Ep145 Why do we compulsively click on ragebait? with Angele Christin

Do algorithms shape our lives? What did clickbait look like before the internet? Why do journalists start writing differently when metrics are introduced? What does any of this have to do with cooking pasta in the bathtub, the actress  Sarah Bernhardt, or Oxford English Dictionary’s word of the year? Join Eagleman with sociologist Angele Cristin to learn how algorithms invisibly sculpt our behavior.

Episode Transcript

Speaker 1 (00:05):
How do the algorithms that were surrounded with shape our lives?
Why does a journalist begin to write differently once metrics
are introduced? What is algorithmic capture? Has clickbait always existed?
What did that look like before the Internet? How could
even vegan content creators go in for rage bait? What

(00:28):
does any of this have to do with why someone
would insist on cooking pasta in the bathtub? Or the
nineteenth century actress Sarah Bernhardt or one weird trick or
Oxford English Dictionary's word of the year. Join me today
with sociologist Angel Christen as we talk about how algorithms

(00:48):
invisibly sculpt human behavior. Welcome to Intercosmos with me David Eagleman.
I'm a neuroscientist and an author at Stanford An. These episodes,
we sail deeply into our three pound universe to understand
why and how our lives look the way they do.

(01:24):
Every brain is trying to navigate itself through a very
social world, and it's always unconsciously making predictions to figure
things out like what should I pay attention to or
what is worth my time?

Speaker 2 (01:38):
What should I care about?

Speaker 1 (01:40):
And for essentially the entirety of human history, those signals
came from other humans.

Speaker 2 (01:46):
You read a lot.

Speaker 1 (01:47):
Of information from the expressions on their faces like a
raised eyebrow. You hear things in the tone of people's voices.
You modify your behavior if you are getting signals of
approof or disapproval, not just from your parents, but more
generally from your tribe. Our brains have worked with these
sorts of signals for millions of years. You're looking for

(02:11):
a nod of approval, or people laugh at what you say,
or they scowl at it or whatever. But now imagine
replacing all that with numbers. Now you've got a dashboard
and a counter that ticks upward, and you've got a
graph that rises or falls, and it's all in real time,

(02:32):
and you have a notification that says three four hundred
people watched this or twelve people watched this. So, all
of a sudden, in this past nanosecond of human evolution,
we are suddenly in an era when the signals that
guide our behavior are not primarily human faces, but instead metrics.

(02:56):
You've got numbers and rankings, you've got dashboards, and it's
based on algorithms that you can't really see. And these
algorithms decide what appears in front of us and what disappears.
And here's the wild part. When those signals change, we change,
often unconsciously. Writers will change what they write based on

(03:19):
the feedback. Creators will change what they're about based on
the feedback. Sometimes there are subtle ways this happens, like
someone tweaks the headline, or they make a video shorter,
or they spend more time thinking about a stronger hook.
But pretty quickly, entire professions can reorganize themselves around these

(03:41):
invisible feedback systems. So what happens now that algorithms are
the primary environment that we're swimming in. What happens to
professional values, what happens to identity? Just think about one
example of this. One of the brand new words that
we all witness the birth of a couple decades ago
was clickbait. As you presumably know, this is one of

(04:04):
those provocative headlines like you'll never believe what happens next,
and this gets you to go down some rabbit hole
that you didn't intend to go down. But now clickbait
has evolved into Oxford English Dictionaries word of the.

Speaker 2 (04:21):
Year last year, which is rage bait.

Speaker 1 (04:24):
This is the next evolutionary step in the attention economy.
Rage bait is all about shouting something, some statement that's
so inflammatory or so absurd that your nervous system lights
up before your more thoughtful neural networks even get their
shoes on. So you feel a surge of whatever anger

(04:46):
or discuss or disbelief, and before you're consciously aware of it,
your fingers are clicking and you're commenting, you're sharing, you're
stitching a video, you're firing back at someone in the
comment section.

Speaker 2 (05:00):
You're all worked up.

Speaker 1 (05:01):
But from the algorithm's point of view, engagement is engagement,
and in a system that measures success by interaction, provoking
emotion is a very effective strategy. So I'm fascinated with
what all this means for us as a society. And
I knew there was no one better to call for
this than Onzel. Christen Angel is a sociologist who has

(05:25):
spent years embedded inside newsrooms, inside influencer culture, inside the
worlds where algorithms and analytics shape what people produce and
how they see themselves on shell studies, incentives, and how
our digital infrastructures are reshaping the social fabric around us.

(05:46):
So I'm so pleased to sit down with her today
to talk about what happens when human brains live inside
algorithmic systems and how those systems are shaping our behavior. So, Angel,
you study algorithms and their social impact on people, tell

(06:08):
us about that.

Speaker 3 (06:09):
For the past ten years, I've been looking at how
algorithmic technologies are playing an increasingly important role in people's
lives and in their work. So now the thing is,
I'm a sociologist by training, so I really care about
people's identities, their work practices, how they communicate and relate
to others, and all of that has been completely transformed

(06:32):
with digital technologies and the computational procedure sustaining them. Everything
we do now is mediated through computational procedures one way
or another, thanks to mobile devices on the one hand
and computers on the other.

Speaker 2 (06:47):
So let's take a specific example.

Speaker 1 (06:49):
So you got interested some years ago in how journalists
were writing their articles and what kind of things they
chose to write, and how that changed once these with
the metrics got introduced.

Speaker 4 (07:02):
Yeah, so that was so interesting. At first.

Speaker 3 (07:05):
I came to kind of study journalism because I was, like,
you know, with the Internet and kind of online advertising,
the business models of the news are changing right in
Big Park because online advertising move to social media platforms
and Google, and so they had news websites had less
money and newspapers had less money, and so I was

(07:25):
kind of broadly interested in that and like, what does
that mean for the future of the news. And I
did what I always do. I'm an ethnographer, which means
that I go spend time with people. I interview them,
I have lunch with them, I observe how they work.
I become kind of part of the furniture at the office.
It's like a yeah, she's here again, you know, just

(07:47):
observing us another day. And what I realized at that
point is that in the different newsrooms that I was studying,
some of them were in New York, some of them
were in Paris, lists kept on looking at these specific
software programs that they were kind of obsessed with. And
so what were the software programs? They were web analytics,

(08:11):
software programs that were providing detailed, real time data about
what readers were doing on the website. Right, So it
was giving kind of social media metrics like how many
readers were coming from Facebook, from Twitter, etc. It was
telling people how many people had read the articles they
had published, right, was it ten people, twenty people, one

(08:34):
hundred thousand, ten hundred thousand, a million, right, So, giving
that it was ranking all the articles in terms of
their popularity on the website and on social media, and
it was also telling journalists how long readers were spending
on average on any given article, which, by the way,
is not a good metric in the sense that like

(08:57):
it's almost always in the range of like six to
fifteen seconds is the average time that readers spent on
any news article. Right, Yeah, not fun for journalists, definitely.
They preferred to ignore that one. And so what I
realized is that journalists kept on talking about that. So
that was the early two thousand tents, right. People didn't

(09:19):
really know. I mean, the internet was still pretty new, right,
and social media was even newer. But journalists were just like, wow,
look like that article is so popular. It's just like it's,
you know, such a hit, everybody's clicking on it. Well
as this article hmmm, disappointing. I spent two weeks reporting

(09:40):
this and no one's clicking on it. And so they
kept on discussing this and making jokes about it, talking
about it and kind of more and more using this
data to make sense of what it meant to be
a good journalist.

