0:00

It's time for Twitter this week. In

0:02

Tech, we have two

0:04

of my favorite people on two geniuses.

0:08

Corey Dottro and Alex Cantor. It's

0:11

pro big tech against big tech.

0:13

It's gonna be a great conversation. We'll talk

0:15

about Google, Killing, yet

0:17

another service Mark Zuckerberg

0:20

fighting in the Octagon Elon

0:23

Musk's robot and

0:26

the plan, Peter Teals plan,

0:28

to buy into Britain's

0:30

national health service that and a whole lot

0:32

more coming up. plus the big scam

0:34

and podcast advertising. It's

0:37

a head on Twitch.

0:41

Podcasts you love. from

0:43

people you trust. This

0:46

is toys. This

0:53

is Twit. This weekend, tech. Episode

0:55

eight hundred ninety five recorded

0:58

Sunday October second twenty twenty

1:00

two eastern blocks.

1:03

This episode of this week in tech is brought to

1:05

you by Neurava, Tired

1:07

of the complexity and cost of traditional pro

1:10

AV solutions for large spaces,

1:12

Nueva has simplified everything about

1:14

meetings and classroom audio. You get

1:16

great audio and plug and play systems

1:19

that are easy to install and manage

1:21

and cost fraction of insealing systems.

1:23

Visit noreva dot com slash

1:26

twist. And by, Eight

1:28

Sleep, Good Sleep is the ultimate game

1:30

changer. and the pod is the

1:32

ultimate sleep machine. Go to

1:34

h sleep dot com slash Twitter to check out

1:36

the pod and save hundred fifty dollars

1:38

to check out Eight Sleep currently ships within

1:40

the US, Canada, the UK, select

1:43

countries in the EU and Australia.

1:46

And by? podium. Join

1:48

more than one hundred thousand businesses that

1:51

already use Podium to streamline their

1:53

customer interactions, see how Podium

1:56

can grow your business, Watch a demo

1:58

today at podium dot com slash

2:00

twins. And by

2:04

policy genius, making it

2:06

easy to compare your from top companies.

2:09

Policygenius can help you make sure

2:11

you're not paying a cent more than you have to

2:13

for the coverage you need. head to

2:15

policy genius dot com slash twit

2:17

to get your free life insurance quotes

2:19

and see how much you could save.

2:26

It's time for twitch this week in tech.

2:28

Show we cover the week's tech news.

2:30

We have such a good panel today. I have

2:33

limited it to just two people, Agile Alex

2:35

Kombrawitz, is here from the big

2:37

tech sub spec. newsletter in

2:39

the big technology podcast. Hello,

2:42

Alex. Good to see you.

2:43

Hey, Leo. Great to see you. Are you plugging

2:46

the Vancouver hockey team?

2:48

This is the Grizzly. So this is throwback.

2:51

So for listeners, I'm wearing this throwback

2:53

Vancouver Grizzly's sweatshirt.

2:55

It's pretty cool. It's by Mitchell and Ness, which is think

2:57

my favorite clothing brand, nice. And I'm not a

2:59

fashion guy, but I did see that the grizzlies

3:01

wore the throwback jerseys, and then this girl's

3:04

coat grizzlies. And are you said are you from

3:06

Canada here. Now the closest

3:08

I've been to Canada is Washington State.

3:10

And of course, I've been to Vancouver's Washington

3:12

State in on the West Coast.

3:15

I do have some family in in Toronto,

3:17

shout out Toronto. But

3:20

but no, I'm a New Yorker. Let's

3:22

go let's go meds. there's nothing weirder

3:24

than New wearing a Vancouver grizzlies

3:26

shirt, but, you know, I will go. I

3:28

have one promise to make it to you today. What's that

3:30

is that I'm here to bring the weird. Let's go.

3:33

You're good. Also,

3:35

here, I'm thrilled to have him, Corey, doctor Ro.

3:37

You know Corey very well, I'm sure. He's

3:40

got a new book. In fact, we did a triangulation,

3:42

Corey, and Rebecca Giblin,

3:45

his co author and I on Thursday,

3:47

you might check that out on the Twitter triangulation

3:49

feed or the Twitter event feed, really

3:52

fascinating conversation. Great to

3:54

see you, Corey. You are a Canadian authentic.

3:57

I I we we walk among you. We are

3:59

like serial killers. We look just like everyone

4:01

else. He seems so normal. How

4:03

did we how could we have known? Although, I

4:05

became an American citizen about ten weeks

4:07

ago. I know. I heard I remember that. Congratulations.

4:11

Are you having regrets yet?

4:13

No opposite, actually. Like,

4:16

the worst things get the more glad I am that

4:18

I have more rights -- Oh. -- that I would

4:20

otherwise not be entitled to as someone who

4:22

is merely permanent resident. Now

4:24

your wife is American. Yes. Now,

4:26

she's British. She's British. But she's also

4:28

American now. So she she is

4:31

a Anglo American. My

4:33

daughter and I are Anglo Canadian Americans,

4:35

but I'm through my father entitled to Polish

4:38

Azerbaijan Belarusian, Russian

4:40

citizenship. Great. So I'm I

4:42

might get some of those. Most of them

4:44

are countries that don't intend to ever. I'd wanna

4:46

go, but But, you know --

4:48

Yeah. -- every now and again, you wanna go someplace

4:50

that was like a normalized nation during the

4:52

Cold War and you got the

4:55

right password for it. Yeah. It's nice

4:57

to have, like, you know you

4:59

know James Bond, a whole stack of passports,

5:01

you know, you can go to in your in your go bag

5:03

just in your Right. Just a kid. That's that's what

5:05

I'm down for. I mean, even if it takes the rest of

5:07

my life to get them and hypothetically, also,

5:10

I could get it as realty password. So, like, eight

5:12

in total. Wow. I can I just

5:14

like to have a big you should act with the rubber

5:16

band around. Exactly. Have it in the safe

5:18

with ten thousand dollars in cash, and

5:20

then just leave it open every once in a while. Case

5:22

anybody. just wondering by

5:25

maybe a a sidearm, and

5:27

you're set. You're ready to go. Someone

5:29

in the IRC San Mateo says Putin

5:31

Mike transcript me. Although, I'm a

5:33

fifty one year old man with two artificial hips

5:35

and cataracts. doesn't matter. It doesn't matter. It

5:37

doesn't matter. We're in court. That's really safe. He's

5:39

in the bottom of that. He'll take you.

5:42

How's your hips by the way? Oh, they're

5:44

getting better. Yeah. Good little by little. Last

5:46

time you were on, you showed us your your

5:48

three d model of your hip bone. Yeah.

5:51

Now I I let I should have brought them in. I've I

5:53

I made a brass cane topper

5:55

-- Yeah. -- cast from my hip bone.

5:57

Yes. And then the bone itself is in a shadow

5:59

box. But they're all they're both out in our

6:01

bar in the backyard. Right. Lisa.

6:03

I think Please talk with it. Leave them there.

6:05

That's good. Google.

6:10

Google. Google. They're added again.

6:13

Stadia. Three years old,

6:15

Google has announced They're

6:18

shutting it down. They're streaming gaming service.

6:21

I guess, really, the only question is,

6:24

whoever thought Stadia would survive

6:26

The good news is they are giving

6:28

you your money back. Why

6:33

did Google Alex even think it

6:35

could get in the gaming. What was the

6:37

what was the process there? Well,

6:40

I think that what we're seeing now is this demarcation

6:42

between old Google and new Google. Yeah.

