28 Scientific American, May 2019
VENTURES
THE BUSINESS OF INNOVATION
Wade Roush is the host and producer of Soonish, a podcast
about technology, culture, curiosity and the future. He
is a co-founder of the podcast collective Hub & Spoke and
a freelance reporter for print, online and radio outlets,
such as MIT Technology Review, Xconomy, WBUR and WHYY.
Illustration by Jay Bendt
Turning Off the
Emotion Pump
Are there better social technologies
than Face book?
By Wade Roush
Strip away the baby pictures, the cat GIFs and the high school
reunion invitations, and what is lurking underneath your Face
book news feed is one of history’s most effective targeted adver
tising platforms.
Face book watches to learn what pleases you and what angers
you, and it uses that information to auction ads to companies that
want to reach consumers with your specific profile. It also watch
es what everyone else likes, then shows you more of whatever is
most en gaging that day—the better to keep you scrolling, so that
you’ll encounter more ads. If the “whatever” happens to be an
Islam o phob ic graphic posted by statesponsored trolls in Russia
saying, “Type Amen if you want Texas to stay Christian,” that’s
what the algorithms will show. Think of it as an emotion pump.
You finish reading a post. Before you can close the app or click to
another browser tab, you scroll some more, almost by reflex. In
that mo ment, Face book injects another post optimized to make
you laugh or get you angry, and the cycle continues. Polarizing
content keeps the pump constantly primed by riling users up.
The side effects of this strategy have become plain in nations
such as Myanmar, the Philippines and the U.S., where misinforma
tion shared on Face book has fueled division and social unrest.
Kathleen Hall Jamieson, a professor of communications at the Uni
versity of Pennsylvania, concluded in her 2018 book Cyberwar that
Russiansponsored Face book ads and posts swayed the outcome of
the 2016 presidential election. Nobody at Face book anticipated
these effects. But they can’t be swept under the rug—and they can’t
be solved through minor algorithmic adjustments, because this is
Face book. The emotion pump is at the core of its business model.
Are there other ways to design social media networks? Yes and
no. For years I enjoyed a smartphonebased social network called
Path, which was conceived in 2010 by Face book alumnus Dave
Morin as a kind of unFace book and a home for smaller groups (it
initially limited users’ networks to just 50 people). Path had an en
chanting interface, but it never found a solid source of revenue,
and it shut down in 2018. The Diaspora project raised $200,000
on Kickstarter in 2010; its vision was to build a decentralized net
work where users would run their own servers, or “pods,” and con
trol their own data. It still exists as an opensource project, but the
difficulty of setting up a Diaspora pod has kept its user base small.
The failure of these smallscale networks doesn’t bode well for
Mark Zuckerberg’s plan, announced in March, to remake Facebook
around messaging within small, private groups.
Without revenue from emotionpumped advertising, Facebook
would wither, and there could never be another socialnetwork
ing company that reaches its planetary scale. But I believe those
would be good things. Face book does only one thing well: it keeps
you from falling out of touch with people you don’t see very often.
There are smallerscale services that serve the same end, howev
er, without the risk of blowing up our democracies.
For instance, after I decided to leave Face book, my family began
using GroupMe, a free grouptexting app owned by Microsoft. It’s
simple, but for the photos and up dates that we formerly shared on
Face book, it’s fine. To share news with people who’ve asked to fol
low my writing or pod cast projects, I’ve used platforms such as
Google Groups and Mailchimp. I even send out an occasional per
sonal email or (gasp!) handwritten note.
I was interested to hear about the finding by University of Ox
ford evolutionary psychologist Robin Dunbar that Face book hasn’t
actually enhanced our capacities as networkers. In a 2016 study of
2,000 users, respondents told Dunbar that only 28 percent of their
Face book friends could be considered “genuine” or close friends.
That fits with my experience. The other 72 percent—let’s be honest
here—aren’t worth a big cognitive investment, and they wouldn’t
be in our circles at all unless the technology made it so easy.
In the past year I’ve lost two dear friends from former work
places. They both died after brief, sudden, shocking illnesses. I did
not learn about their deaths from my Face book news feed. Friends,
colleagues and family members reached out to me directly, and we
shared our memories and grief through email, calls and visits.
That’s how society functioned before Face book. And these
skills can resurface—but not until we reclaim some of the energy
captured by the emotion pump.
JOIN THE CONVERSATION ONLINE
Visit Scientific American on Facebook and Twitter
or send a letter to the editor: [email protected]
© 2019 Scientific American