Speaker 2 (09:55):
Yeah.

Speaker 1 (09:56):
So first, what was the difference before if you you
were a newspaper writer, let's say twenty years before that,
how did you get feedback?

Speaker 4 (10:05):
Yeah, so that's the really interesting thing.

Speaker 3 (10:07):
So what's really fun there, and that's you know, the
great point of being in academia is that other people
have studied print newsrooms in the eighties and nineties, right,
so they have all of these historical material about how
print newsrooms used to work. Now, the print newsrooms, the
way it used to work was two characteristics. First, journalists
and editors did not care much about the audience, right,

(10:29):
And they didn't care mostly because they thought that what
mattered more was the opinion of their peers and superiors.
So there was this kind of strong professional identities that
what was really important was what other journalists were thinking
about you, both in new newsroom and in other newsrooms. Right,

(10:51):
because journalists always tried to move between one newsroom and
the next.

Speaker 4 (10:54):
So let's say you had the New York Times.

Speaker 3 (10:56):
You also care about what people at the Washington Post
think about your work, because if they like it, perhaps
they'll offer your job later on that kind of stuff,
and then what your.

Speaker 4 (11:04):
Hierarchical superiors total of your work.

Speaker 3 (11:06):
So your editors in chief basically, right, and if your
editor in chief was saying like, wow, this is a
great article, well done, that was that you know, that
was kind of the best things that could happen to
you on any given day. Now, they were receiving letters
from the audience, right. So letters to the editor are
like kind of a long established tradition, most newspapers have them,

(11:30):
but there are all of these like historical studies showing
that journalists mostly disregarded these letters and they're like, oh yeah,
people are crazy, they're understand those crazy stuff. They don't
understand how journalism works, right, So a very different kind
of relationship where really what the audience was actually doing

(11:51):
and what it wanted wasn't very well understood or kind
of paid attention to. The second aspect, there was also
that with newspapers there were lots of surveys obviously, because
you had marketing departments and you had like kind of
commercial the commercial side of publications, right, and they cared
about the audience because that's how depending on you know,

(12:12):
readership figures. That's how they were able to charge for
more expensive advertising space, right, And so that's what revenues
depended on. So the marketing departments were doing surveys. But
what's interesting about these surveys is that you know, in
the eighteen and nineties, they only asked like, well, do
you like the newspaper and people would say yeah, But

(12:35):
they didn't have super granular data about well, which section
exactly do you read the most People would say like, well,
based on memory, I tend to read more of the lifestyle.
Say but you know, I really care about international news too, Right,
So people they were relying on this kind of post
hoc memories and recollections of what readers were doing, and
they didn't really know in detail which articles and which

(12:58):
topics they cared the most. Now comes the Internet and
comes web analytics and the kind of software programs and
algorithms kind of displaying with dashboards what these analytics look
like to journalists and editors. Suddenly they realize that no
one's clicking on international news. Very few people click on

(13:19):
political news unless it involves some kind of scandal or
sex or celebrities. Right, that people care a lot about
lifestyle news. They have a lot about celebrity news. They
care not at all about culture and the arts. That's
not really what people click on, right, And so suddenly

(13:39):
they have this kind of like almost tsunami of data
telling them what people are actually clicking on.

Speaker 4 (13:46):
And that's kind of a cognitive.

Speaker 3 (13:48):
Revolution for journalists, right, because suddenly they're just like, wait,
what should we do? Should we only do tabloid news
because that's what readers want?

Speaker 4 (13:58):
Or is our jobs something else?

Speaker 1 (14:01):
And you know, it's interesting because from a behavioral economics perspective,
even if you ask readers in a very granular fashion
what they wanted and so on, they don't necessarily know that.
Consciously they might think I'm the kind of person who
clicks on international news, even though they don't. Okay, So
there's this notion of algorithmic capture, which is that let's
say you're a writer, you start chasing the numbers and

(14:24):
you end up doing what the algorithm wants. Now, one
might say, okay, but maybe there was something like editorial capture.
The editor who's seen or to you, who really admired
you end up doing what he or she wanted.

Speaker 2 (14:37):
The problem is that the algorithm is.

Speaker 1 (14:39):
More opaque, right, and it changes, it's fungible.

Speaker 2 (14:43):
So tell us how you see that?

Speaker 3 (14:44):
Yeah, So that's fascinating because that's exactly what I observed
in newsrooms, and that's what my book Metrics at Work
Journalism is a contested meaning of algorithms is basically all
about right. So what I find kind of due to
that kind of again that wave of granular data that
kind of enter newsrooms, and you know often it entered

(15:05):
newsrooms with the best intentions, right, which was that editors,
journalists and the companies that were that were usually for
profit companies that were providing these kind of dashboards and
software programs. We're like, listen, it's a question of democratization.
We want to give journalists the knowledge of what like

(15:26):
their audiences are actually engaging with so that they can
survive in the digital age, right, so that they can
drive in the digital age. So that was the impulse,
a good one, right. But then what you see is
that once that data is available, you get exactly the
kind of algorithm captures that you find, right, And there

(15:46):
is this.

Speaker 4 (15:47):
Like push towards clickbait.

Speaker 3 (15:49):
It's just like it's and it happens at many different
levels individually it happens because journalists, like all communicators, and
I include academics in that, they want to be read.
What's the point of being a journalist if no one
reads your stuff, you know what I mean? That's kind
of an existential question. And so to me, it makes
perfect sense that you would be like, well, I'm going

(16:11):
to do everything in my power so that you know,
my articles find the audience, right, and if that means
putting a slightly more clickbaita title, so be it.

Speaker 1 (16:21):
And clickbaity is things like here's one weird thing you
can do, and so what else.

Speaker 2 (16:26):
Do you see?

Speaker 3 (16:27):
Yeah, so at the time, Okay, so that's going to
be a bit dated now because again this was you know,
the bright days of the early twenty tents, right, but
the big things where yeah, this one wheel trick you know,
you'll never believe. You know, what happened next was shocking listicles.
We're a big thing, right, like the ten best restaurants

(16:48):
in San Francisco, you know kind of stuff. Slide shows too, right,
because at the time, a lot of newsrooms were still
counting page views, and so a slide show with ten
pages counted as ten phage views. I know, it's really
kind of a cheap trick, so kind of things like that, right,
And you could see these formats becoming widely popular across

(17:12):
newsrooms because newsrooms are also all watching each other and
seeing especially at the time, it was Busfeed and the
Huffington posts using these kind of more clickbaity tactics and
getting the traffic numbers that were coming with from it,
especially on Facebook and Twitter, and everybody was like, oh, okay,
we've got to do this, right, So then a couple

(17:32):
of things kind of happen. So first one, which is
exactly what you said, is that you know, social media
traffic and social media algorithms are fiical. They have gotten
only morephical since, right, But they're opaque, so no one
really understands what's getting recommended. And you know, platforms have

(17:53):
given some modicum of information, but they also don't want
users to gain the system, so they're not given that
much information.