6:44

And all Google said, let's experiment with

6:46

everything we can. Throw money behind

6:49

it because why not? Right? And that's how you end

6:51

up having a thing like stadia get funded in the

6:53

way that it got funded. and become a big

6:55

prong of Google's strategic vision

6:57

is because they were gonna make

6:59

every experimental project part of that

7:01

vision and eventually some of it would

7:03

stick. And that would be the, you

7:06

know, potential next big reinvention

7:08

for Google. Did Google get, you

7:11

know, financial responsibility.

7:14

I mean, did they start getting it? Did they grow up?

7:16

Or Well, that's what I think that's exactly what I

7:18

think we're seeing right now. And this the canceling

7:20

of stevia is, like, maybe just the beginning. Right?

7:23

We had Sumed up Chaya, CEO of Google

7:25

and Alphabet at Code a few weeks ago saying

7:27

Google needed to get twenty percent moderation.

7:30

Yeah. And and I think that, like, rather

7:32

than look at this as, like, you know, Google's bet

7:34

on gaming, I think we can really see this as a signal

7:37

for where this company is going. and maybe

7:39

where all of big tech is going, which is

7:41

that those experimental projects that used to get

7:43

the money because,

7:44

again, why not? They're not

7:45

gonna get the money anymore.

7:47

And I think this is gonna do two things. One, it's

7:50

gonna add bring needed focus to a lot of these companies.

7:52

But two, it's gonna open the door for

7:54

companies who would have otherwise been in these

7:56

incremental areas and gotten wiped out by Big

7:59

Tech to start competing

7:59

with Big Tech in a way that they haven't been before.

8:02

So I think this is, again, so just a very

8:04

dynamic, interesting time for Big Tech

8:06

and Stadia is one example

8:08

of what we're seeing in terms of the change

8:10

that's just going to accelerate, I think, over the

8:12

next few months. Do you think Coriant's maybe

8:14

also trepidation about government regulation

8:16

that maybe Google's saying we should pull some

8:18

of the tentacles back in?

8:21

I mean, III don't think

8:24

that so I think that that

8:26

merger scrutiny is a lot easier to

8:28

imagine the state imposing than merger

8:31

review and unwinding. unwinding

8:34

mergers is really hard. It's very expensive

8:36

and it takes a long time. I think it's probably a

8:38

necessary corrective And

8:40

I I do have one weird trick that

8:42

we can talk about later for how I think we could

8:44

do a lot of it very quickly. But merger

8:47

scrutiny is far more likely. And since they didn't buy

8:49

stadium. There there I

8:51

don't think there would be worried about that. But I

8:53

did wanna say that, you know,

8:56

it's underappreciated the extent

8:58

to which Google bought its way to glory instead

9:00

of inventing its way to glory. You know,

9:03

this is a company that's had one and a half internal

9:05

successes. they made a great search engine and a pretty

9:07

good hotmail clone. All of the other

9:10

things that they built internally crashed and burned,

9:12

and all of the successes that they have are things

9:14

they bought from someone else. And this is just the

9:16

latest example. Right? Google video, stank,

9:19

YouTube, succeeded. Google

9:21

couldn't build mobile OS, but Android

9:23

came along. And people will say, you know,

9:25

that Google Photos is

9:27

an internal success, and it's true, but it's an internal

9:30

success because it comes bundled on android. which

9:32

Google bought from someone else. And they bought Pecassa,

9:35

which is a lot of the back end -- Yeah. --

9:37

as well. I'm looking to know

9:39

your Google's purchase starting in thousand

9:41

one when they bought Desjardins. I forgot

9:43

about that. Desjardins. Yeah.

9:46

Remember Desjardins and blogger, of course, was

9:48

their third big acquisition. They

9:50

bought AdSense. They didn't make that up. They

9:52

deployed their whole ad tech stack, their server

9:54

management, their mobile platform, their video platform,

9:57

you know, customer service HR software,

9:59

like

9:59

all of it, their docs platform,

10:02

all came from other companies. Yeah. And

10:04

what about their room? mean, what about the browser?

10:06

Chrome, you're right. Chrome is Chrome is a browser

10:09

they did build internally. You're right. That's an omission.

10:11

Thank you. And the I take it bad. led

10:13

that project is, of course, the CEO today.

10:15

Sundar was -- Right. -- in charge. -- two and a half successful

10:17

products. You're absolutely right. But

10:19

still, it's an increasing And I don't mean to be

10:21

so. I I don't mean to be flip here.

10:23

Right? I it's but it's not it's also true

10:26

of lots of other companies that they're buying their way

10:28

to glory. And the reason

10:30

I bring that up because you asked me about anti

10:32

trust is that historically, companies

10:35

were prohibited from both merging

10:37

with major competitors and also

10:39

buying nascent competitors that they might relies

10:41

on their way to becoming a threat. And,

10:44

you know, the the modern antitrust that

10:46

was practiced for the forty years of kind of

10:48

Reaganomics that seems to be coming to a

10:50

close was extraordinarily tolerant

10:53

of acquisitions as a growth strategy. But

10:56

as I say, I think that's coming to a close. In

10:58

the UK, the Competition of Markets Authority is

11:00

challenging. you know, even small acquisitions

11:02

like Facebook buying Jiffy, which

11:05

I refuse to call Jiffy. And And

11:09

and, you know, III think that you're gonna see

11:11

more of this. And and I think it's

11:13

necessary corrective. I mean, the extent to

11:15

which VC's have become

11:18

effectively, like, corporate

11:20

recruiters who basically say,

11:22

alright, we're gonna gonna put little money into

11:24

a a startup whose purpose

11:27

is to basically produce a postgraduate

11:29

portfolio piece that we're gonna pretend

11:32

as a product, just to prove that they can work

11:34

together as a team, and then a tech company

11:36

will buy them, throw away the product, and just

11:38

put them to work. And the VC's

11:40

equity will just be like a finder's fee

11:42

it's a it's a grotesque and

11:45

wasteful way to conduct business

11:47

to say nothing of to get, you know, innovative

11:49

products into the market. I'm just looking

11:52

at this. The list of acquisitions

11:54

Google has made, the many great

11:57

companies that I've loved

11:59

when they were around that have

12:01

basically disappeared. Bikasa is a

12:03

good example. Jiku is a good example.

12:06

You you know, you can go on and

12:08

on and on They've been

12:10

they've basically been a graveyard. Feedburner,

12:15

Grand Central, which became I mean, Feedburner's

12:17

still in there. Yeah. And so is so is

12:19

Google Voice as Grand Central. Although

12:22

part of the problem with Google killing stuff

12:24

is it it makes people nervous about adopting

12:26

Google services. because -- Mhmm.

12:28

-- there's always this risk that Google's gonna

12:30

lose interest. So, Alex, you're

12:32

saying this is kind of a sububrious adjustment

12:35

of there. of their financial, you know,

12:37

financial maturity, but it's also risky,

12:40

isn't it? I do think it's risky.

12:42

Yeah. I think this is the I I made this

12:44

point. and big technology a

12:46

couple weeks ago talking about how,

12:48

you know, you might end up seeing stuff that Wall

12:50

Street likes a lot in the short term. things

12:53

like greater profitability. They

12:55

might even make some offensive moves to try

12:57

to take up some competitors. But in the long term,

13:01

I think focusing on profitability, especially when

13:03

it comes to big tech companies that are ten, twenty,

13:05

thirty, forty years old, leaves them

13:07

vulnerable to outside challenges in a way

13:09

that they weren't before. Google

13:12

is at least trying to make a people

13:14

who bought into stadia. They're gonna refund

13:16

your fancy controller

13:19

and any games you bought, which is actually

13:21

a big deal because you had to buy the games to

13:23

play them. on stadia, but they are

13:25

not gonna refund your subscription fees.

