Episode Transcript
Transcripts are displayed as originally observed. Some content, including advertisements may have changed.
Use Ctrl + F to search
0:00
First, the bad news. SAP
0:03
Business AI won't help you generate
0:05
cubist versions of your family's holiday
0:07
photos, but it will help
0:09
you understand which supplier is best to
0:12
help you roll out your plant-based packaging
0:14
in Southeast Asia, identify
0:16
the training your junior project manager needs
0:18
to rise up the ranks, and
0:21
automate repetitive tasks while you
0:23
focus on big innovations, so
0:26
you can be ready for the next opportunity. Revolutionary
0:29
technology, real-world results.
0:32
That's SAP Business AI. First,
0:37
the bad news. SAP
0:39
Business AI won't help you generate
0:41
cubist versions of your family's holiday
0:44
photos, but it will help
0:46
you understand which supplier is best to
0:48
help you roll out your plant-based packaging
0:50
in Southeast Asia, identify
0:52
the training your junior project manager needs
0:54
to rise up the ranks, and
0:57
automate repetitive tasks while you
0:59
focus on big innovations, so
1:02
you can be ready for the next opportunity. Revolutionary
1:05
technology, real-world results.
1:08
That's SAP Business AI. On
1:18
the morning that I spoke to Kyle Chaco, I
1:20
dropped my son off at preschool and stopped to get
1:22
some coffee on the way to work. And
1:25
I wanted Kyle to try to guess
1:27
what the coffee shop that I went
1:29
to looked like. I
1:32
wondered if that was something he could do. I
1:35
think so. I mean, I'm going to
1:37
use my mental magician powers and say
1:39
that there were some hanging
1:41
pendant lamps that may have had Edison
1:43
bulbs in them. There might
1:46
have been some white subway tiles on
1:48
the wall behind the counter,
1:50
which might be like a nice marble
1:52
or even concrete. It's really cool.
1:55
Some nice reclaimed wood furniture or
1:58
some vintage bits and pieces. century
2:00
pieces that make you feel like you're in a
2:02
nice hotel from 1955. And
2:06
then probably some nice ceramic cups that
2:08
your cappuccino comes in. It
2:10
was a latte to go, but the rest
2:12
of this description is pretty spot on. Because
2:16
Kyle has thought a lot about
2:18
what particular signifiers a coffee shop
2:20
today might possess. He's
2:23
a staff writer at The New Yorker
2:25
and has just written the book Filter
2:27
World, How Algorithms Flattened Culture. In
2:30
one part of the book, Kyle talks about
2:32
how whenever he was traveling, he would type
2:34
Hipster Coffee Shop into Yelp and
2:37
Google Maps. Hipster,
2:39
in the 20s at least,
2:41
was the signifier of millennial
2:43
era consumers and tastes. And
2:47
the coffee shops all had to project
2:49
their image online through these apps like
2:52
Yelp or Google Maps. So
2:54
it was a kind of index. I
2:56
could use the phrase through the search
2:59
algorithm, through Yelp's rankings and star ratings
3:01
to find exactly what I was looking
3:03
for. When
3:09
you wrote about this phenomenon in
3:12
2016, you called it airspace. What
3:15
does that mean? At the time,
3:17
I used that word airspace
3:19
to talk about the style that was happening,
3:21
like the generic minimalist,
3:23
white subway tile, mid-century
3:25
furniture thing. And
3:28
I was seeing that not just in
3:30
coffee shops, but in co-working spaces and
3:32
restaurants and hotels and kind of like
3:35
the whole new geography that was popping
3:37
up with the sharing economy. When
3:40
readers saw the term airspace, they
3:42
knew exactly what Kyle meant and
3:45
started emailing in pictures of coffee
3:47
shops and hotel lobbies that hewed
3:49
to this same shareable aesthetic. It
3:52
was a geography and a physical space
3:55
that had a lot in common with the
3:57
digital spaces that we're in. like
4:00
interconnected and driven by the internet.