Speaker 4 (18:01):
Right.

Speaker 3 (18:02):
Then, social media platforms also made sudden changes over the
years that really hurt newsrooms. For example, Facebook, at some
point in the mid twenty tens decided that it was
going to stop prioritizing posts and articles from news organizations
and instead promote relatively more organic content, namely content coming

(18:28):
from individuals, and that's where influencers started driving.

Speaker 4 (18:32):
But that's for a bit later.

Speaker 3 (18:34):
But so from one day to the next, basically you
had like all news websites because traffic was like that,
just like going way kind of further down, just because
in one change in the Facebook algorithm that cut traffic
can of in half for many newsrooms. Right, So it's
fickle and things are changing, and what works well as
a recipe for traffic at one point stops working, and

(18:57):
everybody's kind of in the dark and trying to figure
out what happens. So that's kind of the first problem.
The second problem is that, you know, journalists, we're also
thinking really hard about wait, like, if we follow what
the algorithms want, we're only going to do tablauid news right,

(19:17):
namely against sex, kndle celebrities that like, you know, it
always people are always going to click more on that,
you know, but this is not what our role is.

Speaker 4 (19:27):
At the end of the day. We became journalists because
we care.

Speaker 3 (19:30):
About democracy, holding the state and the powerful accountable, about
covering kind of longer stories, more complex stories, right, And
so what I kind of observed during kind of all
the time I spent in newsrooms is that basically journalists
were torn between these different definitions of what their professional

(19:55):
identity was about. On the one hand, they were like, well,
we're professional communicating, so we want our stuff to be read,
and so if clickbait has to happen, we will do
it because it's important we want to reach our audience.
But on the other hand, they were also saying, well,
we care about sharing important things that will make our

(20:16):
readers more enlightened citizens of the world. And so they
were kind of always torn between these two kind of
definitions and kind of moving back and forth. And what
was fascinating that they would often do like funny little
accounting things where they would do like, Okay, so I
really want to write this article about say, you know

(20:39):
again international news Ukraine, let's say, right, because I really
think it's important and I think our readers need to
know about this. But I know that people are not
going to click a lot right, So to make up
for it, I'm also going to write an article on't
say Rhianna right, and that's going to boost my traffic numbers,
And that way I've kind of subsidized my Ukraine article

(21:03):
with the Rhanna article in terms of traffic, So they
would do like all of these little deals with themselves
to try to optimize both sets of kind of constraints.

Speaker 1 (21:29):
What was the equivalent of clickbait back in the day
of newspapers, because there was what's called yellow journalism. I mean,
there must have been ways that people are trying to
get readers, and of course there were tabloid presses.

Speaker 3 (21:40):
Yeah, absolutely, so there were many equivalents, and so that's
the same clickbait has always existed in many ways. So
at the time, let's see, there was a lot of
kind of as you say, yellow journalism, so extremely outrageous,
like inflammatory, kind of aggressive headlines, you know, really kind

(22:03):
of emphasizing kind of hatred right or like kind of xenophobia, bigotry,
like you know, things like that have always existed in
many newspapers over the past two centuries have really indulged
in that because again it's like stokes kind of patriotic
sentiments for example, and people are like, yeah, like you know,
I'm gonna buy this newspaper. So that's kind of one aspect. Definitely,

(22:25):
coverage of stars and celebrities, for example, at the time
Sarah Bernhardt, you know, in the late nineteenth century was
like the idol. All kind of teenage women were doing clippings,
newspaper clippings of the actress Sarah Bernhart, that was the thing,
and so newspapers would always put lots of pictures of her,

(22:47):
so that like the news that teenage girls would do
the clippings, you know, so they would force their parents
to buy the newspapers so that they could do that.
Same with lifestyle and kind of beauty advice, same things.
That's always been very popular, cooking recipes, things like that,
you know what I mean. So it's not new, it's
just the shape it takes and the specific tricks that

(23:11):
people use to capture the attention changes with the medium,
you know what I mean, from print to broadcast to digital.
But again I think it's a change in degree, not
really a changing kind, if that makes sense.

Speaker 1 (23:26):
So when you published your book in twenty twenty, you
were interested in this issue about clickbait and how people
are getting sucked into this whirlpool of that. Now we're
in twenty twenty six. So last year Oxford English Dictionary's
word of the Year was rage bait, which is close
cousins to clickbait tell us about rage bait.

Speaker 3 (23:45):
So rage bait I think of as a kind of
like vengeful cousin of clickbait.

Speaker 4 (23:52):
Right, well, clickbait was.

Speaker 3 (23:54):
Still kind of light. You will never believe what happens next.
You will be shocked ten weird tricks to whatever do something,
you know. Rage bait is more about provoking you into
kind of negative, kind of arousal, right, negative emotions such

(24:14):
as anger, fear, hatred, again, not things that are radically new,
but really kind of pushing on that. The typical example
of rage bait is going to be an influencer or
someone on the internet just stating something outrageous, right, or

(24:34):
doing something outrageous but with complete confidence and innocence and
not seeming.

Speaker 4 (24:40):
To see the problem.

Speaker 3 (24:42):
Right. So there are a number of examples of that.
I mean, one of the ones that I like are
like these weird cooking videos, but where they cook things like,
for example, like in the bathroom using.

Speaker 4 (24:55):
The baths stub and doing like this is really.

Speaker 3 (24:57):
The most efficient way to cook pasta, things like that,
and like kind of really going going fully in you
know what I mean, and not acknowledging how weird it
is to try to cook pasta in the bastub or
whatever it is, or you know, I mean, there are
lots of examples. And what's really interesting about this type
of content is that it's typically produced by content creators

(25:22):
who know very well that it's going to infuriate people, right,
it's going to make them mad. But because it's going
to make them mad, they're gonna engage with the content.
They're gonna watch it, they're gonna comment on it, they're
gonna share it, all kind of basically yelling insults, you
know what I mean, and saying like how dare you?
How could you use? This is so gross? I can't

(25:44):
believe you.

Speaker 4 (25:46):
Are you a troll? Why are you saying this? How
can you say this?

Speaker 3 (25:50):
But all of this obviously translates into revenues and engagement
on social media platforms, and so that's very much a
deliberate kind of content productions.

Speaker 1 (26:01):
When you think about rage bait, what percentage of that
is political versus I hadn't thought about the cooking past
of the bathtub.

Speaker 4 (26:08):
That's a good question.

Speaker 2 (26:10):
When I think about it.

Speaker 1 (26:11):
I think people saying things politically that are very inflammatory.

Speaker 2 (26:16):
But you I'm glad to see you have a broader
view of rage based.

Speaker 4 (26:19):
I definitely have a broader view.