13:27

And there are some people who are

13:29

a little miffed about Google pulling

13:32

the plug like a YouTuber

13:34

named its Color TV who

13:36

says he had devoted Five

13:39

thousand nine hundred seven hours

13:41

to building up a character in Red Dead Redemption

13:43

two, a character which will

13:46

be lost. That's two hundred and forty

13:48

six days. lost in

13:50

January when Google

13:52

pulls the plug because there's currently no way

13:54

to transfer that character. Yeah. to

13:57

your own copy of Red Dead Redemption.

13:59

I'm sorry about about it. Like, for

14:02

folks who think that they that it's a good idea

14:04

to invest in products other people's platforms

14:06

without thinking that there's a risk that the platform

14:09

might pull the rug. It's

14:11

absurd at this point. Every single

14:13

thing that people build on one of these tech platforms,

14:15

you have to understand that if

14:18

if, you know, for some reason, they decide

14:20

that, you know, they don't wanna support it anymore,

14:23

It's

14:23

done. And

14:23

just do you have my thing? But you don't have

14:25

a Do you have a choice is the problem?

14:28

Right? I mean, where do you go if you're not gonna

14:30

build on at this point, you can't build

14:33

on any you're gonna it's YouTube or

14:35

TikTok. My son is two two point

14:37

one million followers on TikTok is

14:39

starting to build a career as a TikTok. creator,

14:43

but it's it's not like you could do that,

14:45

you know, on your own blog anymore.

14:47

Right? Yeah. Of course. It's different. And even

14:50

in other words, congratulations to him. He's

14:52

done. I appreciate it. You do it you do

14:54

it with your understanding that that it

14:56

might just be moment in time. And and

14:59

I think we all have to be okay with that when we're on

15:01

these other platforms because there's no other

15:03

way around it. They all Oh, by the way, I'd say

15:05

TikTok is get like, providing that opportunity

15:07

for your son to, you know, create that business.

15:09

That's a great opportunity. It wouldn't be there.

15:11

Yeah. So it it goes both ways. But

15:13

all this stuff really needs to be viewed as, you know,

15:15

potentially temporary situations because

15:18

that could always fit over time.

15:20

And often it does. Howard Bauchner: I

15:22

I agree with your cautionary note

15:24

there. I do think that

15:26

when we say there's no other way, you're right that

15:28

there is no other way right now, but it's not like there's

15:30

no conceivable other way. Right? Like

15:33

when RSS was designed, in

15:35

and its core was this idea of

15:37

blocking, lock in, and orphaning, and

15:39

so on. Yeah. So, you know, there's there's an XML

15:41

directive you can put with RSS

15:44

that says this feed is permanently moved to a different

15:46

address. So if you and your hosting company

15:48

part ways, you just send that directive out

15:50

the next time a podcatcher pulls down that XML

15:53

file, it goes, oh, I'm just gonna relocate my

15:55

my bookmark for this to a different server.

15:58

and you can just take your audience with you.

15:59

I

16:00

understand why YouTube hasn't built that,

16:03

but I I don't understand

16:05

why we should why we shouldn't want them

16:07

to build it. Right? You know, like, if

16:09

we're going to get a better deal for

16:11

creators from YouTube, one of the

16:13

ways that we'll do that is by YouTube having

16:15

a legitimate fear that if they give creators a

16:17

bad deal, that those creators will go elsewhere,

16:20

a separate issue to the one about about

16:22

this Red Dead Redemption character. But again,

16:25

at least within a single game, it's

16:27

easy to see how those stats

16:30

could be moved over. I mean, this is just a small

16:32

database entry. But I think that

16:35

firms don't like doing this because

16:37

they like their App Store to be

16:39

like a sole portal into payments

16:42

and other sources of value. And

16:44

if they do create that interoperability, that

16:46

they there is that possibility that customers

16:49

will jump ship. Right? If they're taking a thirty percent

16:51

rate or fifteen percent rate or whatever it is,

16:54

Stadia was taking out of Red Dead Redemption's

16:56

publishers. The inability

16:59

to port a character is a feature and not a bug.

17:02

It's not a technical challenge. You

17:05

know, you that they YouTube

17:07

could disclose to red dead red dead redemption,

17:09

all of that material, and and give

17:12

publishers the the technical means to

17:14

move that character over from all file.

17:17

That's not But they they just choose not to.

17:19

Right. don't you think that if you spend

17:21

the money creating this stuff, you should be able

17:23

to make the rules. Let me give an example like brick

17:25

and mortars to it. Right? If you have a restaurant,

17:28

Yeah. And you you you shouldn't

17:30

just make you'd be required to

17:33

store all the person's food preferences

17:35

so that they can then go to another

17:38

restaurant and basically mimic

17:40

exactly what you've been giving them without having

17:42

that restaurant having to work for it. should be

17:44

something that starts in the ground up. So

17:47

I I mean, that the user should be able to help, please.

17:49

Yeah. They I mean, honestly, I know what my preferences

17:51

are, so I can bring them with me. Yeah. But

17:53

if the if somehow the restaurant was able

17:55

to lock up my preferences, that

17:57

wouldn't be a good situation. Would it? I

17:59

don't Of course. Yeah. Listen. I think that, like, if

18:02

you spend the effort, you run the business, you're spending

18:04

the time. Yeah. I'll never go to that restaurant

18:06

then. If they're gonna deal my preferences and

18:08

keep them both for themselves. Let's let's

18:11

say you'll not sign you go to to let's

18:13

say you go to pizza restaurant number one.

18:15

Right? And you have, like, a you

18:17

know, you have And they name a pie for me.

18:19

The Leo pie, which is using

18:21

great tomatoes, pepperoni, And

18:24

that's all and then I I can't take that to another

18:26

restaurant because it's Oh, you shouldn't. No.

18:28

But the other restaurant can't The other

18:30

the other restaurant should be able to make that.

18:33

but there shouldn't be a requirement to go to this

18:35

restaurant and sort of please send the rest of download

18:37

Leo's rest special. Yeah. Right. And then give

18:39

it to the competitor. That competitor figure out on their

18:41

own. But LaPenski o IHOP. RSS

18:44

is a good example because, I mean, this whole

18:46

network is dependent on RSS. That's why

18:48

that podcasting works. But didn't

18:50

the platforms come along and say, yeah, we're gonna

18:52

kill RSS creating

18:54

some sort of spurious allegation

18:58

that RSS isn't working or it's not right

19:00

or it's not good. They did the same thing with XMPP

19:02

They don't want interoperability, so they

19:05

actively kill it. Don't they? And

19:08

let me be clear. I'm not saying necessarily

19:10

that I want YouTube to be forced to do this.

19:12

But in the absence of meaningful competition,

19:15

YouTube is neither being disciplined by

19:17

firms nor by regulators. And

19:19

that means that they can their corporate preferences

19:22

carry an enormous amount of weight, and

19:24

that furthermore, the lack of competition,

19:27

which arises in part out of this

19:29

anti competitive vertical integration that I was just

19:31

talking about where where Google just buys the companies

19:33

that might later compete with it means

19:35

that it has enormous power over policy

19:38

So while I I support absolutely

19:42

YouTube's fair use claims, I was a

19:44

a staunch supporter of YouTube when Viacom

19:46

was suing them for a billion dollars. I

19:48

think that it's not YouTube's

19:50

job to make sure Viacom's business

19:52

model is intact or to

19:54

respect their business model. I think what sauce for the

19:56

goose is sauce for the gander. And if you

19:59

were to try and make a tool that

20:01

allowed a YouTube

20:02

the

20:03

broadcaster to take their audience with

20:05

them to a rival platform, which

20:08

is I think analogizes to the

20:10

kinds of things that people did that YouTube

20:12

did when they made a tool that allowed

20:14

people to take the video they liked and put it

20:16

on YouTube and and only have to

20:18

respond to takedown and so on. YouTube

20:21

would reduce you to, like, radioactive rubble.