4:05
Was there a point in which you
4:07
realized, oh, this is more than
4:10
this design
4:12
phenomenon of coffee shops and
4:14
WeWorks and what have you,
4:17
that you realized I think
4:19
this is rippling out into something bigger? Yeah,
4:22
like the, to me the coffee shop aesthetic
4:24
was kind of the canary in the
4:26
coal mine. Like it was the most
4:28
visible symbol of this
4:30
homogenization that was happening over the
4:32
internet. So there was that
4:34
design, right? Like it was a cliche, stereotypical
4:37
design that I think a lot of people
4:39
kind of observed. And it became a cliche,
4:41
it was boring. You can kind of see
4:43
it everywhere. But I think the
4:45
like reality of it, the source of it
4:47
ran much deeper. It
4:50
ran, Kyle says, to a handful
4:52
of tech companies creating a larger handful
4:54
of algorithms that drove our behavior. Whether
4:56
we knew it or not. It
4:59
wasn't just about one style, but
5:01
this deeper homogenization and sameness
5:03
that was creeping into a lot of
5:06
different spaces driven by the Instagram
5:08
feed and now the
5:11
TikTok feed and Netflix streaming and Spotify
5:13
recommendations. It was kind of this aesthetic
5:16
that spread everywhere via the feeds.
5:22
Today on the show, are we living
5:24
in filter world? A place where
5:27
algorithms, not people, decide what
5:29
we like. I'm
5:31
Lizzie O'Leary, and you're listening to What Next
5:33
TBD, a show about technology,
5:35
power, and how the future will be determined.
5:38
Sticker it up. Reboot
5:47
your credit card with Apple Card. It
5:49
gives you unlimited daily cash back that
5:51
can earn 4.35% annual percentage yield when
5:55
you open a savings account. A
5:58
high yield, low effort way to grow your... money
6:00
with no fees. Apply for
6:02
Apple cards now in the Wallet app
6:04
on iPhone to start earning and growing
6:06
your daily cash with savings today. Apple
6:10
cards subject to credit approval. Savings
6:12
is available to Apple card owners
6:14
subject to eligibility. Savings
6:17
accounts by Goldman Sachs Bank
6:19
USA. Member FDIC term supply.
6:24
First, the bad news. SAP
6:26
Business AI won't help you generate
6:28
cubist versions of your family's birthday
6:30
photos, but it will
6:33
help you understand which supplier is
6:35
best to help you roll out your
6:37
plant-based packaging in Southeast Asia. Identify
6:40
the training your junior project manager needs
6:42
to rise up the ranks and
6:44
automate repetitive tasks while you
6:46
focus on big innovations so
6:49
you can be ready for the next opportunity. Revolutionary
6:52
technology real-world results. That's
6:55
SAP Business AI. Kyle
7:02
describes filter world as the
7:04
vast interlocking and yet diffuse
7:06
network of algorithms that influence
7:09
our lives today. And
7:11
while you may have thought of that in
7:13
terms of politics, how Facebook
7:15
tends to surface outrageous selection
7:17
content, for example, Kyle's
7:19
book focuses on culture and
7:22
how he says the world
7:24
of algorithmic recommendations favors
7:26
a certain kind of culture. Things
7:29
that are easier, less challenging
7:31
to digest, made for
7:33
sharing and recommending over and over
7:36
and over. I
7:39
want you to try to describe
7:41
filter world as a place to
7:43
me. What are its
7:46
defining characteristics? Like if you're an
7:48
inhabitant of filter world, what is
7:50
your experience? It's almost
7:52
like being a rat, an image. You
7:55
are surrounded or we are surrounded as
7:58
internet users by all of these. these
8:00
different algorithmic systems and
8:02
we're constantly being guided in
8:04
one direction or another by
8:06
recommendations. So Spotify is pushing a
8:08
song at you or Netflix is
8:10
pushing a new TV show. So
8:13
I think as an inhabitant of filter worlds,
8:15
we're just immersed in
8:17
this bubble of our own desires
8:19
or a projection of our own
8:21
desires that come to us through
8:23
digital platforms. I use
8:26
the word filter because these
8:28
feeds and recommendations are a filter,
8:31
like they sort content for you. They
8:33
try to guide you towards something that
8:36
fits with your taste. But I think
8:38
that becomes an inescapable
8:41
environment. It's a
8:43
scrim or a filter that you can
8:45
actually see through, that blocks you from
8:47
experiencing other things. One
8:49
of the parts of the book that I
8:51
found really interesting was this
8:54
moment where you describe a shift
8:56
from chronological feeds, sort
8:58
of the infancy of social
9:01
media to algorithmic
9:03
recommendations. When do you
9:05
start to pinpoint that shift? Yeah, I
9:07
think there really was a shift over the course of
9:09
the 2010s. When
9:12
we all got on social media, when social
9:14
networks were beginning to be popular and say
9:16
like the late 2000s and early 2010s, most
9:21
of the feeds there were linear, meaning
9:24
that they were chronologically ordered. You saw
9:26
the posts that were posted most recently,
9:28
and then you saw the older ones,
9:30
and that's just how all content was
9:32
treated. Then there's
9:34
this watershed moment around 2015, 2016
9:39
where tech companies decided to make
9:41
the feeds more algorithmic, to mix
9:43
in a greater proportion of recommendations
9:45
versus things you chose to follow.
