0:00

Surge Engine is brought to you by PolicyGenius.

0:03

Life insurance is an important safety net for your

0:05

family, but trying to find the right policy on

0:07

your own can be time consuming and overwhelming. PolicyGenius

0:10

is the country's leading online

0:12

insurance marketplace. It saves you

0:15

time and money so you can provide your family

0:17

with a financial safety net starting today. With

0:19

PolicyGenius, you can find life insurance policies that start

0:21

at just $292 per year for $1 million of

0:24

coverage. Some

0:27

options offer same day approval and

0:29

avoid unnecessary medical exams. PolicyGenius

0:32

helps you compare your options from top

0:34

companies. And their team of licensed experts

0:36

is on hand to help talk you through it. PolicyGenius

0:39

gives you unbiased advice from a licensed

0:41

expert support team. They have no

0:44

incentive to recommend one insurer over another so

0:46

you can trust their guidance. Check

0:48

life insurance off your to-do list in no

0:50

time with PolicyGenius. Head to policygenius.com or click

0:52

the link in the description to get your

0:55

free life insurance quotes and see how much

0:57

you could save. That's

0:59

policygenius.com. This

1:03

is Search Engine.

1:25

I'm PJ Vogt. No question too big,

1:27

no question too small. Each

1:29

week, the small staff of our show meets

1:31

in a sunny office in one of the

1:33

tall buildings in New York City's least charming

1:35

neighborhood. And we try to decide what

1:38

we should pay attention to. It

1:40

could be anything, which is sort of tricky,

1:42

actually. Often,

1:44

we settle for trying to understand

1:46

and explain very recent history, stories

1:49

that have unfolded in the past few years,

1:51

which with the benefit of hindsight, we can

1:54

now understand more clearly the rise of

1:56

fentanyl, the fall of Sam Bankman Freed. There's

1:59

one story. though we keep bumping into

2:01

this year a story that were in

2:03

the beginning or maybe the middle of

2:06

which I've I'm is up to curious

2:08

about to resist trying to understand as

2:10

it unfolds. A

2:12

couple months ago in March we spoke to journalist

2:14

and sell a podcast her as recline. The.

2:16

Question we posed to him was. How.

2:19

Do we survive the media apocalypse?

2:23

At the time all these online news

2:25

outlets for dying by speed news had

2:27

been killed. Traditional newsrooms like The Washington

2:30

Post and The Times were setting staff

2:32

through lay offs and buyouts and as

2:34

a person who gloves reading human or

2:36

it and sacked such sentences on the

2:38

internet to depends on this and and

2:41

says. I felt alarmed. I

2:44

wanted to understand this moment and I want to

2:46

hear ideas from smart people about how to prevent

2:48

it. Has or it had insides

2:50

yet. Suggestions for how readers can push back

2:52

if you haven't Listened to please. Second episode

2:54

out. But since then

2:57

or apocalyptic moment it is just kept

2:59

rolling on. The. Scene from

3:01

this apocalypse are so bizarre and

3:03

spectacular a summons can feel myself

3:05

dissociating like was watching this video

3:07

last week. In

3:11

California on a psychedelic stage, a you

3:13

tube or/dj was crawling out of an

3:16

oversized coffee mug while wearing a rainbow

3:18

com o now. The

3:21

Dj than studied howling the name of the

3:23

company is event he was helping. Oh.

3:30

Oh time to get

3:32

up, you silly little

3:34

dogs. way. This

3:39

is how the world ends now with a

3:42

bang that with Adidas at from an internet

3:44

personality. this is google

3:46

the annual developer conference google i

3:48

have the event where every year

3:50

google announces which technological breakthroughs coming

3:52

has in store for us twenty

3:54

one minutes later after the dj

3:56

had tucker himself out the show

3:58

began in earnest Welcome to Google

4:01

I.O. It's great to have all of you

4:03

with us. We have a few thousand... The

4:05

moment that shook me, that shook a lot

4:07

of people, it came after CEO, Sinter Pichai,

4:09

had made his opening remarks. He

4:12

introduced Google's head of search, who walked

4:14

on stage to funky elevator music. Thanks,

4:17

Sinter. With each of

4:19

these platforms, we haven't just

4:21

adapted. We've expanded what's

4:23

possible with Google Search. Liz

4:26

Reed explained that Google Search was about

4:28

to fundamentally change. In a way, unlike

4:30

anything anyone had seen in the last

4:33

quarter century. Now, with

4:35

generative AI, search will do

4:37

more for you than you ever imagined. So

4:40

whatever is on your mind, whatever you need

4:42

to get done, just

4:44

ask. And Google will do

4:46

the Googling for you. Google will do

4:49

the Googling for you? These seven

4:51

words, I'm not kidding. They made me

4:53

feel deeply uneasy, in a way that

4:55

announcements at tech conferences rarely do. What

4:59

Liz Reed means when she says Google will do the Googling

5:01

for you, is that from now on,

5:03

frequently when you ask Google a question, what's

5:06

the best Bluetooth speaker? Or what's happening

5:08

with the war in Ukraine? Instead

5:11

of being sent links to articles written by humans, they

5:13

AI will read those articles, and just provide you

5:15

with its own summary. There will be

5:17

links in the footnotes, but you can skip them. On

5:21

its face, a totally useful feature, but

5:24

as we watched the media apocalypse arrive, this

5:27

seemed like a pretty obvious accelerant. Almost

5:30

without exception, every website on the internet

5:32

depends on Google for traffic, for audience.

5:35

Google now seems to be saying that

5:37

highway we've built will be closing

5:39

the exits. A

5:41

report from the Wall Street Journal suggested that

5:44

online publishers, the average of which is already

5:46

limping and coughing like a 20-year-old cat, could

5:49

now expect to lose as much as

5:51

40% of their remaining traffic. I

5:54

wanted to talk to someone who could explain all this. How

5:57

we got to a place where Google defined so much of

5:59

the internet? and what to make of this new

6:01

change. So

6:03

of course, I called up Casey Newton,

6:05

the genius tech reporter behind the platformer

6:07

newsletter co-host of the Hard Fork podcast.

6:11

I wanted to talk to Casey because I

6:13

knew he actually had some very different ideas

6:15

about possible solutions to this apocalypse. And

6:18

besides, if I was gonna watch the world begin to

6:20

burn, I knew who I wanted to sit

6:22

next to. Hello?

6:26

Casey, how's it going? All right,

6:28

good. Am I coming through

6:30

okay? You're coming through loud and

6:32

clear. I think you're getting into even recording your voice.

6:35

Good, because it's one of the most

6:37

important parts of podcasting that I've learned.

6:40

Without it, it's just purely a theater of

6:42

the mind. If this airs, will I be

6:44

the first returning three-time champion on search engine?

6:47

Not only will you be the first three, Peter,

6:50

Calva and Ezra, I mean, Calva has been

6:52

very vocal about wanting to come back three

6:54

times before Ezra, and I don't think he

6:56

saw you coming in from the other lane

6:58

and knocking out. I'm a dark

7:00

horse, just like that one Katy Perry song. So,

7:04

okay, before we get to this week's news, can

7:06

you just give me the prehistory? Can you

7:08

tell me the story of Google search? Yeah.

7:12

So, PJ, when the internet was young

7:14

and exciting, it

7:19

was just a series of webpages,

7:22

famously delivered over a series of tubes, and

7:24

these webpages were so vast in their number that

7:27

to find them, we needed a box we could

7:30

type into, and there were

7:32

many boxes with names like Excite,

7:35

and HotBot, and InfoSeek, but

7:38

one day, a couple of Stanford grad

7:40

students come along with this thing that

7:42

is better at searching these webpages than

7:45

anything we've seen before, and it's called

7:47

Google. And basically, from

7:49

the minute anyone sees it, people are

7:51

saying, this is the one. They've

7:54

come up with some really clever stuff

7:56

that helped them find webpages better than

7:58

anything else, And the story... The of

8:00

the next. Twenty Five Years

8:02

is Google gradually wrapping it's

8:04

arms around the web, until

8:07

it essentially became synonymous with

8:09

it and why. I mean,

8:11

I remember like I am.

8:14

I. I'm really a not enjoying house. I

8:16

found myself saying I'm old enough to remember

8:19

and and her on a quakes but I

8:21

am old enough to remember the other search

8:23

engines like I remember Altavista. I remember Ask

8:25

Jeeves, I remember using like Aol surged and

8:28

I remember the feeling I got when I

8:30

first use Google. If I remember correctly, For.