20:23

They'd they'd say you violated their terms of service.

20:25

They'd say you violated their computer fraud and abuse

20:27

act. They'd say you violated the DMCA.

20:30

and and they would they would put

20:33

you out of business for doing onto them what they

20:35

did onto others. I'm just saying I'm just

20:37

saying that They should

20:39

face the same competitive pressure. They gave

20:41

rise to something as innovative as and great

20:43

as YouTube so that the next innovative great

20:45

thing can come along.

20:48

That seems fair. You

20:50

wouldn't be against that. Right, Alex? I mean,

20:53

You

20:56

don't have to wealth. I mean, is there a

20:58

bill in the the Senate to promote interoperability?

21:01

Is that And that's the Access Act? It wouldn't

21:03

it wouldn't touch this, but it's The Access

21:05

Act is about exposing APIs for social media

21:07

and few other kinds of platforms, app stores

21:09

as well. I and

21:12

then the EU, there's the Digital Markets Act.

21:14

they're they embody the the punch line

21:16

of that Irish joke. If you wanted to get there,

21:18

I wouldn't start from here. Right. Like,

21:21

you have this, you know, you have this

21:23

situation where these firms are very dominant

21:26

and where they do act as gatekeepers. Right?

21:29

Where they're, like, they're Matt well

21:31

Matt Matt Mullenweg's post about why Tumblr doesn't

21:33

have porn anymore and why it never will was pretty

21:35

instructive here. He was like, you know, we we

21:37

submit updates to Apple three

21:39

times a month. And at any

21:41

time, they might just arbitrarily decide that the

21:44

filters we had last month are no longer good

21:46

enough, and then we just go out of business.

21:48

And he's like, I don't know how how Twitter gets away

21:50

with it. I don't know how Reddit gets away with it, but

21:52

we don't and we couldn't. Apple has

21:54

has chosen to make exceptions

21:56

for some firms and not others. We only have

21:58

hundred million users instead of

22:00

you know, however or we only have a million ten

22:02

million users instead of hundred million users.

22:05

So maybe that's why but

22:07

it it just It just puts Apple in the position

22:09

of picking winners and losers in the marketplace. And

22:12

I don't think the answer is to say Apple, you

22:14

must carry all apps no matter whether you feel

22:16

they're good or bad. but I think that

22:18

its customers should be allowed to

22:20

choose a different app store. Right? It is

22:22

after all their phone, it belongs to them. You

22:24

know, as you as you say, if if you add

22:26

the value to it, EG by opening your

22:28

wallet and buying it, then it should be

22:30

yours to use as you feel like. And,

22:34

you know, the fact that Apple doesn't

22:36

have to face meaningful competition from other

22:38

app stores for the hardware it's sold

22:41

means that a connect in this very high handed and

22:43

opaque way And it it

22:45

does just, you know, it gets to structure

22:47

the entire mobile market, kind

22:49

of, or half of it, the other half

22:51

being structured by Google, but no one elected

22:53

them. And, you know, mostly what

22:55

they use to to attain that structuring

22:58

is not technology, but the law. It's the fact

23:00

that if you were to try to unlock an Apple

23:02

phone and you know, sell a dongle that Jill

23:04

broke your phone and let you choose another app store,

23:07

Apple would sue you. So they're happy to have the state

23:09

regulators step in. and prevent

23:11

people from offering more

23:13

choice to people who own vices who wanna

23:16

use their property in different ways, but they

23:18

abhor regulation when

23:20

someone steps in and tells them how to use their property,

23:22

I would actually prefer to just withdraw the

23:24

legal protections from Apple, not

23:27

impose new obligations on them.

23:29

You say the access act should be modified

23:31

to allow a right of private lawsuit

23:34

-- Yeah. -- which is what Texas

23:36

has done with their gun. laws

23:38

and their abortion laws. Actually, no.

23:40

California did it with the law laws of Texas.

23:42

No. With the laws. It's pretty that's that's slightly

23:44

different because it's disinterested third parties. No.

23:46

This is just, like, you know, there

23:48

are some statutes that only a

23:51

public prosecutor can invoke and some

23:53

that the public can invoke. So imagine if

23:55

you went to the muffler shop and they wrecked your

23:57

car. And the only way you could sue

23:59

them is if you could get your local attorney general

24:02

to sue them. Right. That's what it's like when there's no private

24:04

right of action. So private right of action is if

24:06

you individually were harmed by

24:08

someone who violated a statute, you

24:10

can hire lawyer to sue them. statutes that

24:12

don't have a private right of action require public

24:14

prosecutor to bring action. And sometimes

24:16

that's appropriate, but I think, like with both privacy

24:19

and with the Access Act, a a

24:21

private right of action makes sense.

24:23

Yeah. I think the fear is that

24:25

it would jam the courts with

24:27

a a bunch of frivolous action as well.

24:29

Although, you know, don't see that in Texas. I

24:31

don't see it in California. So maybe And you can

24:33

you can just you can have something

24:36

like a slap act where -- Right. -- you can have early

24:38

motions to dismiss you can also have fee

24:40

recovery, which disc which discourages that

24:42

kind of thing. If you if it's loser

24:44

pays -- Right. -- people aren't gonna

24:46

bring spurious lawsuits because the other side will be

24:48

like, great. I'm just a hire a lawyer on contingency

24:51

to rack up giant billings for your frivolous

24:53

lawsuit. And at the end, I'll take it out of your hide. It's kinda

24:55

telling that we'd have to rewrite our tort system

24:57

in order for this to work. Well,

25:00

it's it's it's mostly illegal protection

25:02

and not a technical one. Right. Right.

25:05

As a technical matter, jailbreaking has

25:07

been of varying degrees of difficulty

25:09

at different times in in Apple's

25:11

device's history. But as

25:14

a legal matter, the the difficulty has

25:16

stayed constant.

25:17

the you

25:18

know, you have you have, like, CheckMate,

25:20

which is a a jailbreak

25:23

against all of the secure end plays for

25:25

eight years with iPhone models that cannot

25:27

be remediated because the secure enclave

25:29

is not field updateable because that's the whole

25:31

point. If you can modify the secure enclave,

25:34

then it's not secure. And

25:36

so, you know, hypothetically someone

25:38

could develop a jailbreak based third

25:40

party app store that leverage

25:43

Checkmate for anyone who bought an

25:45

iPhone over the first eight years of its existence

25:47

or, you know, year four through

25:49

twelve of its existence. but

25:51

they can't because Apple would come after you under

25:53

the DMCA. Is is that

25:56

so Matt Mullenweg, in fact, I wish

25:58

I'd asked him this because he was on our show couple of weeks

26:00

ago. about the porn

26:02

thing. There was a kind of a a movement

26:04

on Tumblr people saying, oh, Tumblr's bringing

26:06

porn back. Look at this. Look at this. Look at this. And Matt

26:08

had to write a blog post saying, no. No. not coming

26:10

back. It's never coming back. Is it

26:12

fair for Apple or for Matt to

26:15

blame Apple for this, Alex?

26:17

I mean, I look I

26:20

I'm not familiar with the with the controversy here.

26:22

I know that Tumblr used to be Philip Warren, I guess.