9:49
That happened with the Facebook feed,
9:51
it happened with Instagram recommendations, it
9:53
happened with the Netflix homepage. They
9:57
all move toward sorting
9:59
content. for you rather than letting you
10:01
choose exactly what you were going to see.
10:04
That has been covered fairly relentlessly in the
10:06
media talking about the 2016 election and
10:09
Facebook showing, you know, more inflammatory content,
10:12
for example, things things that are
10:15
more likely to provoke an emotional response, getting
10:17
boosted by the algorithm, etc. But you
10:20
focus so much on culture
10:22
and taste. That is the heart of
10:24
this book. And there is a
10:27
quote in which you say, the
10:29
overall digital environment is dictated by
10:31
tech companies with ruthlessly capitalist expansionary
10:33
motives which do not provide the
10:36
most fertile ground for culture. Sounds
10:40
bad. It does sound bad. But
10:44
I want examples of that. I want to
10:46
understand where you
10:49
see Filter World sanding
10:51
off the difficult
10:53
edges of art or movies
10:55
or music. Yeah, I
10:58
mean, I think one of
11:00
the issues of Filter World is just
11:02
that so much of our consumption of
11:04
culture has become mediated by
11:06
these feeds. Like whenever we want to
11:08
listen to a song, we're on Spotify
11:10
and getting kind of redirected by the
11:13
platform. Whenever we want to watch
11:16
a TV show or film, we are
11:18
like not just choosing what to see,
11:20
but experiencing the recommendations of Netflix. So
11:22
I think our consumption is being guided
11:24
in manipulative ways. For
11:27
one example, Netflix changes
11:30
the thumbnails of the shows it presents
11:32
to you based on what it sees
11:34
as your personal taste. So not
11:36
only is the selection of content on their
11:38
change, but actually how it's presented to you.
11:41
And so if you're a person who,
11:44
say, only watches rom-coms, but
11:46
Netflix wants to push you toward a
11:48
sports documentary or something, it might show
11:50
you an image that leads you to
11:52
think a sports documentary is a rom-com.
11:54
And that kind
11:57
of fools you into not thinking about your own
11:59
taste. and into consuming something that
12:02
doesn't necessarily appeal to you. On
12:04
the creator side, because all
12:07
of our attention is so mediated by these feeds,
12:09
creators like artists, musicians,
12:12
designers, whoever have to
12:14
tailor their work in a
12:16
way that works for the feed. I
12:19
mean, you see that on TikTok
12:21
where music producers focus on making
12:23
sound bites that are like 10
12:25
seconds long, but are so densely
12:27
packed with sound and drama that
12:29
you just instantly
12:32
stick in your head and they're repeatable
12:34
infinitely. I mean, in one case
12:37
in the book, I spoke with
12:39
an illustrator friend of mine, Hallie
12:41
Bateman, who experienced this algorithmic
12:43
pressure on Instagram. When
12:45
she started doing a series of
12:48
drawings of just little self-help
12:50
commandments or suggestions on brightly
12:52
colored construction paper, and
12:55
immediately that work blew up on Instagram.
12:57
That became her most successful project ever.
12:59
She was getting tons of followers. But
13:02
then she also experienced this pressure to
13:04
keep doing that, to serve the audience
13:06
that was only coming to her for
13:08
that thing. For her, it
13:11
felt like that was just
13:13
a distortion of her own creative practice. That's
13:15
not what she wanted to be known for.
13:17
That's not what she wanted to chase in
13:19
a way to keep her audience's attention. I
13:22
guess that makes sense if you
13:24
are thinking about someone who is
13:26
gaining fame, or gaining followers,
13:28
or gaining money from existing
13:31
on the Internet. But there's a
13:33
section of the book where you seem
13:35
to pause it, and I want to
13:37
dig into this, that it goes deeper
13:40
than that. That the popularity of Sheila
13:42
Hetty or Nausgarde, or
13:45
this particular kind of auto fiction
13:47
is what? Is
13:51
the rise of these books because
13:53
we are used
13:56
to viewing the world in
13:58
an Instagram protagonist way? I
14:01
think Instagram and
14:03
other feeds have really
14:06
shaped how we see everything in the world.