Speaker 3 (26:21):
I mean I think so, I think it's a continuum, right,
always think about it is that it's really a spectrum
from you know what psychologist called kind of high arousal
negative content. Right, So that's going to be anything inflammatory, incendiary,
shocking in a negative way. Right, wh It just makes
you kind of freeze and it kind of bypasses whichever

(26:44):
defenses you may have, and you're just like, this is
serious and bad.

Speaker 4 (26:49):
I need to pay attention.

Speaker 3 (26:51):
Right, It's almost like an embodied response, right that just
like shotcuts kind of a lot of the usual ways
we have of interaction acting with media. So that's kind
of one way. And then like rage bait, I think
it's a bit of a wider kind of type of
content because it also involves other kind of emotions such

(27:14):
as like again discussed weirdness, like you know, a wider
set of reactions that are also negative but are not
necessarily about politics specifically.

Speaker 1 (27:29):
Now, does it make sense for anyone on social media
to do clickbait and rage bait?

Speaker 2 (27:35):
Is it?

Speaker 1 (27:35):
Is it sensible? Or are we all sinking into a morass?

Speaker 4 (27:40):
This is such an excellent question.

Speaker 3 (27:42):
So it depends on the time and frame you're looking at.
And again, to some extent, you know, this is a
question that most kind of content creators have to deal
with at this point, and it also kind of resembles
the questions that newsrooms at the time had to deal with,
but in a bit of a different way.

Speaker 4 (28:02):
So let me walk through that.

Speaker 3 (28:04):
So for content creators, it depends on how you monetize
your online production, right, if you monetize it through platform payments,
so things like the YouTube Partner program, right where if
you qualify you have to have a channel that's kind
of big enough, but if you do qualify, you can

(28:25):
get fifty five percent of the advertising revenues that your
videos gather. Right, So in that case, basically your revenues
are going to depend on how many views you get
and watch time, right Obviously, that's that's going to be
the main thing. So in that case, rage bait in

(28:46):
the short term amazing, right, because you're going to provoke
people into watching you hook them in with something truly
kind of atrocious, you know, or really kind of hair
raising kind of kind of stuff that like makes it
so that like you can't look away. Literally, it's like
looking at a car crash, you know what I mean.
You you kind of have to watch it just to

(29:08):
see what's gonna happen next and which kind of insane
things they are gonna be saying next, And then you
keep on engaging true comments and sharing, et cetera, making
them the video even more popular because that's what recommendation
algorithms are gonna pick up on. They're gonna look at
all of these engagement signals right, and they're gonna promote
it more widely, making it go viral. So all of

(29:30):
this is great for watch time. All of this is
great for platform payments in the short term. Now the
question is can you make a living for several years
based on that type of content? And that's where it
becomes a bit more complicated, because you see, you can
shock people once, you can shock people twice, you can
shock people three times. But then well, likely they're just

(29:53):
gonna stop watching because gonna be like, oh, well that's
that person doing the shocking thing again. But you know, avoi,
I'm not going to click and I'm not going to
give them that attention. So what I've seen kind of
in my world is that often What happens is that
over time, influencers or content creators move from only relying

(30:16):
on platform payments and views and watch time as a
way to kind of really make a living to becoming
more either what I call gurus, so trying to really
monetize their audience directly through especially donations subscriptions, and so

(30:37):
that entails a slightly different way of relating to the
audience and producing content where it's not so much rage
baits and kind of this kind of negative shocking engagement,
but it's more framing things as it's us against them, right,
So basically saying I'm speaking truth to power, I'm saying

(30:57):
things like it is, and you following me because you
like that I'm uncensored. You like that I'm saying things
is the way they should be said. And so if
you want to support my mission, please subscribe to my channel,
give me a donation, subscribe to my Patreon, and kind
of support me financially. But it's a bit of a

(31:18):
different thing than the rage base, you see what I mean,
because it's more about loyalty at that point.

Speaker 2 (31:23):
Yeah, now you've studied that.

Speaker 1 (31:25):
You spend a lot of time studying influencers or content
creators And am I correct that you see there's a
lot of anxiety in that population.

Speaker 4 (31:34):
Absolutely.

Speaker 3 (31:35):
So that was kind of my next project after journalists.
And you know the thing with journalists is that they
themselves could see the writing on the wall and they
were like, oh, we are all going to become influencers basically,
and you know, it's something that newsrooms have been doing
ever since. Basically they show more and more video content

(31:56):
can of on their homepages, and they send kind of
journalists to have more like journalist slash influencer profiles in
order to try to reach young audiences on TikTok for example. Right, So, anyway,
so just like, oh, so what about content creators? So
I spent six years studying content creators and doing interviews

(32:17):
with them, observing them as they were shooting videos and
kind of producing them, doing interviews with brands, with platform employees,
and just trying to understand the careers of influencers, right,
and like, how do you make money as a content
creator and more importantly, how do you keep on making
money over time? Because it's one thing to make money

(32:40):
for one year, it's another thing to make money and
make a living and sustain yourself for five, seven, ten years,
and so yeah, what I found basically is this big,
like one of my first findings, is this big disconnect
between what people imagine being a content creator is like
and what the real reality of social media labor looks like.

(33:03):
So almost everybody at this point thinks that being a
content creator sounds like the most fun job in the world.
In a recent survey from twenty twenty three, more than
fifty percent of Americans said that they would like to
become content creators. So this is a massive kind of

(33:24):
pull on the collective imagination. Right, We're all like, oh
my gosh, they're having so much fun. They get to
follow their passions, they get to have an amazing life,
to talk about what they care about. They kind of
look great, right, They're just having so much fun. They're
hanging out with their friends. I want to do that,
and I want to have the money, the glory, the fame,

(33:45):
the luxurious lifestyle that comes with it. And so what
you see is more and more people entering into social
media creation and trying to become an entrepreneur right and
saying I want to be my own boss, I want
to be a content creator, and I want to follow
and of my passion took about the topics I'm truly
interested in, and that my day job is not recognizing. Okay,

(34:06):
so big pool, a big funnel that kind of brings
in lots and lots of people to social media creation.
What happens, though, is that it turns out making a
living as a content creator is very hard. And why
is it very hard? Well, mostly again because of social
media algorithms and the fact that it's really hard to

(34:28):
figure out what people want to watch and what platforms
are rewarding in terms of content, right, and what's getting promoted,
especially true recommendation algorithms and true feed algorithms that kind
of gonna, you know, are going to present you with
the most kind of viral and popular videos of the day.
And so for content creators, what that means is that

(34:50):
very often they're like, oh, well, you know, I'm going
to try, right, And so they try. And let's say
they get a bit of success at the beginning, right,
and they have a couple of videos that go viral,
and let's say they start kind of getting some followers,
and so that looks pretty good, right, And let's say
is they have brands that start coming to you and
are like, oh, do you want to work with us, Like,

(35:10):
we'll send you sunglasses, you know, and eventually a little
bit of payment. But soon what happens is that, you know,
creators realize that it's a twenty four to seven kind
of job in the sense of in order for some
of your videos to go viral, in order for your
number of subscribers and followers to kind of keep increasing,

(35:32):
which is what brands want to see. They want to
be associated with accounts that are going up, not with
accounts that are going down in terms of audience measurements.

