up the ante by posting things that give us increasingly larger and larger emotional responses. We are at a point where we freely and easily post videos of people getting shot to death to show how upset we are.

Ladies and gentlemen, our society retweets snuff films under some weird guise of justice.

Content producers must respond by creating even more emotionally evocative material and more of it. This constant cycle of being more emotionally evocative and creating more content to fight for peoples’ attention leaves people with less time to consider subjects rationally, to act largely on impulse and emotions.

The more upset you are, the more you click. The more you click, the more these content creators make. It’s a race to the bottom.

Which would all be fine if this emotionality were limited to entertainment content. But unfortunately, contemporary journalism must bank on outrage clicks to pay their bills.

Here is the one NY Times article that uses A/B testing to measure two different headings on the exact same story. Which one would you click?

Source. H/T: Trey Causey

What are they optimizing for? Emotionality — your emotional response (to click or not click) to a stimulus (a lede or headline). And through years of optimization, testing, and writing, we’re discovering that the more emotional response a person has to something, the more likely they are to click.

Here’s the headline they settled on (retrieved October 10, 2016) as being the most profitable. Note they kept the words failing mission, while removing the less evocative words like growing sullen and barge ahead.

Truth be told, that’s actually tame compared to what’s actually happening out there. For example, check out these two headlines written a week apart by the same author:

But moreover, look at the number of Facebook shares. The one calling him a fascist got ~12x more shares than the one calling him a moderate Republican. So, let me ask of you — what type of articles will Jamelle Bouie be writing from now on?

Here’s one from last week, which gained ~100x more shares than the “Moderate Republican” article.

This isn’t just irresponsible, it’s dangerous. By pushing an agenda driven by emotionality disguised as journalism, the author is inciting overthrowing of government by calling it fascist while simultaneously taking absolutely no personal responsibility in the matter, but instead doubling down by saying half of the nation is racist and doesn’t deserve empathy. Regardless of your political leanings, there is simply no way to excuse this type of behavior. And yet, we click and click and click.

No one talks about journalistic integrity anymore. The idea that journalists should report on the facts, not their distortion of the facts is long gone. Data journalism was supposed to change this. Data journalism was supposed to give the people the numbers and let the people decide for themselves. Data journalism was supposed to limit personal bias, stick to the numbers, and give us insights. But that isn’t happening either.

Nate Silver of FiveThirtyEight might be the biggest loser of this election cycle. He correctly predicted 49 out of 50 states in the 2008 presidential election, and 50 out of 50 in 2012 through “algorithms” and “data.” We read his book The Signal and the Noise because we wanted to be like him and think like him and see the world like him. We thought he was invincible.

Silver is credited as being a statistician and a journalist — both disciplines that should require non-partisan beliefs, integrity, and a reliance on the facts. But he too, fell victim to his agenda-pushing, personal bias, and personal taunting, and he did it very publicly:

And let’s not forget (from 538, though not Silver himself)

Nate Silver wrote the article How I Acted Like A Pundit And Screwed Up On Donald Trump back in in May—a false mea culpa, where he admits his model was wrong, he used bad data, and admitted in over-correcting the data for his outcomes, but again screwed up royally in the general election. So bad in fact, that FiveThirtyEight still held on to giving Clinton a more than 50% chance of winning, despite every other major news outlet all but announcing Trump as the winner.

FiveThirtyEight’s forecast
Meanwhile, around 10pm, the New York Times had Trump >90%

And while I do take a slight bit of pleasure from turning Nate Silver into a punching bag, that’s not the point. Nate Silver admitted to fudging the data and has shown a very public personal bias towards one candidate. It’s not hard to imagine that one has had an effect on the other. And it’s also not hard to imagine that trying to increase his readership also has had an effect on his personal views. These personal views now reach millions of people whose personal views also become skewed. And if only those who have the largest reach and make the most money rise to the top, then we enforce an echo chamber of the highest order, encompassing the entire nation with more stories that predominantly exist to make more money.

We thought data-driven journalism would stem much of this bias. We were wrong. We thought opening up data to the public would help lots of people reach the truth. But the time and skill required to reach firm conclusions from this data are out of reach for the vast majority of people. We often forget that journalism requires smart, dedicated, non-partisan journalists, who don’t necessarily write for clicks, or aren’t afraid to speak out because their jobs depend on it.

Though we can’t all become journalists, or take down large media conglomerates single-handedly, what we can do is this:

  • Research claims made, particularly if they are very outlandish.
  • Visit more than 2 or 3 websites. Particularly if they seem to tell the same narrative.
  • Consider the opposite, the truth is almost always somewhere in the middle of two extremes. And always update your priors.
  • Even more dangerous: Stop relying on google or Facebook’s algorithms.
  • Hold journalists accountable for the things they say and write. Vote with your readership and viewership. This includes “fact-checkers”
  • Understand the world is not out to get you, it only appears that way because the business model requires your emotionality.
  • And stop clicking shit like this:

Note: There’s a lot of buzz around how Facebook is now actively trying to develop ways to get rid of “fake news.” Please, please, please do not trust them to provide to you what is “quality” and what isn’t.

I write about logic and reason, data and people.

I write about logic and reason, data and people.