8:33

Almost a feeling you get with like

8:35

a good night's tatty be T de

8:37

products were like oh this is better

8:39

this feels difference. What was happening under

8:41

the heard that made Google work better

8:43

than what preceded at. The. Did

8:45

a bunch of things but the most famous

8:48

is something called page Rank Paid Raped Maimed

8:50

effort One of the cofounders google Larry page

8:52

and the idea was really simple. It was

8:54

just that as they created as index of

8:56

all of the web pages they would look

8:58

to see which web pages were looking to

9:00

other web pages and have a bunch of

9:03

webpages went to the Lepage those a really

9:05

strong signal that it was valuable. So if

9:07

as they're crawling they see one hundred different

9:09

web pages linking to the New York Times

9:11

that's gonna rise up and search results as

9:13

people search for the New York Times. And

9:15

in fact it is the website of the

9:18

New York Times and so everybody sort of

9:20

gets what they what sells. a really good

9:22

and useful thing He i'd it enabled them

9:24

to become the default search I dread for

9:26

most of us. he adds. after that it

9:28

turned into basically the greatest advertising business that

9:30

any was a receipt. And why does it

9:32

hurt? into the greatest advertising business that anyone's

9:34

ever seen? Because. It turns

9:36

out that what search engines do

9:39

is they capture into heads and

9:41

desire if you are typing game.

9:44

New. Car you might be in the

9:46

market for new car, if you're typing issues,

9:48

you might be in the market for new

9:50

shoes and so really quite easily google us

9:53

it as go out to people who wanted

9:55

to advertise, to people who are in the

9:57

market for various products and services and it.

10:00

Started it actually had this idea of

10:02

running an auction so that advertisers could

10:04

bid to be above all the search

10:06

results and just worked incredibly well. Am

10:08

I would argue was actually does the

10:10

really fair bargained for anyone? If you

10:12

are are looking for shoes it probably

10:14

actually doesn't hurt you too much to

10:16

see one ad for shoes on top

10:18

of a list of links to other

10:21

website. So I think it's important to

10:23

say that the first year of decade

10:25

or so of Google Love It had

10:27

various problems. It is fundamentally but with

10:29

a good. Bargain for everyone. People got

10:31

to the web pages they were go into.

10:33

It was paid for by ads and thirty

10:35

five. Yeah. It's funny. Now we

10:38

have so many. I think it's not

10:40

true, but there's that idea of a

10:42

submissive seventy were. So we have so

10:45

many words for either technology making us

10:47

feel bad or capitalism behaving in ways

10:49

that we feel conflicted about. And people

10:51

talk about Extract Daves models. And there's

10:54

all these. Web products for it's

10:56

like you like it but it's doing something

10:58

to someone that's bad or it's offer you

11:00

something but it's like pulling for the have

11:03

your back pocket. Well it's doing it and

11:05

you're right. In the early stages of Google

11:07

in the first chapter of Enemies history it's

11:09

like this is great for everybody actually. Yeah

11:11

and. We. Should say it really

11:14

helps the web grow and establish itself.

11:16

Google Me the web much more useful

11:18

and the more useful the web became,

11:21

the more people rushed into it. Google

11:23

started showing ads odd other web sites

11:25

as if you're a publish or even

11:28

just a blogger that had decent traffic.

11:30

You can just run ads that Google

11:32

would manage as you can begin summit

11:35

money on the Web as the creator.

11:37

So you just see this huge rush

11:39

of talent and capital into the web

11:42

as Google leaves. That hard and making

11:44

it more useful for all of us. And

11:46

so. Another. Question that I've always wanted

11:48

to ask. Someone. Who had Now

11:50

there is a member? Whereas like there's

11:52

a bunch of search engines and Google

11:54

is superior one, What happened That Google

11:56

became. Like. I know not literally

11:58

the only one like when could use being

12:00

but why did Google pull out so far

12:02

ahead and never get caught? A big reason

12:04

is as that the more that people use

12:06

Google the better that it got. So for

12:08

example I you stab at example earlier have

12:11

somebody trying to find a New York Times

12:13

website and Google starts out with the same

12:15

page rank that says the actually have a

12:17

pretty good idea what you're looking for Right

12:19

now let's think about all of the people.

12:21

start visiting Google and they searched for the

12:23

New York Times and they click the link

12:25

and they go to the New York Times

12:27

and they don't go back to Google and

12:29

Google says ah, We serve the direct

12:31

link and it starts speeding that model

12:33

and it does that across every category

12:35

of search for every single thing is

12:37

all of a sudden Google has the

12:39

most accurate indexed by far have any

12:42

of the search engines he and of

12:44

the longer that it goes that as

12:46

becomes more and more troops it starts

12:48

again this memento that nobody else can

12:50

really match and at what point does

12:52

that. News industry to

12:55

the media industries start to enter

12:57

into this relationship with Google here

12:59

so you know from the start.

13:02

People. Were trying to figure out how

13:04

to we're. Optimize our web page

13:07

so that it floats to the top

13:09

of his Google search rankings. Because as

13:11

Google becomes a default place to start

13:13

your day on the internet, one of

13:16

these people are doing is searching for

13:18

news and so publishers. They're changing the

13:20

html. You're talking with people at Google

13:23

about what exactly are you looking for

13:25

A I Had It becomes this dance

13:27

and there are some players in the

13:29

game like I think probably most of

13:32

the publishers were. That was pretty good

13:34

actors. And then there. were a bunch

13:36

of unscrupulous fly by night characters that

13:38

were like just trying to sell you

13:41

a vacuum or whatever and one of

13:43

the like swarm every key word you

13:45

can imagine just on the off chance

13:47

that maybe their web page would get

13:49

at the top of the search results

13:52

and so it becomes is very competitive

13:54

adversarial saying it's he i have an

13:56

effect of that was google just became

13:58

increasingly powerful because faces It's

14:00

not just the publishing industry. It's like every

14:03

industry is beating down a path to Google's

14:05

front door saying, hey, how do I get

14:07

to be the top link on

14:09

the thing? And that becomes one

14:11

of the main drivers of Google eventually

14:14

taking over the web. It's just

14:16

such a strange thing. It happened

14:18

and so it seems normal, but it's weird

14:20

to contemplate the idea for how infinite

14:23

the internet is that

14:25

really the most normal experience you would have

14:27

on it is you search something

14:29

on Google and you visit one of three to

14:31

five links or you go on one of the

14:33

handful of social media websites and then that's it.