14:08
All forms of culture has to kind of
14:10
flow through them. And in
14:13
the case of auto-fiction, my
14:15
book was kind of responding to
14:17
this academic Mark McGurrell's work
14:20
on Amazon, the algorithmic marketplace of
14:22
books that exists on there. And
14:25
he argued that Amazon and
14:27
algorithmic recommendations kind of slot
14:30
authors into very specific genres
14:32
of stuff, whether that's like
14:35
romance or fantasy
14:37
or something like auto-fiction, which has kind
14:39
of become a dominant literary
14:42
mode. And
14:44
to me, the way that auto-fiction resonates
14:46
with Instagram is that we
14:48
want the author to be a
14:50
kind of influencer. We want the author, we want
14:53
to know about the author's life. We
14:55
want to know what kinds of coffee shops they're
14:57
going to. We want to know what's in their
14:59
home. We want to kind of consume
15:01
their lifestyle in a way as
15:03
much as we consume their
15:06
work of literature itself. So I
15:09
think there's a resonance with how
15:11
auto-fiction plays with the presentation of
15:13
the self and authenticity in
15:15
the same way that an influencer on Instagram
15:17
kind of presents some of their self and
15:20
shows off some of their real identity.
15:22
But then it's also kind of inauthentic
15:25
and artificial as well. I
15:27
talked to some of my colleagues about this and sort of
15:30
said, do you see experiences of
15:32
flattened culture? And one question that came
15:34
back to them was like, but isn't
15:37
it kind of fun and cool to learn
15:39
about cottagecore, gremlin's
15:42
core, dark academia?
15:45
Even if these are like weird flash in the
15:47
pan things that circulate via social media, they
15:52
seem organic. No, not
15:55
prescribed by Anna Wintour on high.
15:58
Yeah, I think there's this bottom. generation
16:00
of culture that is really interesting online.
16:04
I think these digital platforms highlight a
16:06
ton of interesting stuff and they allow
16:08
us to experience the breasts of what
16:11
everyone makes. There are fewer
16:13
gatekeepers. And I love
16:15
exploring that field like Grandma or Coastal
16:19
Grandmother aesthetic. Whatever the one
16:21
is that we're talking about this week or
16:23
month, it's always fun to know. And
16:26
I think what becomes overwhelming is the
16:29
specificity and speed that these
16:32
things emerge and pass by with.
16:35
All of a sudden everyone is talking about Mobwife
16:38
core or something. That's the one I've
16:40
been seeing this week. TikTok has spoken.
16:42
The clean girl aesthetic is out and the
16:44
Mobwife aesthetic is in. I am
16:47
so excited. It's time for the fur coats. It's
16:49
time for the makeup girlies. It's time for series,
16:51
all black leather outfits. I
16:53
want dramatic. I want maximalism. So let's talk
16:55
about how to dress like this. And it's
16:58
just like the trends percolate
17:01
up and then pass by so
17:03
quickly that you almost don't have a
17:05
chance to actually engage with what the content is
17:07
there. To me
17:09
it's almost...it creates this
17:11
atmosphere in which there's more pressure to
17:14
keep up with the thing that's new
17:16
and the thing that's going viral this
17:18
second. Then there is to really
17:20
go deep in a single one of these areas.
17:23
So I worry that it's almost like too much flavor of
17:25
the week. And we're
17:27
not allowing ourselves to think
17:30
slower thoughts and develop a
17:32
different sense of identity. I mean,
17:34
I contrast the
17:37
flash in the pan, coastal grandmother, etc.,
17:41
Cottagecore, Tran, TikTok with maybe
17:43
a more long-term cultural identity
17:45
like punks or
17:48
goths. Or even
17:50
the hipster in a way. It was like, produce
17:52
a more meaningful body of culture than
17:54
just this instant adaptation to an aesthetic
17:56
that then passes by when it no
17:59
longer gets engaged. management. When
18:04
we come back, the anxiety that
18:06
comes from not knowing what the
18:08
algorithms are going to do with us. This
18:19
episode is brought to you by S.C.O.C. You
18:21
need to keep operations humming around the
18:23
clock, but potential disruptions are everywhere.