Speaker 4 (35:40):
Right, you have to.

Speaker 3 (35:42):
Post videos all the time because most of these videos
will not go viral, most of them will not find
their audience, right. But you are kind of a production
company of one person, and so that means that you
have to post every day.

Speaker 4 (35:57):
And so again, what.

Speaker 3 (35:58):
I see in my interviews is creator saying I made
myself sick. I try to post every day, twice a day,
just to keep that rhythm. I was on my own
in my bedroom and it just became too much. I
just couldn't deal with it. So that's the first thing.
It's a relentless rhysm. It's a rhythm that would be
very hard for a TV network or like a newsroom

(36:22):
to kind of keep doing. And we're talking here about
people who are alone. They're like a company of one
in their bedroom, right and just pushing themselves and saying, like,
you know, I'm an entrepreneur, I.

Speaker 4 (36:33):
Have to make it work, right. So that's the first thing.

Speaker 3 (36:36):
The second thing that happens is that as they get
kind of more popular and the size of their following
kind of increases, most of them start receiving hate comments,
right and start kind of facing the darker side of
the internet. So at first they're like, oh, I love
my community, I love my followers. We share the same

(36:56):
passion for X topic, right, regardless whether the topic is
like beauty or K pop or you know, DYI kind
of you know, reroof your house. I mean, it doesn't matter.
And they're just like, yeah, there are such good people,
you know, We've become really close. I love them, they
love me. We really get each other in this deep,

(37:18):
authentic way.

Speaker 4 (37:18):
Right.

Speaker 3 (37:19):
But as your fame grows and as you start to
make more and more money from social media production, suddenly
something switches in the relationship with your followers. Namely, is
that First, you can't keep track of all of them,
so it becomes more of a faceless mass of people, right,
and you know still a few of them from the beginnings, right,

(37:40):
but most of them have just become numbers. That's the
first thing. And second, negative feedback starts coming in. Typically,
you're gonna get kind of hate comments, you're gonna get destreats,
You're gonna get themeaning comments about your physical appearance or
your mental health or whatever it is that you know
comes up in your videos. You're going to get stockers,

(38:03):
right who show up kind of in your hometown and
perhaps show up at your doorstep. And suddenly you're just
like wait, like wow, this is really taking a toll
on me, because day after day you get both the
good parts of the Internet and of that kind of
public presence, but also the negative parts and the relentless
like trolling, harassing, insulting, and everything that comes with it,

(38:28):
and that takes such a heavy toll on the mental
health of creators.

Speaker 4 (38:32):
That together with the pressure to publish all the time.

Speaker 1 (38:50):
And from your point of view, the change in the algorithms.
I'm seeing this on X all the time lately, that
creators are complaining, Hey, I was doing so well and
they just change the algorithm. Now suddenly I'm getting shadow
band or whatever term they're using for it.

Speaker 3 (39:03):
Yeah, so I have a whole thing about that. I mean,
so basically, creators talk about the algorithm kind of all
the time, and it's really interesting because it's always the
algorithm singular. It's not like all the computational procedures of
social media platforms, of which are many, right, because you
have feed algorithms, you have recommendation algorithms, you have demonetization algorithms.

(39:27):
So these are basically algorithms that if your video is
deemed to be not family friendly or edgy, or going
kind of being kind of a bit too close to
kind of the edge of community guidelines in a range
of way, you're going to get demonetized, which means that
you won't get any money from the ads getting served

(39:49):
on the videos. Right, So basically for content creators that
means that they're not making money from that video, which
is a big problem for them. And so that's like
demonetization algorithms. And then you content moderation algorithms that are
you know, along the same lines, but with more serious
sanctions such as getting banned, for example, from a given
platform because your content goes against the community guidelines of

(40:11):
the platform. All of these algorithms are kind of changing
all the time they are opake. There is no announcement
of what's happening, right, and so creators are always kind
of like basically putting their finger in the air and
trying to see what does the algorithm want at any
given point in time, right, and trying to intuit what

(40:32):
the algorithm is looking for. And so that's already kind
of quite a thing there. But then what I kind
of argue kind of in my book is that in
a way, the algorithm becomes a shortcut for a more
complex set of actors that you know, influencers and content

(40:53):
creators prefer not to think too closely about. So one
things that content creators do is that they pomorphize the algorithm.
So they keep on saying the algorithm is stupid, the
algorithm is silly, the algorithm doesn't like me, or the
algorithm really likes them. They got really friendly and close
with the algorithm, right. So all of these ways of

(41:14):
projecting human characteristics onto computational procedures that are obviously and
to the best of our knowledge at this point like
not sentient, you know what I mean, they do not
have emotions. But what I argue is that this kind
of process of anthropomorphism is really because so who is
behind the algorithm? Social media companies, right, the alphabets, the meta,

(41:39):
the tiktoks of the world right by dense Now who
else is behind the algorithm?

Speaker 4 (41:47):
The audience?

Speaker 3 (41:49):
Right, I mean basically the algorithm just picking up on
audience signals and now the algorithm, you know, for creators,
blaming the algorithm for everything that goes wrong is a
way to avoid blaming social media companies, which very often
as the one paying them through partner programs, right, and
audiences which also as the one supporting them through their

(42:12):
clicks and their engagement.

Speaker 4 (42:13):
Right.

Speaker 3 (42:14):
So the algorithm is kind of a convenient scapegoat. And
again not to say is that it's not unpredictable, opaque,
and all of these things, But it's kind of a
convenient kind of scapegoat for these broader kind of forces
that really are responsible for the fortunes or lack of
fortunes of content creators.

Speaker 1 (42:33):
By the way, just from a neuroscience point of view,
we always have to do this with systems that are
sufficiently complex, we reduce them to a person because that's
what we're evolved to do, is understand other people.

Speaker 2 (42:43):
Yeah, try exactly, Yeah, yeah, exactly, So it's no surprise
that people do that.

Speaker 1 (42:47):
I want to return to rage bait for a moment
because one of the things I've been so interested in,
and this is on a lot of my podcast, is
about polarization and in groups and out groups and so on.
And one thing that's hard not to notice is what
I've been calling lately the fringe amplification effect, which is
somebody on one side of the political spectrum makes a

(43:08):
totally rage baby video that's just so extreme in their
political view on it, and then the other side says,
look what the other side is doing, and they stitch
in that video and everyone.

Speaker 2 (43:20):
Gets to see that.

Speaker 1 (43:22):
So the videos that you know, one side gets to
see of the other is the most extreme. Yeah, and
that seems to be a real problem because that's driving polarization.

Speaker 4 (43:31):
But again, so I completely agree.

Speaker 3 (43:33):
I mean, I think so the way I kind of
analyze this that all of these stems from the economic
incentives of content creators. And so again we think a
lot of it in terms of like political polarizations, right,
and kind of deeply held beliefs, right about politics, all
about religion, all about immigration, you know, you name it.