26:24

Yeah. Verizon killed it. Right.

26:26

Matt said, basically, that's the old tumbler.

26:29

Back in two thousand six, you could do that.

26:31

But nowadays because Apple would just knock us

26:33

out of the app store. And

26:36

that's forty percent of our sign ups and eighty

26:39

five percent of our page views from mobile.

26:41

We'd be out of business. Yeah.

26:44

I think But look, at

26:46

a certain point, you gotta let the companies that

26:48

are running these products make their own decisions.

26:50

Apple has a reason for not wanting to have Yeah.

26:52

But these point is they like Twitter and

26:55

Reddit, both Mhmm. -- have considerable

26:57

amount of adult content. Well, I do think that

26:59

were gonna wanna be consistent in the application

27:01

of the rules out there. He says they're too big for

27:03

Apple, the block. So they decided to make

27:05

an example out of Tumblr. If

27:08

that's the case, which I think is right. I mean,

27:10

I think Matt knows better than anybody and I like Matt

27:12

and I trust him, that's a

27:14

really good example of Apple misusing its

27:17

market power. you know, I think the one way

27:19

that Apple can really do better job is make sure

27:21

that it gets some of the scams out of the App

27:23

Store. You know, I I don't know if

27:25

you know, Apple's making the best use if

27:27

it's time being, like, morality, please, on

27:29

apps like Tumblr, and if it's going to be better,

27:31

be consistent. But there's so many scam apps

27:33

in the App Store These are well documented.

27:36

You know, they exist in the US and outside,

27:38

largely outside. And, you know,

27:40

if the company is trying to make us think that that

27:42

thirty percent fee is worth it, you

27:45

know, work on get, you know, those scams out

27:47

of the app store first, and then we can, you know,

27:49

go the level down and talk about, you know, decency

27:51

on on the apps. Yeah. And thanks to Costa

27:53

Alvario who

27:56

exposes this makes he's made this

27:58

his job, ever since Apple, blocked

28:02

his very useful

28:04

tool for writing text

28:06

on an Apple Watch, allowing

28:09

others through. He's made it his life work.

28:13

to to find scams on the App

28:15

Store. Apparently, he had a deal with he

28:17

made a deal with Apple over over

28:19

his apps being blocked,

28:22

so he's had to stop talking about that. But

28:24

Well, that's terrible. Yeah. So by the way,

28:26

can can I ask you a question? So I

28:28

think that there's been a very interesting arc

28:31

to these conversations that we've had about Big Tech.

28:33

The first one was a recognition. of

28:36

the fact that these apps have and and companies have

28:38

just become too big, and it all happened

28:40

so fast. Right? Facebook went from,

28:42

like, five hundred million users to a billion

28:45

to a couple billion users in a blink

28:47

of an eye. Amazon went from, you know,

28:49

being a percentage of of online retail

28:51

to being online retail effectively. in

28:54

a moment's notice Apple from a one trillion

28:56

or, you know, a couple hundred billion dollar company

28:58

to a three trillion dollar company in a moment's notice.

29:00

So then everyone's like, okay. These companies are

29:02

the biggest companies in the world, and

29:04

they're squeezing out competition, which is definitely

29:07

true. We need regulate them. And

29:09

and then we had this flood of ideas

29:11

and and bills that have come and tried to

29:14

reign them in. I

29:15

I just wonder if they're if they're going too far.

29:17

Some of them seem like they make a lot of sense

29:19

to me. Right? Like, the idea that a platform

29:21

cannot privilege its own products by

29:24

using the data that it gets from companies

29:26

that have to go through them. That makes sense. But

29:28

the whole idea of cutting off acquisitions,

29:30

okay, most M and A fails. So

29:33

the idea that M and A is the only thing that's made

29:35

these companies successful, doesn't doesn't really

29:37

make sense to me. The idea

29:39

of data portability, like, I I understand

29:42

that the tenants of data portability, but people are on

29:44

Facebook. people on Twitter for the network. It's

29:46

not like you can just take your data and go somewhere

29:48

else and be okay. And all this is happening

29:50

in the context of a market that's punishing

29:53

these companies Ruthlessly. I mean, Facebook's

29:55

down fifty seven percent this year in

29:57

the stock market and getting me capped by

29:59

TikTok.

29:59

So

30:00

I would just wonder if how

30:03

much how far if we've gone too far

30:05

on, let's regulate these the

30:07

fair competition back in the market without

30:10

remembering

30:10

that we have a market economy and

30:12

end letting

30:13

the market do its work as it seems to be doing

30:16

this here. I mean, I

30:18

I agree. I think that it's important to

30:20

distinguish between the dynamics

30:22

that drive growth and the dynamics that maintain

30:24

growth. So growth, I think, is driven

30:27

it's well understood by network effects. You

30:29

know, you joined Facebook because you wanted to talk

30:31

to people who are there. They joined Facebook because

30:33

they wanted to talk to you. You

30:35

made an app for the App Store because you wanted to

30:37

sell it to Apple customers. Apple customers

30:40

bought iPhones because they wanted to use your

30:42

app. And so there's this virtuous cycle that

30:44

that drives growth. But intrinsically, technology

30:46

has really low switching costs because

30:49

Computers are universal. The only computer we

30:51

know how to make is the computer that can

30:53

run all the software we know how to write. It's the,

30:55

you know, Turing machine. The von Neumann

30:57

machine And so, historically, you

31:00

know, when you had firms that had choke points in the

31:02

market the way Microsoft did in the early two thousands,

31:04

when it wouldn't maintain the

31:06

Mac office product and It became

31:08

harder and harder for CIOs to justify

31:11

having Macs in the office. was a CIO back

31:13

then. We started putting Windows machines

31:15

on designers desk so that they could access

31:17

Word files and Exile files and PowerPoint

31:19

files without corrupting them. Eventually,

31:21

we just put bigger graphics cards in them threw away

31:23

their Macs and and bought them Adobe Suite

31:26

for for Windows. And the way

31:28

Apple resolved that was not by

31:31

asking the the law to regulate Microsoft

31:34

nor was it by telling

31:37

people that they like Max Veterans, so they should

31:39

just hang in there. They reverse engineered

31:41

Microsoft products. They made pages,

31:44

keynote, and numbers. which are reverse

31:46

which reverse engineered the file formats of

31:48

Excel, PowerPoint, and Word.

31:51

And they made them feature compatible, and they

31:53

kept a team on that so that every time Microsoft

31:56

updated their file formats. Apple updated

31:58

their file formats in parallel

31:59

so that they could maintain compatibility.

32:02

that's a thing that has been ended.

32:05

Right? The

32:06

mechanisms under which we used to do that have now

32:09

been made illegal under purchase interference

32:11

with contract under non disclosure

32:13

and non compete under the computer fraud

32:15

and abuse act, section twelve one of the digital millennium

32:17

copyright write act and so on. We've created

32:20

this like a number of laws that boil down

32:22

to felony contempt of business model.