18:27
You can find and fix problems and find and fix
18:30
issues fast so you can reduce risk and
18:32
ditch the time. The world's largest
18:34
enterprises rely on S.C.O.C.E.'s unified
18:36
security and observability platform to
18:39
become more efficient, resilient, and
18:41
innovative. With S.C.O.C.E., you can
18:43
react quickly, evolve faster, and be ready
18:45
for anything. Stay ahead
18:47
of disruptions. Learn more
18:50
at S.C.O.C.E..com/resilience. First,
18:56
the bad news. SAP Business
18:58
AI won't help you generate cubist
19:01
versions of your family's holiday photos.
19:04
It will help you understand which supplier
19:06
is best to help you roll out
19:08
your plant-based packaging in Southeast Asia, identify
19:11
the training your junior project manager needs
19:13
to rise up the ranks, and
19:16
automate repetitive tasks while you
19:18
focus on big innovations so
19:21
you can be ready for the next opportunity. Revolutionary
19:24
Technology Real World Results.
19:27
That's SAP Business AI. This
19:32
podcast is brought to you by Progressive Insurance.
19:35
Hey listeners, whether you love true crime
19:37
or comedies, celebrity interviews, news, or even
19:39
motivational speakers, you call the shots
19:41
on what's in your podcast queue, right? And
19:44
guess what? Now you can call the shots
19:46
on your auto insurance too. Enter the
19:48
name your price tool from Progressive. The
19:50
name your price tool puts you in charge of your auto
19:52
insurance by working just the way
19:55
it sounds. You tell Progressive how much
19:57
you want to pay for car insurance, then they'll show you
19:59
a variety of things. of coverages that fit within
20:01
your budget, giving you options. Now
20:03
that's something you'll want to press play on. It's
20:06
easy to start a quote, and you'll be able to choose the
20:08
best option for you, fast. It's
20:10
just one of the many ways you can save with
20:12
Progressive Insurance. Quote today at
20:14
progressive.com to try the Name Your Price
20:17
tool for yourself, and join the over
20:19
29 million drivers who trust Progressive. Progressive
20:22
Casualty Insurance Company and Affiliates, price and
20:24
coverage match limited by state law. In
20:27
one of the most fascinating sections of the book, Kyle
20:30
writes about something called algorithmic
20:32
anxiety. Because by now, we
20:34
all know that algorithms are affecting our lives,
20:37
but we don't always know exactly how. The
20:41
term itself was coined by an
20:43
academic to describe the way Airbnb
20:45
hosts tried to alter their property
20:47
listings to appeal to the company's
20:49
search algorithm, changing the
20:52
photos of their listings or describing them
20:54
with slightly different keywords so they would
20:56
be shown to more potential renters.
20:59
And these things didn't actually work. Like
21:01
they were just strategies of
21:03
coping with the unknowability of
21:06
that digital platform. And
21:08
I think we all kind of experienced that unknowability.
21:12
Like we don't know what the TikTok feed
21:14
is thinking of us. We don't know exactly
21:16
how the Netflix homepage works. And
21:18
so we're constantly negotiating with how
21:21
our behaviors are being perceived by
21:23
the technology, what signals we're
21:25
sending or not sending. And
21:28
I think we also feel anxiety because we
21:30
can't talk back to these systems. Like they
21:32
exert so much power over us and
21:35
we exist through them so often, but
21:37
we can't change how the X
21:39
algorithm works. We can't tell Netflix
21:42
to stop recommending one thing or
21:44
another. Right, I have no
21:46
idea how Spotify is gonna put this podcast
21:48
up today, right? It could
21:50
display it prominently or not. Right, right,
21:53
and those are technological systems that we
21:55
just don't have any agency over. And almost,
21:57
I mean, no human does in a way.
22:00
They are quite black box
22:02
systems that are operating at a
22:04
scale that is beyond any one person.
22:07
There maybe is not a person choosing to
22:10
highlight one Spotify song or another. It's just
22:12
what's popular and what gets automatically recommended.
22:15
But I think what I struggle with is the
22:18
human response to that, right? Like you talk about
22:20
this woman who bought leg
22:22
warmers because she saw a bunch
22:24
of influencers by leg
22:26
warmers. And
22:29
it seemed to really inspire like
22:31
a crisis of confidence in her that she sort
22:33
of felt like she didn't know who she was
22:35
and she couldn't figure out her
22:38
own style or identity. And
22:40
maybe this is because I'm 48, but I read that
22:42
section of the book and thought, why are you
22:44
letting this do this to you? Just
22:48
wear what you want to wear. Who cares what
22:50
the internet thinks? Yeah,
22:52
I mean, it sounds so simple. Like
22:54
to just say, oh, we'll think about
22:56
what you like, kind of separate yourself
22:58
from the feeds. But
23:01
I mean, as a
23:03
millennial and talking to a lot of
23:05
younger people, like Gen
23:07
Z people on TikTok,
23:09
particularly, their identity is almost
23:12
more manipulated by the
23:14
feeds online. Like they have experienced
23:16
this for more of their lives.