14:36

Yeah, I mean Google did do things to put

14:38

itself at the center of the news conversation. The

14:40

first thing it did was it had a product

14:42

called Google News where it just started to aggregate

14:44

headlines and you can visit Google

14:46

News and get a rundown of what was

14:49

happening around the world. Another thing that it

14:51

started to do, and this happened much later

14:53

in the mobile era, but eventually by

14:55

the time you Google something on your phone,

14:57

even before you search for anything, Google would

15:00

know your previous searches and they would show

15:02

you new stories that you might be interested

15:04

in. And all of a sudden

15:06

that was starting to send

15:09

a flood of traffic people's way. A

15:11

third thing that happened was that publishers just

15:14

started to pay attention to what people were

15:16

searching for. Like there are various tools that

15:18

let people understand, oh wow, a lot

15:20

of people are searching for the Game of Thrones trailer. It's

15:23

gonna take us four seconds to embed

15:25

the Game of Thrones trailer in our

15:27

website. Let's just go ahead and do

15:29

that so that when anybody searches Game

15:31

of Thrones trailer, maybe we'll rise to

15:33

the top and we'll be able to

15:35

gain that ad revenue. And here's where

15:37

I do think the publishers just made

15:39

a mistake because there was a lot

15:41

of easy traffic that was available. The

15:45

output wasn't actually that high

15:47

in journalistic quality, but

15:50

the revenue that was coming in from all

15:52

those visits was

15:54

pretty good. And so this dynamic was just

15:56

created where these big digital publishers just saw

15:58

this ocean of traffic. available to them. And

16:01

all they had to do is figure out what

16:03

are people searching for and what's the cheapest web

16:05

page that we can quickly get up to take

16:07

advantage of the traffic. Yeah. And it's

16:09

like sometimes as

16:12

a person who has worked in online media for my

16:14

entire adult life and spends most of my time thinking

16:16

about what tech companies have done wrong and not what

16:18

media companies have done wrong. It is

16:20

funny how much in that

16:22

chapter of internet, how much

16:25

of what got published would just be every

16:27

single website, whether they're a video game review

16:29

site or a national

16:31

newspaper or a blog or whatever

16:33

would just be like, Hey, the

16:36

people on the highway of the internet today

16:38

want to look at the trailer for the

16:40

movie. Let's slap that on our website. Like

16:42

it was so undifferentiated. Everybody posts like the

16:44

John Oliver clip, like everybody posts Saturday Night

16:46

Live, like just everyone's selling the same product

16:48

with very little differentiation. The example I

16:51

always think of is it felt like a decade

16:53

ago, every single news site was writing articles just

16:55

called what time is

16:57

the Superbowl. Do you want to tell

16:59

that story? Yeah. One of the most

17:01

popular queries on Google, as you might

17:03

imagine is what time is the Superbowl

17:05

because that is a day I'm told

17:07

when people who do not ordinarily watch

17:10

football games will watch a football game

17:13

and they don't know what time it is. PJ. Could you, if

17:15

without looking, do you know what time football games are on? I

17:18

have no idea what time it is. 5pm when's

17:20

the kickoff? So

17:23

the Huffington Post

17:25

realizes that it can

17:27

write an article that answers the question what

17:29

time is the Superbowl and it will be

17:32

a traffic bonanza akin to

17:34

the Superbowl itself. The what time

17:36

is the Superbowl post is the

17:38

Superbowl of SEO traffic for the

17:41

Huffington Post. And of course,

17:43

this idea lasts for about 30 seconds

17:45

because that every other publisher is like, wait, we

17:47

know what time the Superbowl is. We can just

17:49

put that on the web too. And

17:52

you can probably guess what the ultimate conclusion of

17:54

the story which is that Google says, we

17:57

also know what time the Superbowl is. We're just gonna

17:59

start showing it. on top of search results.

18:01

And it is that shift. Google

18:03

sort of realizing if what people are looking

18:06

for from us are just answers, we don't

18:08

have to leave it to the Huffington Post

18:10

and all these other hangers on to answer

18:12

people's questions. We can just start doing it

18:14

for them. And if you are the frog

18:17

in the pot of water that the entire

18:19

media industry has been for the past 25

18:21

years, this is when the temperature

18:23

went up by five degrees. After

18:29

the break, the temperature keeps

18:31

rising. As the quality

18:33

of search results declines over the years, as

18:35

websites become generally crappier in an effort to

18:37

get noticed by Google, the

18:40

death spiral continues. More

18:42

death spiral after these ads. So

19:40

search engine is brought to you by June's

19:42

Journey. June's Journey is a mobile

19:44

game you can download on your smartphone. Everybody

19:47

loves a good family mystery, especially one

19:49

with as many twists and turns as

19:51

June's Journey seven to the role

19:53

of June Parker and search for hidden clues

19:55

to uncover the mystery of her sister's murder.

19:58

Engage your observation skills to quickly uncover

20:00

key pieces of information that lead to

20:02

chapters of mystery, danger, and romance. Where

20:05

will each new chapter take you? June's

20:08

Journey is a hidden object mystery game with

20:10

a captivating detective story taking you back to

20:12

the glamour of the 1920s

20:14

with a diverse cast of characters. Each

20:17

new scene takes you further through a thrilling

20:20

murder mystery story that sets the main

20:22

protagonist, June Parker, on a quest to solve

20:24

the murder of her sister and uncover her

20:26

family's many secrets. I

20:29

have never met an iPhone game that I don't

20:31

like to play instead of thinking. Try

20:33

June's Journey. You can download it for free

20:36

on iOS or Android. Discover your

20:38

inner detective when you download June's Journey

20:40

for free today on iOS and Android.

20:43

Search Engine is brought to you by NetSuite. Okay,

20:46

quick math. The less your business spends

20:48

on operations, on multiple systems, on delivering

20:50

your product or service, the

20:52

more margin you have and the more money you keep. Obvious.

20:56

But with higher expenses on materials,

20:58

employees, distribution, and borrowing, everything costs

21:00

more. So to reduce

21:02

costs and headaches, smart businesses are

21:04

graduating to NetSuite by Oracle. NetSuite

21:07

is the number one cloud financial

21:09

system, bringing accounting, financial management, inventory,

21:12

HR, into one platform and one

21:14

source of truth. With

21:16

NetSuite, you reduce IT costs because NetSuite

21:18

lives in the cloud with no hardware

21:20

required, accessed from anywhere.

21:23

You cut the cost of maintaining multiple

21:25

systems because you've got one unified business

21:28

management suite. And you're improving

21:30

efficiency by bringing all your major business

21:32

processes into one platform, slashing

21:34

manual tasks and errors. Over

21:37

37,000 companies have

21:39

already made the move. So do the math.

21:41

See how you'll profit with NetSuite. Backed

21:44

by popular demand, NetSuite has extended its

21:46

one-of-a-kind flexible financing program for a few

21:48

more weeks. Head

21:51

to netsuite.com/pj. netsuite.com/pj.

21:55

netsuite.com slash

21:57

pj. Welcome

22:08

back to the show. So,

22:12

Casey had told us the story of the What Time

22:14

is the Super Bowl era, the chapter

22:16

in which August American news outlets were competing

22:18

against each other to be at the top

22:21

of a predictable annual Google search. I

22:25

remember when Google changed its website so that Google

22:27

itself could just tell you what time the Super

22:29

Bowl was, and I remember thinking,

22:31

does that make sense? I'd

22:33

understood why the publishers had wanted the web

22:35

traffic, but a news industry designed to

22:38

tell you what time the Super Bowl is is just not

22:40

that healthy of an industry. So no big

22:43

deal. I was not

22:45

savvy enough, however, to notice what it

22:47

might mean, as Google gradually put more

22:49

and more of the information that would

22:51

have lived on various websites onto

22:53

its own front page. When

22:57

you saw them start to answer those questions

23:00

themselves, as someone who studies

23:02

the power that platforms have relative to the industries

23:04

that depend on them, did

23:06

you make note of that shift? Yes,

23:09

but only in the sense that I thought,

23:11

well, here is a place where Google's power

23:13

is increasing. I've been

23:15

writing about Google for more than 10 years, and

23:17

I would say the whole time, they've been trying

23:20

to figure out how can we answer

23:22

more people's questions on what they

23:24

call the SERP or the search

23:26

engine results page. It's an acronym.

23:29

And to me, one of the most

23:31

interesting statistics about Google over the past

23:33

two decades is the rise

23:36

of what they call the zero-click

23:38

search, which is the search

23:40

that does not result in any

23:42

outbound traffic to anything. Oh, interesting.

23:44

Right. You sort of flash

23:47

back to the first days of Google, I would

23:49

guess that almost every search resulted

23:51

in a click to somewhere because Google itself didn't know

23:53

anything except for maybe where the webpage was that you

23:55

were looking for. But then you get into

23:57

the 2010s, and all of a sudden sudden,

24:00

it's not just answering what

24:02

time is the Super Bowl, it's

24:05

pulling snippets out of Wikipedia. It

24:08

is telling you what movies actors

24:10

are in. It's telling you what

24:12

movies directors have directed. And

24:14

all of this is appearing in various

24:16

little boxes and carousels on top of

24:19

the classic 10 blue links that have always

24:21

been the heart of Google. And

24:24

so, yeah, what I noted over

24:26

the past decade was every year,

24:28

there's another box, there's another widget,

24:30

there's another answer, and there's one

24:32

fewer, what time is the Super Bowl

24:34

bonanza for publishers to count on? I mean, I

24:36

feel like the other experience you could have in

24:39

the last 10 years on Google was that sometimes

24:43

when you search something that could

24:46

appear in one of those boxes, but didn't appear in one of those

24:48

boxes, you would end up on a

24:50

website that gave you that information that had been

24:52

so designed for Google that the experience actually landed

24:54

on that website. The example that I see people

24:56

refer to a lot is recipes where, for

24:59

whatever reason, the Google algorithm decided it

25:02

liked longer articles. The most privileged link would

25:04

not be a recipe for tomato soup. It'd

25:06

be like a five-page essay about

25:08

what tomato soup meant to someone and their grandmother

25:10

who gave them the soup recipe and blah, blah,

25:12

blah, blah, and you're scrolling down and you're like,

25:14

why is this written this way? This is completely

25:16

insane, but it's written that way for Google. And

25:18

at that point, you're like, could Google just please

25:20

tell me what tomato soup has made out of?

25:22

I'm pretty sure it's tomatoes. Right.

25:24

And it is a great example because we all

25:26

ran into it and we were all annoyed by

25:29

it. And this was just one of many things

25:31

that Google did to take over

25:33

the web experience. They also

25:35

created the Chrome web browser.