32:24

And what that's done is it's made it

32:26

possible for firms that have attained dominance through

32:28

network effects but would historically have

32:30

faced the risk of losing

32:32

that dominance also through network effects because

32:34

the corollary of a service that gets

32:36

more valuable every time joins someone

32:39

joins as a corollary as a service that gets

32:41

less valuable every time someone leaves. And

32:43

so you're you're prone to these like bank

32:45

runs on your users as we sort of see happening

32:47

with Facebook now. -- where people -- Yeah. -- even Facebook

32:50

and advertisers are leaving and then people leaving

32:52

and advertisers leaving. So, you

32:54

know, restoring that interoperability, the right

32:57

to interoperate when Facebook extended

33:00

membership to non EDU addresses, all

33:03

of the users that had hoped to court were already

33:05

on my space. And rather than telling them,

33:07

you know, you should pick a day when all of you quit

33:09

MySpace and come to Facebook or you should

33:11

maintain two separate clients, it gave

33:13

them a bot. and you could load that

33:15

bot with your login and password for myspace,

33:18

and it would go and scrape your waiting messages

33:20

and put them in your Facebook inbox. and then

33:22

it would like to reply to them, push them into your MySpace

33:24

Outbox. And, you know, if you try

33:26

to do that to Facebook today or if you try to

33:28

reverse engineer Apple's App Store, and

33:31

produce a feature compatible app store

33:33

today the way Apple did to Microsoft, they

33:35

would destroy you. And so you're

33:37

right, there's a market dynamic that drove

33:40

this growth, but it's not a market dynamic

33:42

that maintained the growth. The thing that maintained the

33:44

growth was the capture of regulation.

33:47

to prevent new firms from doing

33:49

to these firms what they did when they were new

33:51

firms. Well, I also

33:53

wonder Leo, you

33:55

can tell. You can ray me in, but No. No. No. This

33:57

is good. I'm I'm I'm and then I'm gonna bring up

34:00

some other congressional legislation

34:02

that is probably misguided, but go ahead.

34:05

If this is the case, then how do you explain

34:07

and by the way, you know, just just for the

34:09

I'm trying to learn here. So how do you explain

34:12

the notion that that Figma, which effectively

34:14

does what Adobe does, but does it on the

34:16

browser, just sold to

34:18

Adobe for twenty billion dollars. And what was

34:21

absolutely a defensive move

34:23

because Adobe knew that Figma

34:25

was gonna kick its butt if it let

34:27

it continue to to grow. And

34:30

another thing but by the way, I'm surprised,

34:32

but the consensus seems to be that that's gonna

34:34

be allowed to go through. When to me,

34:36

Corey, this is exactly what

34:39

you're talking about. But but Figma's ability

34:41

to succeed is effectively also like

34:43

pretty impressive. Right? Yeah. And then there's the

34:45

other like, with with this acquisition,

34:47

Adobe is gonna become fig effectively. The

34:50

other thing that I wonder about is what happens when

34:52

we, you know, we we we have these

34:54

conversations in the context

34:57

that, you know, nothing's gonna

34:59

change. But what happens when we move platforms? and

35:01

we go to, you know, augmented reality,

35:04

for instance. Like, the the

35:06

people who are building the operating systems in the

35:08

hardware for augmented reality right

35:10

now. You know, we don't we don't know who's gonna

35:12

win that. And that can just

35:15

like we move from desktop to mobile and

35:17

and you know, even, you know, downloaded

35:19

to to the cloud. You know, that

35:22

could also throw a few Well, this is a isn't this an

35:24

opportunity with met with the with the Metiverse

35:26

to say, let's do this differently. because otherwise,

35:28

you're gonna have Apple's metaverse. You're

35:30

gonna have a meta's metaverse. You're

35:33

gonna have maybe Microsoft's metaverse. and

35:35

you're gonna have to choose one of the other. One will

35:37

win over the life forever. Second

35:40

life and don't don't

35:42

laugh at me, but I think that The

35:44

metaverse is gonna be largely enterprise,

35:46

and there's this Facebook advertisement that

35:48

I think I've been right about soon. That's all about

35:50

enterprises use. of enterprise

35:53

uses of the metaverse. I'm I I

35:55

really believe that the metaverse is gonna be enterprise,

35:57

not social. And if it's Well,

35:59

I think suck is hoping it will be more than

36:02

just enterprise. But well, exactly. But you

36:04

look at their advertising and they're starting to Well, they

36:06

want they want it all. And Microsoft

36:08

clearly made their decision. They said, yeah. The

36:10

HoloLens isn't gonna be a consumer product. Exactly.

36:12

So but let let's throw another competitor in.

36:15

And again, this is where I'm hoping no one laughs,

36:17

but magic leap. You know, their second

36:20

generation device is not bad, really.

36:22

And it's heared entirely towards

36:24

enterprises. Yeah. So Like, I think capital

36:26

overhang, though. I mean, I think a lot of money

36:28

they owed to Virtis, but when you go to -- Yeah. --

36:30

ships little Voice openings for competitors.

36:33

Yeah. Sorry for that. I mean, I think

36:35

that the like, starting for the

36:37

end of working backwards, I think you're there's

36:40

a good case to be made that the metaverse

36:42

if it if it ever succeeds won't be

36:44

for entertainment. If for no other reason then

36:46

walking around with a brick on your faces an invitation

36:48

for someone to come up and kick you in the ass. So

36:51

I could see why it would only be used by people

36:53

who were, like, sitting comfortably. What about augmented

36:55

reality where you're I I maybe.

36:57

But I don't wanna get too caught up in there. I think

36:59

that you that the Figma story is really

37:02

interesting that that what you see is exactly

37:04

what I'm describing where there

37:06

are elements of of PSD that were

37:09

not which is the Photoshop document for

37:11

file format that were not within

37:13

this containment vessel. And so

37:15

Figma was able to make feature

37:17

compatible Photoshop replacement

37:21

that could read and write your Photoshop files That

37:23

was key. Right? Because people have a lot invested

37:26

in their existing Photoshop files. They

37:28

can't just abandon them. They need

37:30

to be able to open them and read them. PSD

37:32

was reverse engineered and that's why you can read it

37:34

in the Gimp and that's why you can read it in in other

37:36

programs and so on. But,

37:39

you know, their response was

37:41

to use their access to the capital

37:43

markets to snuff at a competitor before

37:46

it could grow to become a significant and

37:48

meaningful business all on its own.

37:50

And so, you know, that we're

37:53

we're left with this kind of tautology, which

37:55

is that if the capital markets will give you enough

37:57

money to buy a company, you must be

37:59

the best person to run it. And the way that we

38:01

can tell you're the best person to run it is you have enough

38:03

money to buy it. But going back to the Google

38:06

graveyard, it's pretty clear that

38:08

that you know, there are lots of people whom the

38:10

capital markets will entrust with the money to

38:12

buy a nascent competitor who

38:14

either, you know, treat that as predatory acquisition

38:17

and snuff it out. or who are just not

38:19

qualified to run it. We were making jokes about

38:21

the flicker URL at the beginning

38:23

of the show before we went on the air. you

38:25

know, Flicker was the first service that had

38:27

mobile photos. Right? Years before

38:30

Instagram, Yahoo bought it, and

38:32

then it became a play thing

38:34

that was dueled over among

38:36

the various, you know, venal princelings of

38:38

the Yahoo empire, and it suffered

38:41

and fell into, like, you know,

38:43

neglect and And now it's belongs

38:45

to smug mug who are gradually digging out

38:47

more than, you know, decades of of

38:49

technology debt, but it was, you know,

38:51

catastrophic. to to

38:54

be acquired. Yeah. But don't you

38:56

think that if a company gets acquired and then

38:58

ruined, that opens up the door for

39:00

another company to come and compete? So for instance,

39:03

you know, Figma is not the Adobe

39:05

isn't killing the cloud developer

39:08

design. You know, there's

39:10

a lot of competition. Yeah. Exactly.

39:12

which is an open source version. Just

39:14

raising, like, twenty million. Yeah. So exactly.

39:17

this is my point. So if Adobe kills Figma,

39:19

can't Penpott come in and start

39:21

to compete. Like you you, of course,

39:24

there is always the risk when you do

39:26

M and A that you're going to end up killing

39:28

innovation. but it's not the end of the story.