23:19
Their tastes or preferences were already
23:21
developed within these systems. And
23:24
so I think there's a sense in which they haven't learned
23:26
to distrust them, or for a long time it's
23:28
worked and shown them what they did like, or
23:31
they believed it showed them what they liked, unless
23:33
they trusted the feed more than they
23:36
might trust a human tastemaker. They
23:38
haven't had those kinds of conversations or
23:41
developed like almost the skill of thinking
23:43
about what you as a person like,
23:45
as opposed to what's popular. I
23:48
mean, to me, that feels extraordinarily naive.
23:52
I mean, tech companies
23:54
have occupied so much of our mental
23:56
landscape in a way. Like the more
23:58
online you are, the more you're... influenced
24:00
by these systems. And
24:02
so I think it is easy to forget
24:04
that there's a world outside of that, to
24:06
forget that you can figure
24:08
out what you like, not on the
24:11
TikTok feed, not just chase what influencers
24:13
are doing or what Twitter pushes at
24:15
you. So I said same
24:17
woman Valerie Peter, she
24:20
was on Twitter or on X and
24:22
was looking at astrology content and
24:24
had been interested in it for a while. And
24:28
all of a sudden she got bombarded
24:30
with all of this astrology that she
24:32
then started to dislike and felt like
24:34
kind of misled or anxious
24:37
about getting. But she couldn't
24:39
tell Twitter to stop giving her this
24:41
content. She couldn't send a strong enough
24:44
signal to stay away from the astrology.
24:46
And I feel like it shows in
24:48
a way how the feeds direct you
24:51
in ways that you don't necessarily like.
24:54
You bring me to the idea of pushing back and I
24:56
feel like there are sort of two
24:58
buckets there. One is on
25:00
the individual level and one is on the
25:02
societal or regulatory level. The
25:05
final chapter in this book is about your algorithmic
25:08
cleanse or algorithm cleanse. Tell
25:11
me about that. I
25:14
wanted to do a kind of algorithm
25:16
cleanse, like almost like a diet
25:18
or like giving up something
25:21
for a length or a dry January,
25:23
like just getting outside of the feeds.
25:25
In part because I just become so
25:27
overwhelmed with them. Like I am as
25:30
online as anyone. I was on X
25:33
for work. I was looking at Instagram all the time.
25:36
And so I kind of decided that the only way
25:38
I could truly see who I was without these things
25:40
was to just get off them for a while. And
25:43
so I deleted everything from my phone, logged
25:46
off all the apps. I
25:48
was kind of confronted with this sudden
25:50
silence of the absence of
25:52
the feeds and the absence of these thousand
25:54
posts a day that I was
25:57
used to consuming. But I think
25:59
like cutting that stuff off. off and going cold
26:01
turkey, did reframe how
26:03
I think about digital platforms. It
26:06
reminded me in a way that you don't
26:08
need them. That there's a world outside of
26:11
what gets recommended to you. I
26:13
think it lessens that algorithmic anxiety
26:16
simply because I didn't have to
26:18
contend with this image
26:20
of myself that existed online, that
26:22
I was constantly being manipulated by.