25:37

The Chrome web browser helps

25:39

to dictate HTML standards,

25:41

how web pages are built, how

25:43

browsers interpret them. It's able to

25:46

exert pressure in that way. So

25:48

it's not just like

25:50

in what order do links appear on

25:52

web pages? Like Google is actually dictating

25:54

the shape of the web itself through

25:56

all these different things. It's

25:59

weird. It's like I can't. I'm by always regular

26:01

with anniversary things and I'm like I wanted

26:03

to be like okay like the American highway

26:05

system. like the highway system which is is

26:07

meant to connect towns eventually you know it's

26:09

like people that upstairs on the highway in

26:11

the highway itself reshape set but I still

26:13

like what happened with google and internet is

26:15

more than that like a feeling it's like

26:17

as if the mouth became the saying instead

26:19

of the the thing it was describing the

26:21

only mean. Yeah. They they built as

26:23

like the greatest I was the so that the

26:25

internet had ever seen an ad over time it

26:27

is just shrug to the size of a parking

26:30

lot and anybody who searches that is is like

26:32

driving around in a circle the parking lot and

26:34

why did that happen because like everyone is experience

26:36

at the what me that happen. I. Mean

26:38

this is my answer to that

26:40

Would be that Google just wound

26:42

up being the arguably the biggest

26:44

economic victor from the Internet in

26:46

terms of surly the amount of

26:48

digital advertising revenue that they were

26:50

able to generate from the internet.

26:52

A digital advertising revenue is like

26:54

the single biggest category of revenue.

26:57

I think that's well. I don't

26:59

know. we should look that up.

27:01

Bullet Setup. We. Did like

27:03

the sub for professional podcast years. Alphabet.

27:06

Google confusingly named parent company made

27:08

three hundred seven billion dollars Last

27:11

year. Google Search alone accounted for

27:13

one hundred seventy five billion. The.

27:16

New York Times by comparison. Two.

27:18

Point Four billion. The. Publishing

27:20

industry is. It was really amazing when it

27:22

was as like newspapers but nobody was making

27:24

one hundred billion dollars a year right? like

27:27

school was able to sort of go out.

27:29

He had. over the years

27:31

more more the advertising revenue just a

27:33

crude to look like google just became

27:35

like z most powerful c s publishers

27:37

just became disempowered a laid people off

27:39

the scramble your whatever they could do

27:41

to like get up higher the surf

27:43

results they would use it would work

27:46

for a times than the our the

27:48

machines that you know more people would

27:50

be laid off like all this does

27:52

have a downward pressure on the quality

27:54

of things like people couldn't afford to

27:56

take big swings anymore they couldn't afford

27:58

to hire bid staffs anymore you just

28:00

get more of these generic websites telling

28:02

you about that week's movie trailer. So

28:05

basically, Google got too much of the

28:07

money and the rest of the digital

28:09

media ecosystem, in my opinion, did not

28:11

get enough. And there's a very

28:13

solid narrative that has unfolded over the past

28:15

few years that Google just isn't as good

28:18

as it used to be at searching for

28:20

things. In part, that is just

28:22

because there are so many more ads now

28:24

on the high value searches that people often

28:26

do on Google. A predictable

28:28

way that Google has added revenue over

28:30

the past couple years when they need

28:32

to show growth to Wall Street is

28:34

they'll literally just add one more sponsored

28:37

link to mobile search results. So

28:39

maybe it used to take five

28:42

links before you would see what they call an organic

28:44

result, so a result that has not been paid for.

28:46

Now I think it's up to seven, right? Now

28:49

maybe most people don't even realize that those are sponsored links

28:51

and they're perfectly happy to click on ads all day. But

28:53

for people who are a little bit savvier and you just

28:55

kind of wanted to see a web that

28:58

wasn't totally corrupted by commercial values,

29:00

that just feels like it is

29:03

harder to find. And it is, in part, because people

29:05

are not making as much stuff for the web as

29:07

they used to because there is not as much money

29:09

in it. Right. It's funny. It's

29:12

like Google's really two businesses that even within themselves

29:14

are sort of in competition. There's

29:17

the search part of it where we're serving people who

29:19

want to find stuff on the internet. The

29:21

advertising part of it, which is like we want people who

29:23

are trying to find stuff on the internet to get distracted

29:26

on their way and stop at our store or stop at

29:28

the advertiser store. And it's like I

29:30

wonder if even within the company they feel like

29:32

search and ads are in combat. They

29:35

always have that. This is kind of

29:37

what led a former Googler in the

29:39

early days to coin their famous catchphrase,

29:42

don't be evil. This was what don't

29:44

be evil was about. It was about

29:46

not compromising the integrity of what they

29:49

were doing by reaching for the easy

29:51

revenue. And over time,

29:53

I think that they have just reached

29:55

more and more for the easy revenue

29:57

and have not thought enough about the

29:59

health. of the broader web ecosystem

30:01

that ultimately they do depend on. So

30:05

tell me about this most recent news.

30:07

What happened? So this week at Google

30:09

I owe they laid out some changes

30:11

to the way that search results will

30:13

work and there's the way that it

30:15

will work in the near future and

30:17

then there's the way that it will

30:19

work in the medium term and

30:22

the way that it's going to work in the near-term

30:24

future and in fact, a lot of people have had

30:26

this feature already and preview. I've had it for several

30:28

months now is when you

30:31

search for some things Google

30:33

will just show you

30:35

an AI generated summary of the results.

30:37

So if you say something like what's

30:39

the best laptop I can buy right

30:41

now before this feature rolled

30:43

out you would see a list of

30:45

links to sites like Wirecutter that had

30:47

done a lot of rigorous testing of

30:49

laptops now with what Google is

30:52

calling AI overviews. It'll say like here's some of

30:54

the best laptops of 2024 as well

30:57

as that's what Google will judge by experts and they'll sort of

31:00

look at 50 different companies that have written

31:02

a page like this and

31:04

they will summarize it and they will sort of show you

31:06

in little footnotes maybe

31:08

who wrote that story but most people of course

31:11

are not going to click on the footnotes. They're

31:13

just going to see a little summary. So

31:16

why does that matter? Well, this

31:18

is one of the places where publishers

31:21

are still making money.

31:23

Affiliate links if they do these sort

31:25

of Wirecutter style tests of products and

31:27

if people buy something because they read

31:30

that web page then the publisher gets

31:32

a little bit of a kickback now

31:34

those kickbacks are probably going to start

31:36

going away too. And

31:38

so this is this again one more

31:40

place where publishers aren't going to see

31:42

revenue but it's actually much bigger than

31:44

that because the real idea here PJ

31:46

is that whereas browsing the web used

31:48

to be considered something of a pastime

31:50

to older folks like you and me

31:53

now. It's being sort of presented as a chore something that

31:55

you shouldn't have to do something that you should just like

31:57

Google read the web for you show you a bunch of.

32:00

results, and you'll never have to leave

32:02

Google. So the reason that this is

32:04

so important is this is really the

32:06

first step toward you

32:08

not having to visit the web anymore, because

32:11

Google is going to read the web for

32:13

you. And like, I

32:16

feel like what you just said is, let me put it

32:18

this way, but it this way, recently, I had seen a

32:20

different search engine, I think was perplexity, that was sort of

32:22

doing the same thing. And we talked about this on search

32:24

engine, we were talking with Ezra Klein, it was like, this

32:26

seems kind of bad. They're like taking the journalism, but they're

32:29

not paying for it. What's gonna happen to underlying journalism? That

32:32

seemed like a moment. Google

32:35

unveiling this functionality. Would you

32:37

say this is a bigger deal? Yes.

32:41

So I hated what perplexity was

32:43

doing. I hated what Arc

32:45

search, another company was doing that was basically

32:47

exactly the same thing. And the

32:49

reason I hated it so much was that I knew

32:52

that Google would do it. Because

32:55

in some ways, it is a

32:58

better user experience, right? There's

33:00

a reason that people really like asking

33:03

chat GPT questions. And

33:05

it is that they do not get a big

33:07

research project back when they say, show me the

33:09

best shoes, right? Chat GPT will just say, Oh,

33:12

if you're a man, here's like 10 kinds of

33:14

shoes that should be in your wardrobe. Google will

33:16

show you 4800 links to websites. It's

33:20

clear to me, what is the better user experience,

33:22

right? So I knew when I saw

33:24

what perplexity and arc and some of these others

33:26

were doing that Google was going to feel pressure

33:29

to do the same thing. But

33:31

still, it had to happen. And then this

33:33

week, it happened. I

33:35

wondered if Google wouldn't do the same thing,

33:38

because I feel like, okay, as a journalist,

33:40

I understand why I don't want this to

33:42

happen, because it makes us less valuable. It

33:44

makes our jobs more precarious. As

33:46

an internet user, I understand why people like

33:49

I have used perplexity, I have found perplexity

33:51

to be useful. But I also feel like

33:53

I'm like pirating music

33:56

except for the thing I make, so I

33:58

feel worse about it. Sorry, musicians. I

34:00

thought Google might not do it because I'm like, yeah,

34:02

in the short term, you're giving people a better

34:04

user experience. But if you roll this

34:06

thing out, and I don't

34:08

know how to estimate how many journalists lose their

34:11

jobs from this, but 10%, 20%, 40%, you're killing

34:13

the input for the machine that you

34:17

need. And it seems that Google needs that

34:19

machine more than a perplexity or an arc. So I thought

34:21

maybe they wouldn't. Yeah. But I

34:23

do think that they will move a little

34:26

cautiously here because to

34:29

some degree, that is almost certainly true.

