False economy What’s in it for me? Understand what makes the

particular political candidate less appealing 
when her picture is bordered in yellow than 
when it’s bordered in green.
This may not seem like a spectacular or sinister 
discovery, but let’s say that this politician’s 

---

campaign team gets its hands on that 
information. If they send you campaign ads 
featuring her green-bordered likeness, you will, 
statistically speaking, be more likely to vote for 
her.
And social media companies have no qualms 
about selling your information. After all, you’re 
not their client; you’re their product.
Their clients are advertisers – the companies 
that buy the data about you and then use it to 
convince you to buy certain products or vote for 
a certain candidate. In the author’s view, this 
amounts to direct manipulation of your behavior.
Sure, advertising has always been manipulative, 
but it’s only recently that ads could be tailored 
based on your personal preferences and online 
actions. Of course, this tailoring only has a 
statistical effect – that is, it’s not 100-percent 
accurate. You might, unlike most people whose 
diet is similar to yours, hate green, and therefore 
not vote for that green-bordered candidate.
However, over an entire population, statistical 
effects are reliable. Therefore, it’s more likely 
than not that you are being manipulated.
Social media platforms are designed to be 
addictive.
Imagine you’re a child, and whenever you say 
please, you’re instantly given a piece of candy. 
Unsurprisingly, this would prompt you to say 
please quite often.
Now imagine that, sometimes, saying please 
fails to result in the desired sweet. Given this 
occasional failure, do you think you’d start 
saying please more or less?
Though it may seem counterintuitive – after all, if 
an action fails to produce the desired result, why 
engage in it? – research suggests you would 
probably start saying please much, much more.
Behaviorists discovered this phenomenon 
decades ago, and it holds true for both animals 
and humans: moderately unreliable feedback is 
often more engaging than perfectly reliable 
feedback.
As we all know, social media want to keep us 
engaged, and they do this by taking advantage 
of this bit of behaviorist knowledge.
Facebook’s first president, Sean Parker, called it 
a “social-validation feedback loop.” Sometimes, 
someone will like your post or photo – but not 
always. And it’s this element of randomness that 
gets people addicted.
Furthermore, social media algorithms are usually 
designed to incorporate a bit of randomness, 
too. These algorithms are called adaptive 
algorithms, and they’re constantly adjusting 
themselves in order to be as engaging as 
possible.
How do these algorithms adapt? Well, let’s say 
an algorithm shows you an ad three seconds 
after you watch an adorable cat video. 
Sometimes this algorithm will conduct a little 
test. It might show you the ad two-and-a-half 
seconds after the video, for instance, to see 
whether that makes you likelier to buy the 
product in question.
If showing you the ad two-and-a-half seconds 
after the video doesn’t prompt a purchase, then 
it might try three-and-a-half seconds – but what 
if this doesn’t have an effect either?
In order to keep from getting stuck at three 
seconds, the algorithm sometimes makes a 
semi-random leap. It will try, say, waiting five 
seconds, or one. This randomness ensures that 
the algorithm never stops adapting.
And just like random social feedback, algorithmic 
unpredictability also contributes to social media’s 
addictiveness. Indeed, social media are so 
addictive that many parents in Silicon Valley 
send their children to Waldorf schools, where 
electronics usually aren’t allowed.
Addiction can cause a kind of craziness, one 
that can make you lose touch with the people 
and the world around you. And social media is 
turning us all into addicts.

---