26:24
Yeah. I have to admit that there's a part of
26:26
me that was like, dude, just log off. Easier
26:30
said than done sometimes. Well,
26:32
that brings me to the macro stuff. I
26:34
mean, and this is
26:36
where I think you
26:39
get to some of the most interesting parts
26:41
of the book, is that perhaps culture
26:44
and society is
26:46
being manipulated, even if you have
26:49
logged off, even if you
26:52
are not a part of this on
26:54
a micro level, even if you're Marty
26:57
Scorsese, you feel the influence
27:00
of these companies. It
27:03
makes me wonder if this
27:05
is a question of algorithmic
27:07
influence or monopolistic influence. If
27:09
there were 500 companies
27:14
driving social
27:16
media feeds, would it be different? I
27:20
think it would be. I mean, as users,
27:22
we feel that we don't have that much
27:25
choice of what we experience. Meta
27:28
as a company owns Facebook and
27:30
Instagram and WhatsApp, which are three
27:32
of the biggest platforms in
27:34
the world. Even if
27:36
you are not on those platforms, the
27:38
actions and decisions of billions of people
27:40
around the world are being guided by
27:42
them. That impacts what
27:44
music gets made, it impacts what
27:46
films get produced. The
27:49
response of the Martin Scorsese concern, influences
27:53
how restaurants advertise themselves, and
27:55
what kinds of dishes
27:57
they put on their menus. Because even if you don't
27:59
care, about how something works on TikTok,
28:02
a lot of other people will. So
28:05
I think we are all kind
28:07
of living in this world shaped
28:09
by algorithmic recommendations. And I think
28:12
increased competition, if we can break
28:14
down the monopolies of
28:16
meta and Google, would
28:18
actually lead to a wider diversity of
28:21
experiences. We could change
28:24
how algorithmic feeds work perhaps, or
28:26
we could drift to
28:28
different flavors of platforms that might better
28:30
serve our personal preferences. At
28:32
the end of writing this book and talking about
28:34
it, it just strikes me so much that
28:37
the idea that a billion people should
28:40
be on one digital platform and all
28:42
being pushed through the same pipes in
28:44
a way is deeply disturbing. That's
28:47
the biggest homogenizing force
28:50
of us that we've ever seen
28:52
in human experience. It's
28:55
kind of gross to me at this
28:57
point. LISA K. H United
29:21
States, we've had a really hard time
29:24
passing any regulation on social media, just
29:26
because there isn't fully the political
29:29
willpower and the platforms themselves have
29:31
become so dominant. They spend so
29:34
much money on lobbying. But if
29:36
you look outside of the US
29:38
at European Union regulation, they
29:41
are actively putting laws into place
29:43
that give users rights to their
29:45
data, for example, or give you
29:47
the ability to opt out of
29:49
surveillance, or even with
29:51
the more recent Digital Services Act
29:53
in the EU, it mandates
29:55
that you can opt out of algorithmic
29:58
feeds. Like if you don't have a chance to do that, you can't just do that. don't
30:00
want to experience recommendations, you will have the
30:02
option to not do that. And
30:05
that's almost unthinkable in the United
30:07
States right now. You
30:10
see the platforms kind of responding to
30:12
those regulations and changing how they work
30:14
for everyone because they use. It's easier
30:16
for them to just make one big change. Yes,
30:19
yeah, yeah. But I think actual
30:22
regulation in the United States where these companies
30:24
exist could accomplish a lot more and just
30:26
give us more options for how we live
30:28
online. How should one
30:30
arrive at the idea of taste and culture?
30:32
Because you talk a lot
30:35
about the local DJ or the art critic or
30:39
whatever. But I think about
30:42
going to the movies in the early 90s,
30:45
and my family had this joke about our
30:47
movie theater as the Bethesda monoculture, that
30:49
it catered to
30:52
upper middle class white
30:54
people. And those were
30:56
the tastes that spread
31:00
society wide. And
31:02
so the democratizing
31:04
aspects of the internet are
31:07
incredible. I
31:09
worry that if we
31:11
go the other way, it's racist,
31:14
it's classist, it's sexist, that
31:17
you're subject to the taste of just a handful of people.
31:20
Yeah, there were these gatekeepers in
31:22
the past, like these human gatekeepers.
31:25
And there still are. Yeah. And
31:28
now I think we have this
31:30
algorithmic gatekeeping mechanism. But we do
31:32
have the total democratization of publishing
31:35
and of allowing anyone
31:37
to put something online and access an
31:39
audience. And I think that's amazing
31:41
and really a positive force
31:43
in culture. What I
31:46
don't like is the total
31:48
domination of algorithmic gatekeeping in
31:50
a way. Anyone can
31:52
put something online. But that
31:54
doesn't guarantee that you're going to get an
31:56
audience for it. You don't have a right
31:59
to other people. people's attention. And
32:02
algorithmic feeds push people's attention toward
32:04
a very specific set of content
32:06
that works for the platforms that sparks
32:08
the right kinds of reaction on people.
32:11
So I think, you know, we gain
32:13
something in the system, in the ecosystem of
32:16
the internet that anyone can put something out
32:18
there, but we also lose it with
32:21
just the sheer extent that
32:23
our attention is directed by algorithmic feeds.