34:32

It's really every publisher in the

34:34

world went away and

34:37

restaurants stopped creating websites

34:40

and dry cleaners stopped posting their

34:42

phone numbers online. This

34:44

does create a problem for Google. I'm

34:48

just not sure, one, that

34:50

it's as big a problem for

34:52

them as a journalist I would like

34:55

it to be. And

34:57

two, they are going

34:59

to be in control of this entire process,

35:02

right? Like they have their fingers on the

35:04

knobs and the levers. And

35:06

so they can just tweak it like 5% this

35:08

way or 10% that way, they can

35:10

see what happens. And

35:12

if nothing really breaks for them, then

35:15

they can dial it another 5% or 10%. Every

35:19

other business on the internet might be

35:21

kicking and screaming the whole time, but

35:23

there is almost truly nothing they can

35:26

do because Google is in control. So

35:28

to me, what this moment has meant is

35:30

that on stage this week at

35:33

Google I.O., the company essentially put the

35:35

web into a state of managed decline

35:38

where they said, without saying it, that

35:41

the web was really useful for 25 years,

35:44

but we don't need it anymore. Because

35:46

with generative AI, we'll be able to tell

35:48

you anything that the web could have told

35:50

you and you're not even going to have to

35:53

leave Google to get the information. You're

35:55

not somebody who is like, like one of

35:58

the things you and I talk about sometimes We're

36:00

not talking into a microphone, but I guess

36:02

also talking to a microphone is that like

36:04

to be clear We mostly talk to each

36:06

other via microphones mostly their relationship you've set

36:08

up But

36:11

like I feel like we're both Tech

36:14

journalists in our generation for the most part can

36:16

be Incredibly skeptical of tech

36:18

companies incredibly paranoid about what they're releasing

36:21

and like with good reason that's like

36:23

earned skepticism Earned paranoia. I feel like

36:25

you and I are a little

36:27

bit unusual in that. We're still

36:29

sort of like stubbornly We have some optimism in

36:31

us We have some optimism in us like

36:33

my feeling about AI has been like I'm

36:36

not just gonna like try to go destroy the machines with an

36:38

axe. I want to see how this is good I want to

36:41

see how this is bad and Maybe

36:43

this is just my solvism where it's like

36:45

why you weren't worried about Dolly, but now

36:47

you're upset about this But like it's

36:50

unusual for me to hear you Talk

36:53

about things in this dire way Yeah,

36:56

and I'm a little nervous that

36:58

I am over rotated here Right

37:00

and yet if you look

37:02

at like the trajectory of the journalism industry

37:04

since I got into it in 2002 It

37:08

pretty much just is a line falling off a

37:10

cliff. I Want

37:12

to say yeah correlation is not always

37:14

causation not always causation and I'm not

37:16

saying it's all the Internet's fall I'm

37:18

not even saying this was Google's job

37:21

to fix this necessarily It

37:24

just did become the economic engine

37:26

that powered the web and

37:29

so the moment when it says This

37:32

honestly just is not that important to us anymore Like

37:34

regardless of what you think of like whether that is

37:36

good or bad or like what Google straps is you

37:39

it just Is a big deal

37:41

for publishers, you know, there's been some reporting on this

37:43

in the Wall Street Journal And

37:46

analysts believe that publishers might lose

37:48

between 20 and 40 percent of

37:50

their traffic over That it's

37:52

year as this stuff rolls out right because we should

37:54

say what happened this week was Google Took

37:57

this AI overview experience that they've been testing

38:00

now rolled it out across the United States. By

38:02

the end of the year, they say a billion

38:04

people around the world are going to have it.

38:06

So it's gone from this very small test to

38:08

now a billion people are going to have it

38:11

by December. And once that

38:13

happens, if people are really losing 20

38:15

to 40% of their traffic,

38:18

it's just is going to be we're

38:20

just going to see so many more

38:22

publications go out of business. Last year,

38:25

a bunch of publications went out of

38:27

business, BuzzFeed News, Vice as we know

38:29

it, the new gawker protocol sites that

38:31

just kind of disappeared. And

38:34

when I think about the few and the

38:36

proud big publishers that remain, if you walk

38:38

into any of their C suites, and we're

38:40

like, what's your plan to have 40% left

38:44

traffic by December? I

38:46

don't think anybody has a really good plan for that. After

38:57

a short break, if Google shutting

38:59

down a huge chunk of traffic to new

39:01

sites is as big a deal as Casey

39:03

fears, what are the possible solutions

39:05

here? How can someone

39:07

putting their work on the internet safeguard their ability

39:10

to make a living? That's

39:12

after these ads. Rocket

39:51

Money is a personal finance app that

39:53

finds and cancels your unwanted subscriptions, monitors

39:56

your spending and helps lower your bills that

39:58

you can grow your savings. With

40:00

Rocket money I have full control over my subscriptions and

40:02

a clear view of my expenses I can

40:05

see all my subscriptions one place and if I see

40:07

something I don't want rocket money can help me cancel

40:09

it with a few taps Rocket

40:11

money will even try to negotiate lower bills

40:13

for you That's 20% all you have

40:16

to do is submit a picture of your bill and rocket money takes

40:18

care of the rest They will deal

40:20

with customer service for you Rocket

40:22

money gave me a free trial because they're on the

40:25

podcast. I started using it. I really like it I'm

40:27

now actually paying for a subscription for them It's

40:30

not very much money But the funny thing is that

40:33

it shows me my rocket money subscription on

40:35

rocket money and I have

40:37

considered trying to get rocket money To

40:39

negotiate the prices of their subscription down

40:41

against themselves We'll see if

40:44

it works Rocket money has over 5

40:46

million users and has saved a total of 500

40:49

million dollars in canceled subscriptions Saving

40:51

members up to 740 bucks a

40:53

year when using all of the apps features Stop

40:56

wasting money on things you don't

40:58

use cancel your unwanted subscriptions by

41:01

going to rocket money.com/search That's

41:03

rocket money.com Slash

41:09

search search engine is brought to you by Squarespace

41:12

Squarespace is the all-in-one website platform

41:14

for entrepreneurs to stand out and

41:16

succeed online Whether

41:18

you're just starting out or managing a growing brand Squarespace

41:21

makes it easy to create a beautiful

41:23

website engage with your audience and sell

41:25

anything from products to content to time

41:27

all In one place on

41:29

your terms Despite reporting a

41:31

lot of stories about the internet. I

41:33

find websites hard to make Squarespace

41:36

offers blueprint AI and SEO tools

41:38

You can start a completely personalized

41:41

website with the new guided design

41:43

system Squarespace blueprint They

41:45

also have flexible payments You can make

41:47

checkout seamless for your customers with simple

41:49

but powerful payment tools Squarespace

41:52

also offers the fluid engine the

41:54

next generation website editor from Squarespace

41:57

It's never been easier for anyone to unlock creativity.

42:01

Head to squarespace.com for a free

42:03

trial and when you're ready to

42:06

launch go to www.squarespace.com/engine to save

42:08

10% off your first purchase of

42:10

a website or domain. That's squarespace.com/engine.

42:13

Search Engine is brought to

42:15

you by LinkedIn Jobs. When

42:17

you're hiring for your small business you want to find

42:20

quality professionals that are right for the role. That's

42:22

why you have to check out LinkedIn Jobs. LinkedIn

42:25

Jobs has the tools to help find the

42:27

right professionals for your team faster and for

42:29

free. LinkedIn isn't just a job

42:31

board. LinkedIn helps you hire professionals you can't

42:33

find anywhere else. Even those who aren't actively

42:35

searching for a new job but may be

42:37

open to the perfect role. In

42:40

a given month over 70% of LinkedIn

42:42

users don't visit other leading job sites.

42:44

So if you're not looking on LinkedIn you're looking

42:46

in the wrong place. On LinkedIn

42:48

86% of small businesses get

42:51

a qualified candidate within 24 hours.