(43:54):
But I guess I mean, I guess I'm a bit
more cynical kind of in a way. And that's like
the sociologist in me. I'm like, sure. But that's also
because platforms have put in place economic incentives that basically
encourage content creators to produce this kind of polarizing content.
Again because of like the logic of engagement. So if

(44:15):
you want people to engage with your content, the tendency
and I've seen that in my interviews over and over
and over again. Creators start by saying, Okay, I have
one example. One of my case studies was vegan creators. Okay,
you'll tell me vigan creators.

Speaker 4 (44:31):
How nice.

Speaker 3 (44:32):
Right, They're going to be talking about plant based eating.
It's good for the planet, it's good for the animals,
it's good for your health, you know, with kind of
some precautions around it. So what could go wrong? How
could that lead to polarization?

Speaker 4 (44:49):
Well, it turns out that, you know, almost.

Speaker 3 (44:51):
All the creators I talked about were like, well, okay,
so there are a couple of really outrageous personalities that
are extremely rage baity within the vegan kind of influencer world,
and that basically just like attack other creators and other
kind of people for how they are doing veganism wrong, right,

(45:13):
so extremely kind of contentious, kind of like you're doing
it wrong, this is not the right way of being
kind of a vegan. What they also see is like
the rise of extremely restrictive diets that become more and
more niche, so typically FRUITIVL diets where you only eat fruits,
or raw vegan diets where you only eat row fruits

(45:36):
and veggies, or you know, things like that that become
like increasingly niche and kind of extreme right. And what
like all of my content creators were saying is that,
And often I will add, the creators that are the
most kind of outrageous also start kind of verging into
conspiracy theories Q and non you know at the time

(46:00):
Pizza Gates, I mean, you know, things where it's like,
you know, really blaming the system for everything that goes
wrong and saying like, oh, you should live off grid
and be kind of on this extremely restrictive kind of
vegan diet and everything kind of going together. And what
like my you know, my more moderate interview is were saying,
is like, listen, what you realize is that as soon

(46:23):
as I start covering or reacting or responding or stitching
one of these more extreme influencers and what they say,
my number of views and my watch time go way
up because basically everybody wants to read that, Zaka, You're
going to take down that outrageous influenza. Yes, please, like

(46:45):
you know, I'm going to watch it. And so what
they were saying my more moderate kind of you know,
vegan creators, is that it's really a slippery slope towards
more aggressive, more incendiary, more extreme kind of content. And
even if you start kind of somewhere in the middle,
after months and months of producing videos and seeing what

(47:06):
performs well with the algorithm and audiences and kind of
you know, social media platforms, you kind of go more
towards polarized content. It's just kind of what the system
is designed to kind of reward you for. And so,
you know, we've had social media now for ten to
fifteen years, and what you see is this kind of

(47:27):
war of all against all. But that's in part because
of these economic incentives you know.

Speaker 2 (47:32):
Let me ask you this.

Speaker 1 (47:33):
So in the world of let's say, stand up comedy,
young comedians find that if they do a lot of cussing,
that gets them a lot of attention. But then you
have comedians like let's say Jerry Seinfeld who never cussed
and just kept slow and steady and did did his
thing without blue material and got really successful. What is
is there an equivalent in the algorithmic world.

Speaker 3 (47:56):
So it's really interesting because that's that's a very good point, right,
And I mean I think that in general, you have
this moderating force a bit like my journalist at the time,
you know what I mean that even as some of
them are going the road of clickbait, there was also
a significant kind of pushback, you know, saying this is
not who we are, you know, and this is not

(48:16):
what we care about. And at the end of the day,
I became a journalist for different reasons, not to cover
like you know, X scandal, like this kind of tableauid
manner day after day with influences. What I do find,
and it is kind of interesting because it's you know,
slightly different, is that there is a moderating force, but

(48:36):
it's not a neutral one. The moderating force is brands
and sponsored content. So, as an influencer, how do you
make money? Either you do platform.

Speaker 4 (48:47):
Payments, right, and you receive money.

Speaker 3 (48:49):
Directly from meta or YouTube or TikTok based on the
number of views and the watch times that your videos get.
Oh the kind of actually the main way which influencers
make money is by contracting with brands to promote whatever
the brand is doing and kind of promoting it kind

(49:10):
of in their videos, in their posts to their followers. Now,
what's interesting about that? And you know, for anyone who
kind of knows a bit about advertising, it's not that surprising.
But brands hate polarization, They hate scandals because they want
to appeal to the wider possible customer base that's out there,

(49:34):
and they do not want to be associated with things
that could raise red flags among their customers or for
that matter, amongst their boards or you know kind of
financial kind of stakeholders.

Speaker 4 (49:47):
Right.

Speaker 3 (49:48):
And so what I find is that over time you
kind of see two different profiles of content creators that emerge.
You have the ones who really depend on virality, and
they are gonna go more and more extreme. It's really
hard to fight against that. Again, there are some exceptions,
but it's just really hard to fight against it because

(50:10):
everything pushes you in that direction. It's like not nuge
after nudge. But you have a second category where they
go mostly towards branded content, right, and for people who
go more the route of sponsored content and working with brands,
actually avoiding any whiff of you know, scandal, harassment, polarization,

(50:35):
anything and of outrageous is really important because otherwise brands
are gonna stop working with them and they're gonna lose revenues.
So you do have these different polls there.

Speaker 1 (50:45):
So is that enough of a force that we can
have hope that we won't just sink completely into clickbait
rage bait world? Or are there things that social media
companies can do, slash should do, slash might do someday
that would change this?

Speaker 4 (51:02):
This is an excellent question.

Speaker 3 (51:03):
So I think brands in themselves are not enough for
two reasons. First, because you know, especially over the past
year and a half, brands themselves have become politically polarized. Right,
So you have big companies taking sides on one side

(51:25):
or the other, and especially I mean I'm thinking on
the kind of conservative side. The number of brands have
aligned with you know, extremely misogenistic, anti immigrant and violent
forms of content on social media and actually take pride
in it because they're like, these are going to be
our markets, right.

Speaker 2 (51:43):
Like, which brands can so right now.

Speaker 3 (51:45):
Actually I forget them, and I remember there is a
coffee brand that's like specifically marketed like you know, far right,
and they're basically like, yeah, the more extreme the better.

Speaker 4 (51:56):
This is where we want to advertise.

Speaker 3 (51:57):
This is the audience we want to get, and so
they're kind of finding a bit of a niche market there.

Speaker 1 (52:01):
I mean, presumably there's brands on the lab Yeah.

Speaker 3 (52:05):
Definitely obviously like brands on the left to our costco right,
like also kind of becoming more aligned with kind of
you know, liberal kind of attitudes kind of in recent years,
so you know, you do have you do see kind
of that the transformations of politics, in part with social media,
are also affecting brick and mortar kind of companies, right

(52:25):
that themselves kind of end up having to pick aside
a bit more. So that's not so good, right in
terms of reducing polarization. If anything, it might be that
like this is the future we're moving towards.