32:25
We're not just seeing everything that everyone
32:28
puts out there. And from
32:31
the perspective of creators, like
32:34
the independent kid who might
32:36
upload music to Spotify, Spotify
32:38
literally put into place
32:40
a new rule that they are
32:42
not going to monetarily reward any
32:44
song under a thousand streams. So
32:47
that essentially argues that
32:50
if you can't reach over a
32:52
thousand streams, if you can't game
32:54
the system and get an audience,
32:56
then your art is literally worthless,
32:58
that you shouldn't get rewarded for it.
33:00
And that means you can't make it in
33:03
the future. Like I think this
33:05
algorithmic ecosystem is unsanable for
33:07
the production of culture. That
33:11
makes me wonder how much
33:13
of this might be transient,
33:16
either because as a consumer, you
33:19
age out, right? You
33:22
reach a point and you just
33:24
think like, this is who I am, and I'm okay
33:26
with that. Or because if
33:28
you're a younger person, and you've
33:31
been brought up with this, you're pretty
33:33
savvy. Like you might start to question
33:36
what the algorithm is doing. Did
33:38
you find any points in
33:41
your reporting process where you thought like, maybe
33:44
I'm capturing a moment in time, but it's not always
33:46
going to be this way? Yeah,
33:49
yeah. I mean, the internet changes so
33:51
much that one fear I had
33:53
in writing the book was just that I would be
33:56
capturing something that wasn't relevant anymore that
33:58
had like disappeared and I
34:00
added into the ether. And I
34:02
think change has happened a bit slower
34:04
than I expected. I
34:06
think the way that I framed
34:09
the book or the way that I think about it
34:11
now is that it's about the internet of the 20
34:13
times. It's about how a
34:16
certain ecosystem evolved, how
34:19
we responded as consumers to it,
34:21
how digital platforms developed to
34:23
be the way they are. And
34:26
now I kind of see the on-wee and
34:28
the anxiety growing to the point that
34:30
people are rebelling against these systems and
34:33
looking really intentionally for alternatives.
34:36
I don't think the alternatives exist yet
34:39
necessarily, like in part due
34:41
to the lack of monopoly regulation and
34:43
things like that. But there's
34:45
a desire for a new internet. There's
34:47
like a desire to not
34:50
exist in this filter world and to
34:52
move beyond it. And I think by
34:54
the end of the book, when I finished writing
34:57
a year or so ago, I
34:59
started to watch Twitter fall apart
35:02
under the pond ask. I started
35:04
to see the Facebook scrambling
35:06
and trying to do the metaverse. That
35:09
didn't work either. So
35:12
I think by the tail ends, I
35:14
got the sense that we're moving out of the phase
35:16
of the internet that happens. But
35:18
that phase dictated so much of what culture
35:20
was popular in the past eight to 10
35:23
years. Does that leave you hopeful? Yes,
35:26
it does. I mean, the
35:29
internet's always changing. Like there's always
35:31
new things popping up and new
35:33
mechanisms and experiments to try. And
35:36
like human creativity, there's no depth of
35:39
culture here. There's no end to the
35:41
human desire to make art and to
35:43
make original stuff. I
35:46
just think the digital ecosystem
35:48
could support that more. And
35:50
what I'm hopeful for is like different kinds
35:53
of platforms that are more sustainable, both
35:55
for our attention and for the careers
35:57
of the artists and creators. Kal
36:05
Chayka, thank you so much for coming on and
36:07
spending this time with me. Thanks for
36:09
having me. Kal
36:12
Chayka is a staff writer at
36:15
The New Yorker and the author
36:17
of Filter World, How Algorithms Flattened
36:19
Culture. And that is
36:21
it for our show today. What Next TBD
36:23
is produced by Evan Campbell, Anna Phillips, and
36:25
Patrick Farr. Our show is
36:28
edited by Mia Armstrong Lopez. Alicia
36:30
Montgomery is vice president of audio for
36:33
Slate. And TBD is part of the
36:35
larger What Next family. We're
36:37
also part of Future Tense, a partnership
36:39
of Slate, Arizona State University, and New
36:41
America. And if you are
36:43
a fan of this show, I have a little
36:45
request for you. Become a Slate
36:48
Plus member. Just head on over to
36:50
slate.com/what next plus to sign up.
36:53
All right, we will be back next week
36:55
with more episodes. I'm Lizzie O'Leary. Thanks for
36:57
listening.
Podchaser is the ultimate destination for podcast data, search, and discovery. Learn More