42:53

Hire professionals like a professional on

42:55

LinkedIn. LinkedIn knows that

42:58

small businesses are wearing so many hats and might

43:00

not have the time or resources to hire. LinkedIn

43:03

is constantly finding ways to make the process

43:05

easier. They even just launched a feature that

43:07

helps you write job descriptions making the process

43:09

even easier and quicker. Post

43:11

your job for free at

43:14

linkedin.com/PJ search. That's

43:16

linkedin.com/PJ search to post your

43:18

job for free. Terms and

43:21

conditions apply. Welcome

43:32

back to the show. I have

43:34

to say I was feeling unusually defeated hearing

43:37

Casey describe what he thought was about to

43:39

happen with the internet. I

43:41

wanted him to walk me through the possible solutions here.

43:44

How other people were thinking about the problem

43:46

of these AI bots chowing up the work

43:48

of human beings and spitting out

43:50

good enough summaries. Most

43:52

things are fixable. I wondered what the

43:54

possible fixes were. For instance the most

43:56

tried and true American response to anything.

43:59

Lawsuits. Couldn't journalists and

44:01

publishers just sue Google or sue

44:03

whoever's nap stringing their work? So

44:07

there is a big lawsuit

44:10

filed by the New York Times against

44:13

open AI, essentially for

44:15

the same reasons. And

44:18

that is now unfolding. It

44:22

is an open legal question whether

44:24

it can be

44:26

permitted for a company like

44:29

Google to go

44:31

in and look at all

44:33

of the articles that a publisher like

44:35

platformer has ever published and use those

44:37

to train a large language model. The

44:40

case for it being illegal is,

44:42

hey, you're stealing my work, knock it off, don't do

44:45

that. You're just like taking my labor and using it to make

44:47

something valuable. I don't want you to do that. So maybe it's

44:49

illegal. The case that

44:51

it is legal, though, is, I don't

44:54

know, all of us are allowed to go

44:56

and read web pages and form thoughts and

44:58

do other things based on that information. And

45:01

these LLMs are not reproducing what they are

45:03

ingesting perfectly in most cases. And so are

45:05

you really going to tell computers that they

45:07

can't read the internet? Because guess what? Computers

45:10

are already reading the internet in all sorts

45:12

of ways. So this is just going to

45:14

have to get litigated. But of

45:16

course, all the big tech companies are

45:19

just making the bet that courts are

45:21

going to side with the big corporations

45:23

here and the publishers are going to be out of luck. Right.

45:26

And it seems like of all the possible futures you

45:28

could imagine, there's ones where the courts are like, you

45:30

do kind of have to pay them some money. There's

45:32

ones where they say, no, no, no, this is aggregation.

45:34

This is what you guys were doing with humans. But

45:37

there's no world where they say turn off the machine. Some

45:40

people do think that some people will

45:42

say that chat, GPT and Google Gemini

45:44

are like the fruit of a poisoned

45:47

tree, that because you use all this

45:49

copyrighted material in the creation of them,

45:51

you will actually have to destroy them.

45:55

But vanishingly, few people I've spoken to

45:57

think that is a likely outcome. You're

46:00

a person who publishes on the

46:02

internet. You're a publisher. You're not an enormous

46:04

company, but you're someone who is swimming in

46:06

the same ocean that you're worried is being

46:08

destroyed. What does it mean for you? I

46:11

can't figure that out yet. I had an

46:13

opportunity back when Platformer was still on Substack.

46:15

If you want to know why we're not

46:17

there, that's another great story that you can

46:19

Google. For

46:21

now. And they had some

46:23

kind of toggle where I could

46:25

say, don't train your

46:27

LLMs on Platformer. I'm a anti-rogues.

46:31

I didn't switch it on because,

46:33

number one, Platformer publishes three articles

46:35

a week. We're a

46:37

very low volume publisher. We've

46:40

never frankly even relied on Google for very much

46:42

of our time. I started a newsletter because I

46:44

didn't want to have to fight platform algorithms for

46:46

the rest of my life. I just wanted to

46:48

write about them. So I've never thought that whatever

46:52

happens in Google is necessarily going to

46:55

be like curtains for Platformer. Sometimes

46:57

I think I've weirdly worked up for this given

47:00

how little I actually expect this to affect me

47:02

in the near term. Except

47:04

that I love the web.

47:06

I grew up on the web. The web brought

47:08

me everything that I have. And

47:11

Platformer is much better when there is more

47:13

journalism for me to read and have thoughts about

47:15

and inspire questions in me and send me chasing

47:17

stories of my own. I don't think

47:19

this is going to have a huge effect on me

47:21

directly. But indirectly, it feels

47:24

like the only story. Yeah. Which

47:26

is what is the internet going to look like

47:28

in five years? And you think it's

47:30

that fast. This

47:34

stuff is moving very quickly.

47:36

Every few weeks, it seems,

47:38

one of these AI makers

47:41

comes out with a more

47:44

efficient version of a model or a

47:46

model that has been tuned for some

47:48

specific purpose. Like maybe it's better at

47:50

education or it's better at science or

47:52

it's better at health care. And

47:56

we're just kind of in that

47:58

liftoff stage. And this stuff is... starting

48:00

to accumulate these chatbot assistants

48:03

are getting better and better. A

48:05

lot of people probably watch the

48:07

demo that OpenAI had this week

48:09

where they've made this assistant that

48:11

has this uncannily emotional voice. It's

48:13

like actively flirting with you now.

48:15

I was watching their videos, it's

48:17

crazy. They had one, we did

48:20

a story like last year about

48:22

Andrew Leland, interviewing Andrew Leland about

48:24

his blindness and he was saying

48:26

how he uses this app called

48:28

Be My Eyes where you can

48:30

connect with a human being and they will tell you

48:32

like hey, this shirt doesn't match your pants or whatever. And

48:36

one of the things OpenAI demoed was this video

48:38

where Be My Eyes is now AI powered and

48:40

they were showing a person who did not have

48:42

sight walking down the street holding their phone and

48:44

their phone was saying the cab's here, put your

48:46

hand up. Buckingham Palace is in front of you.

48:49

The flag indicates the king is there. And I'm

48:51

like, I'm not against this.

48:53

This is progress, this is amazing. I

48:56

just want the work I love and the web I

48:58

love to exist in the future.

49:00

Yeah, another way I would put that

49:02

is just like I want the benefits

49:04

to be a little bit more evenly

49:06

distributed, right? I know that the majority

49:08

of the spoils here are going to

49:10

go to the companies that do the

49:13

most innovation and I'm basically fine with

49:15

that. But again, when you

49:17

flashback and think about what Google was

49:19

like in the early 2000s, I

49:22

just feel like we had a better bargain. We got

49:24

a guide to the web that was really fast and

49:26

easy to use and reliable. They got

49:28

a bunch of advertising revenue and

49:31

there was a rising tide that was lifting

49:33

all boats. And what I'm worried

49:35

about is that tide has now sort of,

49:38

I don't know, come in and washed

49:40

out a lot of what was on

49:42

the shore and Google is

49:45

going to be the last boat

49:47

standing or whatever, I've sort

49:49

of lost track of this metaphor. I don't know

49:51

if Google is a boat now. It's

49:54

gone completely away from me. But if you

49:56

just sort of imagine them as a large

49:58

thing that survived whatever I was... talking about,

50:00

that's what it would be. When

50:03

we talked about this on Search Engine last, and

50:05

it's weird, I really feel conflicted where I'm like,

50:08

when you said that this feels to you like the only

50:10

story in some ways, that's how I feel. And I'm like,

50:13

I don't wanna belabor people's patience

50:15

with my curiosity with this, but it

50:17

really feels like global

50:20

warming's a bigger deal, climate change is a bigger deal.

50:22

But it feels like climate change for the thing I

50:24

love, and it feels like it's happening so quickly, and

50:26

it's very hard for me not to think about it

50:28

a lot. When we spoke to

50:31

Ezra about it, I was like, what do you do?

50:33

And Ezra being like an ethical

50:35

person who believes that people should act ethically was

50:37

like, look, if you love journalism, it

50:39

is incumbent on you to pay for it. And I agree,

50:41

people should pay specifically for Search Engine, and they have money

50:43

left over in their budget, they should pay for platformer,

50:46

maybe search engine twice, platformer once, I

50:48

pay for platformer. Thanks, BC. Of course.

50:50

I pay for search engine. Thank you.

50:53

We're modeling good behavior to the internet.

50:55

That also feels like, yeah, people should

50:57

do that, but it feels

50:59

like the problem is larger than $5 a month. What

51:02

do you think the solutions are? Well, I mean,

51:05

set the journalists aside for it. I mean, this

51:07

is something else I think is important to say.