39:30

The story continues. If you mess something

39:32

up, you can oh, you are gonna

39:34

open it up to competition, And if

39:36

there is a need in the market for this thing

39:38

to be built, it will be built. Well,

39:41

but you get things like so

39:43

there's a great effluorescence of RSS readers.

39:45

at one point. And Google decided

39:47

to launch its own reader. And

39:49

that created what what venture

39:51

capitalists call the kill zone around

39:53

RSS readers and wanted to back them because

39:55

there was something that was priced at a you know, it's

39:57

cross subsidized from another business at a price

39:59

that was so low, free. that

40:02

there was no way to compete with it. And

40:04

as a consequence, we saw, you know, a decade

40:06

of neglect, which would have been fine

40:08

if Google had not then killed reader.

40:10

but readers never recovered. Right? That we

40:12

we we still you know, that was the end of RSS

40:15

effectively. It's it's now this kind of rump.

40:17

It was actually just in New York doing a book

40:19

event that Neelik Patel,

40:21

who's the editor in chief of the verge, was

40:25

was hosting. And the verge's new

40:27

redesign is amazing. and it one of

40:30

the cool things about it is it has a feed of

40:32

articles that people in the Virgil's newsroom

40:34

think are cool that's there on the

40:36

front door of the Virgil's new read as

40:39

And and I said, where can I

40:41

get an RSS for that? And he was like,

40:43

we discussed it and thought nobody would wanna

40:45

stop the RSS. That's exactly right. What

40:47

they're doing is an RSS feed. Can

40:49

I can I tell a funny story, though? So

40:52

I think that think there's a lot

40:54

of the idea of a kill zone is

40:56

is real and legit. I think there's a lot of truth

40:58

to what you said. Okay. maybe

41:01

I'm wrong here, but but I do think

41:03

that that we ended the tech world

41:05

ended up building a replacement for

41:08

RSS. and that was Twitter and

41:10

that the person who built Google reader

41:13

ended up going and working with Twitter

41:15

for a while. And he ended up unfortunately,

41:17

Bill and he regrets this building the retweet

41:19

button, which I think is a source of and

41:22

he thinks is a source of a lot

41:24

of the negative effects of social

41:26

media. So it's a mixed bag, but

41:28

it it also the market did come in

41:30

and say here is another way

41:33

that you can get, you know, your your

41:35

stories via a feed, and that was Twitter.

41:37

Reddit also has I I have to, by

41:39

the way, I think, Neil, I answer

41:41

to you, Corey is disingenuous. I

41:44

don't think that has you know, how much

41:46

does it cost to do in RSS feed? It's nothing.

41:48

You can Well, this is too far. I think she was telling

41:50

the truth. I think I think it was just basically,

41:53

like, they they built a lot of features

41:55

It's you know, they're not an engineering organization.

41:58

They had to prioritize it. But Niela knows.

41:59

He knows better than that. And

42:02

then they're Yeah. Remember,

42:04

the the verge launched to to be --

42:06

Right. -- reference customer for its own

42:09

CMS. Right. And so we like,

42:11

I have no idea. Like, if we were talking about Droupe,

42:13

I'd say, yeah, for sure. You just, like, you

42:15

know, type in the obscure search map. Yeah.

42:17

says make this thing be an RSS feed

42:19

now. Right? By the way,

42:22

the guy when

42:25

Google killed reader, there were so many

42:27

clones that came out. The market did try to

42:29

end up. you know,

42:31

building every placement. I remember, I

42:33

I'll be honest, I wasn't really into our assess

42:35

readers. And then Google killed reader

42:37

and I saw the outpouring of all this English.

42:40

and said, oh, I gotta try that. And so

42:42

I downloaded or or set up an account

42:44

on something called the old reader. And

42:47

I'm I gotta that was I

42:49

I still use RSS. I use Sumi dot

42:51

news. And by the way, I gotta point out

42:53

Sumi discovered an RSS feed at the

42:55

verge. So it's not that

42:57

feed on the column there on the right, but it's

43:00

the all posts feed from the verge.

43:02

So -- Alright. -- they're still doing

43:04

RSS. Maybe Neil, I didn't know

43:06

that. No. No.

43:08

No. They put RSS for the main feed. For the main

43:10

feed. Not for that side feed. Yeah. For the not

43:13

for that side bar. That side bar is amazing.

43:15

It's just it's just not convenient for me to keep

43:17

a browser to open some of my screen. It's just

43:19

Yeah.

43:20

You know?

43:22

But but III don't think he was be asking

43:24

me. I think they just sat down and did a triage and

43:26

they were like, nobody uses to RSS anymore.

43:29

Why will we bother? And I tried to convince

43:31

them that RSS is like, what

43:33

the people who read the news to talk about the news

43:35

and amplify the news. Yes. I use. Yep.

43:37

And so, you know, it's like it's a we're

43:39

we're a small but proud people, you know?

43:43

Like an influential, I think influential.

43:45

Like Canadians. Yeah. Like Canadians.

43:49

He seemed like such a nice guy who would have

43:51

known. We're gonna take a person that's

43:53

gonna have to take a break. Hold on, Alex. because

43:55

I we k. Otherwise, we'll be here

43:57

little eight or nine or ten. in on

43:59

your time that's, like, three in the morning. So

44:02

Alex Kenterowitz is here. The big tech big

44:04

technology podcast technology dot subset

44:07

dot com I didn't think of it this way, didn't

44:09

intend it this way, but you're

44:11

not here to defend big techs, so don't feel like

44:13

you have to. especially

44:15

when you get Corey Doctor Rowan, the author

44:18

of the latest chokepoint capitalism,

44:20

which is absolutely an indictment of not

44:22

just big tech, but big business in

44:25

general. You point out in the book and I think

44:27

it's, by the way, a great read, highly recommended.

44:30

the that

44:31

it's not just tech. It's

44:33

it's publishing. It's the

44:35

record in business. It's the live

44:37

concert business. It's every business that is out there.

44:39

In fact, we got Story will talk about

44:42

a little bit about the podcasting business

44:44

as well is becoming a

44:46

big tech. Not I'm not a part of that, unfortunately.

44:49

I want more just a butt a bit our show today

44:51

brought to you by Nueva. Nueva's

44:55

great audio simplifies today's IT

44:57

pros. Yeah. It can't have been a tough spot.

44:59

we're we're now we're doing this hybrid thing. Right?

45:01

Hybrid working and learning means you

45:04

have to equip and support more spaces

45:06

with audio and video conferencing systems.

45:09

because you've got people in the office, you've got people

45:11

elsewhere. They are all

45:13

working together. and,

45:16

you know, your IT department still got to do network

45:18

security, which is a big issue these days. And, of

45:20

course, you're shifting to cloud based solutions. You

45:23

got infrastructure issues. Product

45:25

shortages and delays add that on top.

45:27

It's put so much strain in IT's resources,

45:30

people time, expertise, and budgets.

45:33

So here's a product. Customers

45:35

will love that requires minimal

45:38

effort from IT to deploy and manage its scale

45:40

with a bonus of requiring zero end

45:42

user training, but it

45:44

solves a big problem in hybrid

45:47

work. When it comes to audio

45:49

conferencing in larger spaces, it's very

45:51

common to be faced with complex

45:53

and expensive multi component systems.