Speaker 4 (52:35):
So that's the first thing, become.

Speaker 2 (52:37):
Part of your identity to drink this coffee.

Speaker 3 (52:39):
I'll go to that supermarket, right, So that's kind of
that's kind of one one way, So that's not perfect.
The second thing with brands is that you know, again
brands hate politics in general, right, most brands hate politics.
And do we want content creators who i are incredibly

(53:01):
incendiary and politicized or are not politicized at all, you
know what I mean? You also probably want a healthy
middle of people who are voicing kind of political questions
and concerns and kind of hashing it out kind of
on social media in a respectful and civic way. But
this is not what brands want. Most brands again, do

(53:23):
not really want that. So it's not perfect. I think
for me, you know, thinking about what could be done,
it's a complicated thing. And as often, you know, social
scientists are very good at kind of diagnosing what's going wrong,
but less good as saying like and this is what
we should do now. But I am kind of mildly
optimistic about subscription systems.

Speaker 4 (53:47):
I mean, you know, I'm thinking.

Speaker 3 (53:48):
Of substack, for example, where you know, a lot of
content creators, a lot of writers, have kind of gone
so low right and created newsletters on substanles and monetizing
that true kind of annual subscriptions. What I like about
that is that these kinds of annual subscriptions do not

(54:09):
create exactly the same kind of incentives as clickbait, right,
because you have your loyal audience and you want to
keep on delivering quality content for them, right, And so
again it's a bit more like traditional newspapers and memberships.
It's like you kind of have a sense of who

(54:30):
your loyal audience is and you're trying to do the
best you can for them.

Speaker 4 (54:36):
And so I see that as like, you know, one
way in.

Speaker 3 (54:40):
Which the incentives are a bit different, But then it
also raises lots of questions.

Speaker 1 (54:44):
So I think it's a really good point about substack.
But not knowing the numbers of the time ahead, I
have to imagine that substack is so much smaller than
say TikTok or Instagram. What do you see if you're
thinking five ten years out about the future of this.
Presumably short form media is going to stay around. I
think people love the choice involved in it.

Speaker 2 (55:05):
Is anything going to change in some meaningful way?

Speaker 3 (55:08):
So one of the big changes that has happened over
the past ten years and which I think is going
to continue. And it's a fascinating change because it's one
that social media companies never talk about, is that average
people have stopped sharing on social media, normal regular users.

(55:31):
The bets that may Facebook successful, right, which was like
you were going to connect with your friends and share like,
you know, pictures of you, like whatever, going to the
coffee shop or like having kind of a moment of
you know, going skiing. The average, like the percentage of
kind of overall content that comes from people who are

(55:52):
not professional producers has gone from like, you know, eighty
percent to like less than five percent over the past
ten year. It's something on which like it's really hard
to find data because social media companies do not publicize
that because again, part of their appeal is like, oh,
you may once in a while you are going to
see content from your friends, right, And that's kind of

(56:12):
that was the whole goal of social media platforms in
the beginning, right. But this is not what you see
on social media now. What you see is professionally produce
content by kind of influencers and creators who make a
living from their production.

Speaker 1 (56:28):
It's like cable television where you have lots of production companies,
lots of channels, and and they've gotten quite good. I mean,
the content is extraordinary. I wonder if they'll have to
change the name social media at some point because it's not.

Speaker 3 (56:43):
Really that that's exactly the point, right. So I think
that if you look at the ten years like to come, right,
you're going to see this decline of like just friends content,
organic you know, non monetized content is going to keep
on declining because you know, who to share a poorly
shot picture of themselves when you have the Kardashians kind

(57:06):
of next to you, or for that matter, like all
content producers who are doing quality, high quality videos and
posts rights, I mean, you should, just you don't really
want to do that, and it doesn't feel like the
space for it either. So that my question is like, well,
are people going to stay on social media or you know,
are the way in which they interact with social media

(57:28):
going to be the same if it's all professional content
producers ten years from now? And I guess I do
think that things are going to have to change when
we get closer to exactly what you are talking about,
which is more of a cable television system, except that
you have like, you know, five hundred million channels, you
know what I mean, not thirty.

Speaker 2 (57:50):
And you have the option and you have the option
to swipe right.

Speaker 3 (57:53):
But I do think that for platforms, once that's kind
of clarifying, and once people kind of understand that, you know,
we're not in the world that Facebook created circa like
two thousand and you know ten, we're already in the
world that's closer to cable television, where you get, like again,
this infinite kind of array of professional production on an

(58:14):
infinite array of topics. I do think that things will
get clearer, you know what I mean, Or that both
the economic incentives, the algorithmic infrastructure, and the expectations of
the audience are going to change, But I don't exactly
know in which directions.

Speaker 1 (58:32):
Yeah, we're really just in the adolescence of all these platforms.
This is so new, and I'm really interested in what's
happening with politics. I mean, it's clear that we're very
polarized now. I do want to say so you and
I are both very interesting in how algorithms play a
role in that. I do want to say that people
have always been polarized, and if you just look at

(58:52):
let's say last century twentieth century, the polarization was so
extreme that hundreds of millions of people were killed by
their neighbors, whether that's Nazi Germany or the Chinese and
Russian Communist revolutions, or Rwanda with the Tutsi against the Hutu,
or in Cambodi with Paul pot What. It doesn't mean
people always are very prone to get polarized. So I
don't think it can be blamed on social media. But

(59:15):
I am hopeful that if we sort of get out
of our hormonal adolescence with these platforms, we can at
least calm things down a little bit.

Speaker 2 (59:25):
What are your thoughts on that?

Speaker 3 (59:27):
I agree, I mean, I think polarization. You know, I
think for the past twenty years there has been and
I mean pre dating social media, there has been this worry,
right that polarization is increasing, and I agree with that. Historically,
it's a bit of a myopic view, right in the
sense that again, this very long history and as you

(59:47):
as you were saying, extremely deadly of kind of people
hating each other with such faults that it led to
actual civil war or kind of revolutions, et cetera.

Speaker 4 (59:58):
So that's kind of point number one.

Speaker 3 (01:00:01):
I agree also that social media is not the only culprits.

Speaker 4 (01:00:05):
There is no doubt about that.

Speaker 3 (01:00:07):
There are a number of other forces at stake, including
a big kind of macro economic and socio demographic transformations
in the United States among other places. But you know,
in many countries that everybody is just kind of grappling with,
right and with the role of technology in society. And

(01:00:28):
I also think that social media really doesn't help, right
in the sense of, like, again, it's not like it's
not the only culprit, it's not the only cause, but
it is enough to kind of tilt, you know, the
thing from a three, four to six, seven.

Speaker 4 (01:00:46):
You know, on the scale of zero to ten.