51:09

So people don't just think we're navel

51:11

gazing about our own industry. Google

51:14

does not only deliver traffic to publishers,

51:16

it's how people discover all sorts of

51:18

businesses, right? You move to a new

51:20

town, you need a dry cleaner, you

51:22

need a dog walker, you wanna know

51:25

some cool restaurants or cool bars. Right

51:27

now, imperfect as it is,

51:30

all those things can like jockey for a

51:32

position. They can go to the search engine

51:34

optimizers and they can get some tips. And

51:36

hopefully if they're a really good restaurant or

51:38

really good dry cleaner, they'll pop up to

51:40

the top of search results. And they might

51:42

not even have to buy a Google ad,

51:44

right? Like they can compete just by being

51:46

really good. We

51:48

are talking about the beginning of a future where

51:51

all of those webpages, whatever all those other

51:54

businesses are doing to sort of wave their

51:56

hands and say, hey, like we exist, that

51:58

is all just getting sub- assumed into

52:01

an even more complicated and

52:03

mysterious set of algorithms that's just

52:05

going to be spat out and

52:07

you're just going to be told,

52:09

yeah, here's the three dry cleaners

52:11

in your town. And hopefully

52:13

one of those will be good. So

52:16

I don't want to overly romanticize like the present

52:18

state of affairs, because I do think that SEO

52:21

has like ruined a lot of things. But

52:24

it still

52:26

seems preferable to me to a

52:28

world where there is just this

52:30

kind of mystery AI giving you

52:33

the answer to everything and actively

52:35

discouraging you from visiting websites to make up your

52:38

own mind. And I just keep going back to

52:40

like at this conference this week, they just kept

52:42

coming back to this phrase, let Google do the

52:44

googling for you. And what they were telling us

52:47

was searching the web as a chore using the

52:49

internet as a chore. Google is now the thing

52:51

that stops you from having to do that. Sure.

52:53

Google is just going to be the Star Trek

52:55

computer. It's going to tell you whatever you need to know. Don't

52:58

worry about visiting the web anymore. And

53:00

while they protested us not really what we meant, and

53:03

we're still going to send lots of traffic, and we

53:05

believe in the web. At the end of the day,

53:07

I was like, No, like you're telling us what you

53:09

want to do. You've been building it for 20 years.

53:11

And now you're really close. So I

53:13

just think it's time that we take them seriously about that. What

53:15

does it mean to take them seriously except for to worry? I

53:20

have the best, worst, dorkiest

53:23

answer to that question PJ, which

53:25

is that we have to finish building the Fediverse.

53:33

Really? Yes. You're already so

53:35

upset that I'm making you talk about this.

53:37

And that's fine. We should all

53:40

be upset that we have to talk about the Fediverse. But

53:42

that's where Google has driven us to talk

53:44

about the Fediverse. But in a way

53:46

that my mom can understand it. Yeah. So

53:49

the Fediverse is

53:51

a way for people

53:54

to take back the internet for themselves.

53:57

It's a way to have a identity

54:01

and connect to other

54:03

things that are important to you

54:06

online and just not worry

54:08

about having to fight through a Google algorithm

54:10

or a Facebook algorithm. In fact, you can

54:13

bring your own algorithm if you want

54:15

to. I'm already doing such a bad job of

54:17

explaining what the set of things are. I would say honestly,

54:19

my mom is kind of a false flag in this because

54:21

the truth is like, here's what I understand about the set

54:23

of verse. Yeah, I understand that there were people who watched

54:26

Twitter go up in flames and said, never

54:28

again should one man be able to control

54:30

the algorithm. And so from

54:32

now on, among Twitter clones, which there

54:34

must now be 1000, each

54:37

one bad in its own specific way, the

54:40

Fed averse means that you will be able to have

54:42

an account that is not linked to any one of

54:44

those sites. And that I could post a boring

54:46

post on threads, but it could be read by

54:48

angry people on blue sky, the idea that these

54:50

things are federated amongst each other, but not centrally

54:52

controlled, right? But honestly, like if you said keep

54:54

explaining the Fed averse for five more minutes, I'd

54:57

be like, I've run out of steam. I don't

54:59

really get it. But that's

55:01

pretty good, right? It's a

55:03

collective term for these

55:05

various web platforms that

55:07

use open source and

55:10

decentralized protocols to let

55:12

different platforms communicate and

55:14

interact across these like different

55:16

hosting services. That's like probably about

55:18

as technical as you need to

55:21

go. But the way I think of it is, it's

55:23

just like a way to bring some

55:26

humanity back to the internet. It's a

55:28

way to sort of rest it back

55:30

from these giant mega tech platforms. It's

55:32

a way to personalize things through liking

55:34

to like sort of customize them. And

55:37

so it is starting with these social

55:40

platform mastodon was the first thing in

55:42

the Fed averse threads, which is actually

55:44

now much bigger than mastodon is a

55:46

meta product, but it is part of

55:49

the Fed averse flipboard is

55:51

joining it. WordPress is joining it ghost, which

55:53

is this hosting provider that I use for

55:55

platformer is going to join it. And so

55:57

Sunday, you might just like I have an

55:59

app on your phone and instead of just

56:01

going to Google to see what's the news

56:03

of the day, you just open up your

56:05

app that links you to the Fettivers and

56:07

you might be following some publishers there, you

56:09

might be following some creators there, there might

56:11

be some ads in it so those folks

56:13

are getting money, maybe you do pay a

56:15

subscription to some of the publishers in there

56:17

so you get to see all of their

56:19

paywall posts and they just kind of show

56:21

up right in your feed. And

56:24

while there's a lot to figure out in terms of

56:26

how do you create a good user experience, how do

56:28

you make that kind of more fun and useful

56:30

than Google? That just kind of feels

56:32

like the direction to go to me

56:35

because instead of one giant walled garden

56:37

that is just keeping you there, keeping

56:39

all the revenue for itself, it is

56:41

a way of rebuilding a web where

56:43

there's just a lot of organic connections

56:45

between people and publishers who like each

56:47

other and have ways about how we

56:50

can make and share money with each

56:52

other. And so if it works,

56:54

we're gonna have something I think that feels much

56:56

better than the world we have today. But

56:58

could it work? It feels like this

57:01

is like so under informed and I should not be saying

57:03

it into a microphone and putting on the internet and maybe

57:05

I won't but like it feels like

57:07

one of those ideas where you're like yeah it'd

57:09

be nice but like you

57:11

know like things

57:14

that are civic and volunteer and like parks that

57:16

end up being trashed and then everybody just goes

57:18

to the mall, do you think it could really

57:20

work? Here's my case that

57:22

it could work. Okay. Thres is an

57:24

app that has a

57:27

hundred and fifty million monthly users.

57:30

It is ten months old and

57:32

it is part of the Cetiverse. So that

57:34

means as hard as it is to believe,

57:36

a hundred and fifty million people every month

57:39

are in the Cetiverse. For the most part,

57:41

they don't know about it and they don't

57:43

care and that's actually a great sign because

57:45

as we've just established through our

57:48

tortured explanations of the Cetiverse, nobody

57:50

wants to understand what it is

57:52

or how it works. So

57:55

we're already working on one of the biggest problems

57:57

with the Cetiverse. The

58:00

thing is, PJ, I'm not the only

58:02

person who's worried about this. Yesterday,

58:04

I met with two folks.

58:06

One is guy Eugene Rachko, who's the

58:08

founder of Mastodon. The other guy's Mike

58:10

McHugh, who's the co-founder of Flipboard. And

58:13

these two guys are running at this Fettever

58:15

stuff at 100 miles an hour. And the

58:17

main thing they wanted to let me know

58:20

was just how many other people are building

58:22

this stuff with them, right? There's a lot

58:24

of old timers and even young people around

58:27

who remember the early promise of the internet,

58:29

who remember how exciting it was that we

58:31

were going to have this thing that was

58:33

decentralized, that was open, that shared the wealth

58:36

with a lot of people. And they're

58:38

going out and they're picking

58:40

off these name brand websites

58:42

like WordPress and Tumblr, the

58:44

Verge, the site where I

58:46

used to work, they're pushing

58:48

into federation. So at this

58:50

moment, is it a crazy

58:52

band of insufferable obnoxious rebels?

