Modern Life Skills: Reverse-Engineer the News
The idealized model of the way the media operate is this: a reporter makes a bunch of observations, slowly comes to a conclusion, and then writes up what they learned in order to let the reader draw their own conclusion. The cynic says this is 100% backwards: a reporter starts out with a conclusion, gathers evidence to support it, throws out anything that contradicts it, and makes their case — they don’t explicitly tell you what to think, they just do everything they can to lead you to think a certain way.
Traditionally, it would have been exhausting to double-check the media. A lot of people in 2003 were pretty convinced that the Iraqis had weapons of mass destruction and were prepared to use them imminently. What were the rest of us going to do? Go to Iraq? Get inside Saddam’s head? The closest you could get to that was trying to parse out everybody’s motives to see who wanted to lie — but anyone motivated to lie is equally motivated to tell you the truth, so while untrustworthiness plus motive is circumstantial evidence, it’s not proof.
We’re still stuck with that kind of sad agnosticism with most stories in the media, but there’s a particular subcategory that anyone can double-check: the Internet-outrage story.
The template goes like this: somebody got real mad and said something on Twitter or Facebook. You can actually write two stories about this: “Claps back at” and “Completely freaks out about” are two ways to describe the same reaction depending on which side the person doing the describing has taken. (Interestingly, Twitter’s trending topics section often tells visitors which side to take in this way.)
What’s amazing about these stories is that:
Anyone can investigate them — since the sources are all public, you can just read the quotes in context, and
If you do this, you’ll find that these stories are basically fake.
Here’s a great thread on a recent outrageous quote. Who was outraged? Some guy with… a total of 18 followers.
Ideally, if journalists are going to use Twitter to sample the vox pop, they should be choosing representative opinions from representative people. In this case, it’s representative in an entirely different way — it’s the kind of person the journalist expected to object, and the sort of objection he hoped to find.
In fact, you can go further: in both of the major Twitter-based articles (“People are saying X” or “One person is always saying X,”) you can skim the tweets and guess what search term the writer used to find that tweet. “Gillette” + “Boycott” is a good guess, here.
In this piece on the reanimation of Gawker.com, you can actually see how the article evolved:
First, the author searched for “@CarsonGriffith” and “Black”
The only remotely offensive tweet was about a black coat — maybe stolen by the maid! Let’s search “maid,”
Search the name of any protected class.
Search for “Trump”
Search for “fat”
Most of the tweets are just kind of blandly tacky; a few of the offensive ones are retweets or quotes. But if you’re skimming the article, it looks like nonstop meanness, in tweet after tweet — but these are a bunch of random, disconnected out-of-context tweets from the last decade.
Whether you’re skimming the dumbest 0.01% of one person’s Twitter brain-farts or the dumbest 0.01% of the entire Twitter userbase’s reaction to a headline, you’re guaranteed to find enough material for an article.
And that leads to a problem for writers. Economic efficiencies happen, but they don’t persist, and businesses try to fix any kind of situation where a job requires no skills but pays above minimum wage. It’s even worse, actually: unlike fast food and entry-level retail, there are people who assemble compendia of stupid tweets for free. Imagine thinking you could earn a middle-class income performing unskilled labor while competing with volunteers.
It’s no accident that there’s been a mass bloodletting in the segment of online media that specializes in this sort of low-effort, low-value content: as I argued a while back, the economics of digital media cause big cyclical swings in the number of people employed as writers.
If you work as a writer, or recently worked as one, it behooves you to stay aware of what parts of your job are value-added, and what parts are easy to automate. The only way you can do that is to learn, firsthand, about automation. Really, seriously, learn to code.