Speaker 3 (01:00:48):
And what I find is that it's not only polarization,
it's also this kind of like twenty four to seven
news reism, which you know, if you look at that
kind of the timeline of kind of news production. Again,
it's a longer history, right in a way it started
with cable television, right that we're broadcasting twenty four to seven,

(01:01:14):
you know, so it has a longer history. But certainly
the reactivity of social media the fact that you can
post a video just right when it's happening, the fact
that you can live stream. All of these development just
means that the ReSm of kind of news, kind of
circulation and reaction and reaction to the reaction has just
accelerated so much. And you see that it has definitely

(01:01:37):
changed politics as well. I mean, political life used to
be pretty subdued kind of business. Like the reasm, for example,
of Congress is a very slow ReSm, as is the
judicial ReSm. You know, the time for the Supreme Court
to make decisions a long time ago that was taking

(01:01:59):
like literally years, you know, and now we're just like
kind of in this you know, puff puff puff, kind
of every second there is a new kind of thing
that is even more shocking than the previous one.

Speaker 4 (01:02:10):
And that you know, you can't help but kind of
see that.

Speaker 3 (01:02:14):
It is correlated with the affordances of social media platforms. Right,
and so the question then is are we gonna are
we gonna move towards kind of a more stable kind
of reason. Well, as you said, it's like, I think
the parallel with adolescence is really interesting that right now
it's like everything's running really high, you know, what I mean,
and everything's gonna kind of came down a bit. And
I think that for that we need to turn to

(01:02:36):
kind of technology companies and just say, like, what are
the incentives that you've created for people to post, right,
and what are the guard rails that you're putting in place.
And again this is not to blame social media companies,
because it's really hard business. Like hundreds of millions billions
of people are interacting with this algorithmic structure, right, and

(01:03:01):
trying to kind of, you know, adjust things for that
scale of kind of people. It's just incredibly hard. And
so they are also on a learning curve, just like
the rest of us are. So they're trying in some
cases to do the right thing. In other cases, they're
also making decisions that are kind of irresponsible, right For example,

(01:03:27):
you know, getting rid of content moderation guidelines, I mean
you kind of know that that's going to encourage like
much more kind of heinous, hateful behavior. Again because that
not necessarily because creators believe in that heinous and hateful content,
but because that's what's going to capture the engagement and
attention of online audiences. And so I do think that

(01:03:49):
social media platforms at this point are at a bit
of a crossroad in terms of, like, so you have
become the key infrastructures for kind of media and communication,
what are you going to do. Are you going to
keep on chasing engagement at all costs right potentially kind

(01:04:11):
of destroying everything in your wake, or are you going
to start looking at the long term satisfaction? The long
term kind of benefits a long term outview of what
you can do for your customers. And I really think
it's kind of a short term, long term kind of
trade offs that's happening right now, and I really hope

(01:04:33):
that the long term considerations are going to gain traction.

Speaker 1 (01:04:37):
I love that, And I feel like social media companies
don't necessarily have the economic incentive to do that unless
one can stand out by saying we're going to take
less advertiser revenue for the purpose of making a better
community here.

Speaker 2 (01:04:54):
And I think they can really stand that that way.

Speaker 4 (01:04:55):
I agree.

Speaker 3 (01:04:56):
I mean, I think a couple of things that are
pretty obvious. I mean another one and again, and I mean,
I know that it's not necessarily a popular recommendation, but
we kind of got used to getting stuff for free.
Audiences Also, have to change, We all have to change,
and we kind of have to be willing to pay
for access to social media platforms because you see, if

(01:05:18):
we all had kind of subscriptions, well, that would kind
of reduce the part of advertising right in the kind
of overall revenues of social media companies, So they wouldn't
be chasing engagement metrics in quite the same way, so
they wouldn't be putting quite the same incentives in place
for content creators, and so everything would look a bit different.

Speaker 4 (01:05:40):
You know what I mean.

Speaker 3 (01:05:41):
So part of this is also due to the advertising
only or advertising mostly revenue structures of social media companies.
And I do think that all of us, as customers,
as audiences, as participants, we also have kind of a
role to play in saying, like, you know, we should
would be willing to pay for this service. And again,

(01:06:03):
it raises lots of questions about access democratization.

Speaker 4 (01:06:07):
You know, there are lots of it.

Speaker 3 (01:06:09):
It's not an easy debate to have, but I think
it's a debate for us having because a lot of
what's going wrong right now comes from this original I mean,
I don't want to call it an original sin, but
this original deal right that advertising was going to be
funding social media period. Yeah, and that is not the

(01:06:29):
healthiest way to kind of make a living as a company.

Speaker 1 (01:06:38):
That was my conversation with Angel Christen, and what we
talked about today was this notion of algorithmic capture, which
is the gradual reshaping of behavior in response to opaque
systems that reward certain choices and ignore others. If you
zoom out far enough, you can see that we are
living inside a huge experiment. Billions of brains, billions of

(01:07:04):
feedback loops, billions of tiny reinforcements happening every second. And
in this environment, some headline performs well, or some video
sparks outrage and multiplies. Maybe a creator feels the uplift
of attention. Maybe a newsroom watches a dashboard climb or
stall out, and gradually everybody's behavior adapts. So, as we

(01:07:28):
heard from Anzell, she studies journalists who entered the field
with a sense of civic mission, or content creators who
began with passion for a topic or a community. And
by the way, this is just as true with lots
of professions. Maybe a musician who enters wanting to create
something really new, and then the metrics arrive and suddenly

(01:07:50):
they're dealing with the gravitational pull of the algorithm. I
just want to zoom out for one second. As you'll
know if you've listened to a lot of these episodes.
Lea's pointing out that polarization predates social media by millions
of years, and obviously, any cursory reading of history will
tell you that outrage predates the Internet. Newspapers in the

(01:08:13):
nineteenth century, for example, they ran inflammatory headlines. This was
known as yellow journalism, and this all thrived long before
Wi Fi. Human attention has always been drawn to scandal
and spectacle and celebrity, but the scale of the quantification
of today's systems do introduce a new texture to public life.

(01:08:37):
Everything accelerates, the incentives intensify, and presumably the emotional temperature rises.
But I do want to say, as a techno optimist,
that nothing here is inevitable. Incentives can be redesigned, Revenue
models can and will evolve, Platforms can make different trade offs,

(01:08:57):
audiences can decide what there willing to pay for, and
creators can choose which signals they respond to. None of
this is simple, but it is important to realize that
we are in the earliest chapters of a story about
how human psychology interfaces with computational systems. We are in
the infancy phase of a technology that now functions as infrastructure.

(01:09:22):
So the next time you feel outrage when you're scrolling around,
or you see a headline that's engineered to hook you,
or you watch numbers tick upwards somewhere on a screen,
just pause for a second to ask yourself, what behavior
of mine is getting reinforced, what is being trained? How
do I want to participate in the system a little

(01:09:44):
more thoughtfully? Understanding the machinery just gives you a little
more agency inside of it, and that might be one
of the most important skills for living in this second quarter.

Speaker 2 (01:09:58):
Of the twenty first century.

Speaker 1 (01:10:03):
Go to Eagleman dot com slash podcast for more information
and to find further reading. Join the weekly discussions on
my substack, and check out and subscribe to Inner Cosmos
on YouTube for videos of each episode and to leave
comments until next time. I'm David Eagleman, and this is
Inner Cosmos.