58:54

Absolutely. But I ask

58:56

you, PJ, what movement in the history

58:59

of the world has not begun with

59:01

a band of insufferable obnoxious rebels? I

59:05

can't think of one. OK, so maybe the Fetteverse

59:07

saves us. It might be the Fetteverse. I mean,

59:09

look, a lot of people are going to use Google. Again,

59:12

one of the reasons why I'm so mad,

59:14

PJ, is that this is going to work,

59:16

OK? It's like I'm mad because it feels

59:18

like game over. I'm mad because

59:21

most people are going to be totally happy

59:23

to get the sort of Star Trek computer

59:25

answer and not give two thoughts to any

59:28

of the labor that went into producing the

59:30

answer. And I'm sure Google

59:32

will have a great business for itself. But

59:36

some of the people that worked at

59:38

Google in the early days were really

59:40

idealistic about what the web could be. And

59:42

I believed in that optimism. And I'm

59:44

not ready to give up on it. So if that

59:46

means that I have to learn what the Fetteverse is

59:49

and explain it to other people, that's what I'm going

59:51

to do because a better world

59:53

has got to be possible here. Casey, you're

59:55

making me feel things. I'll

59:57

tell you this. I felt things talking to Eugene and

59:59

me. I was honestly

1:00:01

shocked at how emotional I was at Google

1:00:03

I am. Like it felt weird that I

1:00:05

was as upset as I was walking around

1:00:07

this developer conference. And I think I was upset

1:00:10

because I felt gaslit honestly, because nobody at Google

1:00:12

would just stand up and say, we

1:00:15

actually do have a long term plan to replace most

1:00:17

web visits with our wild garden. So

1:00:19

that's like kind of why I was mad. But I

1:00:21

was also really just pessimistic about the future. Then

1:00:24

I sat down with those two and they were like, here

1:00:26

are the next three things that we're going to build. And

1:00:29

here's like the next three big platforms that we're

1:00:31

going to go after and get them to federate.

1:00:34

And I'm like, this

1:00:36

might work too because the thing

1:00:38

about Google taking most of the winnings of

1:00:41

the internet for itself is, there's

1:00:43

a lot of other people on the internet that would also like

1:00:45

to eke out a living. And

1:00:47

they're highly motivated to make it work for them. So

1:00:50

there will always be a rebel

1:00:52

alliance and I would not count

1:00:54

them out because companies that are

1:00:57

old and have a lot of

1:00:59

money, they get really lazy and

1:01:02

they can't move as

1:01:04

quickly as sometimes they need to adapt to

1:01:06

the future. So if the Fediverse

1:01:09

folks can build a better future that is truly

1:01:11

more fun to use, it'll be really

1:01:13

small for a while relative to the size of Google.

1:01:16

But there's no reason why it couldn't grow

1:01:18

very large in the end. All

1:01:20

right, Casey, I'll see you on the Fediverse. Do

1:01:25

you have a Fediverse account? Are you

1:01:27

on mastodon.social? Okay, so I started a

1:01:29

mastodon account and then I forgot the password. Yeah,

1:01:32

that's the single most common story about the Fediverse,

1:01:34

by the way. Really? Pretty

1:01:36

much. I went on threads and I

1:01:38

was like, this is very boring. You

1:01:40

do post there. I see your threads post. Don't

1:01:43

pretend you're above threads. I post

1:01:45

episodes of search engine, but I don't like hang out and

1:01:47

make funny jokes. Well, maybe you should try it. Maybe you

1:01:49

didn't try it. Maybe I'll try

1:01:51

to make a joke on threads and see how it feels. Casey,

1:01:54

thank you for talking about the past and future of the

1:01:57

internet we've developed. You're welcome. Casey

1:02:02

Newton, his newsletter platformer is essential

1:02:05

to understanding our quickly changing internet.

1:02:08

You can also find him on the wonderful

1:02:10

weekly technology podcast, Hard Fork, with his co-host

1:02:12

Kevin Roos. Did

1:02:15

we do a good enough job explaining the set of

1:02:17

verse this week? I feel like I'm still a little

1:02:19

confused by it, so maybe you're also still a little

1:02:21

bit confused by it. As

1:02:24

far as I can tell, one way to think

1:02:26

of the problem of Google search is that it's a

1:02:28

problem of Monopoly. If the

1:02:30

internet had more than one popular search engine,

1:02:32

the entire web wouldn't have been somewhat

1:02:34

corrupted by trying to appeal to Google.

1:02:37

And Google wouldn't then have had to replace

1:02:39

many of its useless web results with AI

1:02:41

summaries. On a

1:02:44

federated internet, people with followers can take

1:02:46

their followers with them from platform to

1:02:48

platform. So maybe that internet

1:02:50

resists Monopoly more easily? I

1:02:52

don't know. Honestly, I'm still a little confused

1:02:55

how that fixes search and AI. You

1:02:58

know what? If people want more Fediverse talk,

1:03:00

email us, and we'll consider revisiting this in

1:03:03

a future episode. If you want less

1:03:05

Fediverse talk, email us and let us know too.

1:03:08

How in the weeds should this show go? This is

1:03:10

a question for you to help us answer. Journalism

1:03:13

is a service industry. You can reach us

1:03:16

at searchengine.show. After

1:03:18

the break, we have a podcast recommendation, which

1:03:20

has to do with the themes of this

1:03:22

episode. Stick around. Thanks

1:03:54

for watching. I'll see you next time. So,

1:04:37

as we've been thinking about Google's effect

1:04:40

on the internet and about small publishers

1:04:42

trying to survive, we've been

1:04:44

closely watching the launch of one of

1:04:46

my favorite new news websites, 404 Media.

1:04:50

They cover the internet, they are brilliant

1:04:52

reporters, and they're a small, independent outfit

1:04:54

like us. They

1:04:56

feel like a canary to me. Like if

1:04:58

they can make a business succeed online, I'm a

1:05:00

little bit less worried about the future. And

1:05:03

I know other journalists who are watching 404 with

1:05:06

the same question right now. What's

1:05:09

relevant here though is that the team at

1:05:11

404 are very much at the mercy of

1:05:13

Google's algorithms. And recently

1:05:15

they posted an episode where they just

1:05:17

talked candidly about what it's like for

1:05:19

a small group of humans to try

1:05:21

to survive while fighting AI

1:05:23

websites that are constantly scraping their work.

1:05:27

You should go listen to that episode. It's

1:05:29

called Why Google is Shit Now. I

1:05:31

found it really fascinating, and it just goes way more

1:05:33

in depth to answer the question a lot of people

1:05:35

have. How come even before this

1:05:38

AI thing, Google search just seemed to

1:05:40

mostly stop working? You

1:05:42

can find out there. I'm going to put a link in our

1:05:44

show notes. Also, as

1:05:46

we mentioned last episode, we

1:05:49

are right now heading towards the end of season one

1:05:51

of Search Engine. We're doing a

1:05:53

board meeting with all of our paid

1:05:55

subscribers on Friday, May 31st to discuss

1:05:57

show business. That's business about the

1:05:59

show. not news about Hollywood,

1:06:01

the meeting Friday, May 31st, 1

1:06:04

p.m. Eastern time. We will be sending out

1:06:06

a Zoom link to join WeGov. This

1:06:09

is only for our paid subscribers, people who

1:06:11

are members of Incognito Mode. If

1:06:13

you're not signed up, there's still time. Go to

1:06:16

searchengine.show. You can also send us questions there that

1:06:18

you'd like to hear answered at this board meeting.

1:06:20

And if you sign up, you'll get a lot of other

1:06:23

stuff, which you can read about on our website. Again,

1:06:25

that URL is searchengine.show. If you're

1:06:28

a paid subscriber, look out for

1:06:30

an email with a link next week and

1:06:32

mark your calendar, May 31st, 2024, 1 p.m. Eastern. Search

1:06:43

Engine is a presentation of Odyssey and

1:06:45

Jigsaw Productions. It was created by

1:06:47

me, PJ Vogt, and Shruti Pinamaneni, and is produced

1:06:49

by Garrett Graham and Noah John. Back

1:06:52

checking this week by Holly Patton. Theme,

1:06:55

original composition, and mixing by Armin

1:06:57

Bazarian. Our executive producers

1:06:59

are Jenna Weiss Berman and Leah Reiss-Dennis.

1:07:02

Thanks to the team at Jigsaw, Alex Gibney,

1:07:04

Rich Perrello, and John Schmidt. And

1:07:06

to the team at Odyssey. JD Crowley,

1:07:09

Rob Miranda, Craig Cox, Eric Donnelly,

1:07:11

Kate Hutchison, Matt Casey, Mora Curran,

1:07:13

Josephina Francis, Kurt Korten, and Hilary

1:07:16

Schott. Our agent is

1:07:18

Oren Rosenbaum at UTI. Follow

1:07:20

and listen to Search Engine for free on the

1:07:22

Odyssey app or wherever you get your podcasts. Thanks

1:07:25

for listening and see you soon.