45:55

You gotta bring in people to design them and

45:57

install them. And then they've gotta continue to

45:59

maintain them and update them and manage them. Noreva

46:02

is the opposite. Great

46:05

audio simplifies as changed by

46:07

offering solutions that deliver

46:09

a high level of simplicity with Noreva

46:11

you you put in what is looks like a sound

46:13

bar that gives you

46:16

full room mic pickup. If you got a really

46:18

big room put in a couple, if you got regular sized

46:20

room put in one, It's a a soundbar, it's

46:22

got speakers, it's got microphones, and it

46:24

uses Nextiva's patented microphone

46:27

mist technology to in effect fill

46:30

the space with

46:31

microphones. And

46:32

so easy to install, you could do it. You don't even

46:35

have to call the IT department, install the RAVA system

46:37

in less than thirty minutes.

46:39

maybe an hour, if you've got giant room,

46:41

it's simple. You don't have to configure

46:43

it. You don't have to tweak

46:46

it. You could it's so much

46:48

less expensive and so much easier than

46:51

those traditional systems that could take your room

46:53

offline for days. In fact, you could start

46:55

right now Get an array of system, install

46:57

it, be ready for your very next meeting.

47:00

With an array of you can monitor, manage, update,

47:02

and adjust all the systems. no

47:05

matter how many rooms with a powerful cloud

47:07

based platform to raise a console, IT

47:10

doesn't even have to walk down the hall. NARABA

47:13

is very scalable. They could bring

47:15

their simplicity to large organizations too.NARABA

47:17

systems, NUR EVA, They

47:20

cost a fraction of traditional systems give you

47:22

a better result, solves

47:24

this problem of people in the office and people

47:26

elsewhere trying to meet and work

47:28

together. Great audio is so important

47:30

to that. Check it out noreva dot com

47:32

slash tweet and you are EVA

47:35

noreva dot com slash

47:38

tweet. This is solution. I don't think people I

47:40

mean, I guess if you've been watching our shows, you know

47:42

about it, but I I wish it were better known because

47:44

I still see conference

47:45

room so poorly equipped for audio

47:48

as he could do so much better. No Rava,

47:50

dot com slash Twitter. Thank you,

47:52

Narayva, for supporting this week

47:54

in tech, and thank you listeners

47:57

and viewers for supporting us. by

47:59

going to that

47:59

special address. So I know you saw it here. sharva

48:02

dot

48:03

com

48:04

slash twitch.

48:06

Did

48:11

Mark Zuckerberg fight in the UFC show at

48:13

all? It should have.

48:16

So I don't Saturday,

48:19

I don't know what happened. I mean, I'm hoping somebody

48:21

subscribed to this. UFC

48:23

closed its a fight card at the apex

48:25

of facility owns in Vegas to the press in the

48:27

public. They wouldn't

48:29

say initially who did it, but an

48:31

MMA insider, Aria Hawani, said,

48:34

A very good close close to the event had told

48:36

him it's something to do with Mark Zuckerberg.

48:39

Speculation is Mark's written out the

48:41

event, maybe just so they could

48:43

watch. maybe

48:46

they could record it for the metaverse. Maybe

48:48

Mark's gonna get into the Octagon. I don't

48:50

know. Wouldn't put it past him.

48:52

Oh, they have been already happened?

48:55

Yeah. So I'm I'm asking, what happened?

48:57

It was just Marcus Hooker. I feel like it was Marcus

48:59

Hooker Berginis. Great news. cronies.

49:01

watching. smoke in those mics.

49:04

And there's this amazing reaction,

49:07

Jeff, of his wife, Priscilla. Just

49:09

kind of losing her mind as the

49:11

fight goes on. But I think the more telling thing

49:13

is that there were all the UFC fighters who

49:16

were talking about how off foot was that

49:18

the entire fight could be bought out by one person.

49:21

And, yeah, I think that that that spot

49:23

on. And I don't know what Marillyn took a

49:25

break. You got enough money what could I can't

49:27

you but Yeah. Of course. But just the a, the optics,

49:29

BD Act itself, I find, like, fairly

49:32

fairly wrong. And, you know, sports is a

49:34

game for the people, you know, I I

49:36

just, you know, I can't really

49:38

see any justification for

49:40

wanting to, you know, buy, you

49:42

know, the entire seating and then

49:44

lead the public out of it. If you

49:47

wanna go watch if I go watch your fight,

49:49

but don't buy out. Has anyone else ever

49:51

done? I've never seen it. How much do you think it cost?

49:53

to buy and get started to gun. I

49:55

don't know. You probably I guess, you probably do it for a million

49:57

dollars. They make most of they make most of their

49:59

money on the on the cable

50:02

for the streams. Yeah. Yeah. Yeah. They've reviewed.

50:04

They I mean, it's like getting Kanye to

50:06

play or your corporate event or your kids

50:08

buttress -- Yes. -- or or, you know, Kincinera

50:10

or something. It's just such a

50:12

it's such an oligarch. It

50:15

really is. Yeah. So, you

50:17

know, that I was told that

50:19

when the sultan of Brunei visited

50:21

Disneyland, he had thirteen tour

50:23

guides and the vice president of operations with

50:25

a retinue of over a hundred people and a flying

50:28

wedge. ahead of them.

50:30

And they just went from right to right

50:32

place. during parts of Disneyland as they went,

50:34

which was hilarious. It's a military to,

50:36

like, ten grand a day. It

50:39

you know what? It'd be worth it not to have to get

50:41

in line. That's all I'm saying. Absolutely. Yeah.

50:44

That's best deal Disneyland. you see every

50:46

once in a while, you see videos on insta

50:48

and and TikTok of celebrities

50:51

getting ushered to the front

50:53

of the line, you know. You just have to, you know,

50:55

It's good business. I

50:58

did like how David Beckham waited the

51:00

full what was it? Six He waited for

51:02

the queen to go. That's blessing. Yeah.

51:04

God bless him. Ten hours. You know? Yeah.

51:06

I think a number of celebrities actually got

51:09

in line for that. That's respectful. I like that. That

51:11

is respectful. William

51:13

Gibson once

51:14

told me that he feels like he has just the

51:16

right amount of fame. You know, like,

51:19

what is the right amount of fame? People

51:21

enjoy his work and they tell him so

51:23

he can earn a living from it, but he it's not

51:25

like he can't eat dinner in a restaurant. Right?

51:28

I I once Angela

51:30

was visiting London and we went out for lunch

51:33

and, like, got interrupted, like, three times.

51:35

No fun. is a moderately famous

51:37

person. But, you know, we are eating at

51:39

a restaurant in London. It's not like Pen and Penantell

51:42

are really well known in the UK. And

51:44

nevertheless, like, he couldn't get through a forty minute

51:46

meal without being interrupted three times. So,

51:48

you know, I I feel like there's probably

51:51

it's there's there's definitely a threshold where the

51:53

fame gets pretty toxic. Steve

51:55

Martin once told me that he knew it

51:57

was over for him as a real person

51:59

when he attempted to ride this

52:02

subway, this was many, many years

52:04

ago, to go see a show in

52:06

Brooklyn or Queens are

52:08

somewhere. And he said, I can never he said on

52:10

that. I can't ride. I can't do it. And

52:12

I've been to dinner with him. He rents, basically,

52:14

you take over a private room. You don't eat in public.

52:17

And you're still harassed by the chef, made

52:19

her tea. You know, how how do you talk

52:22

about this relationship you have with Steve? Oh, it's

52:24

well known. I've told the I've told people about this

52:26

before. he he used to listen to my radio, so I

52:28

don't think he does anything. Mhmm. And he he

52:30

d m to me about ten years

52:32

ago saying, you don't don't have to respond

52:34

to this, but I really like your radio show. I said, yeah,

52:36

I'm not gonna respond. Steve Martin, who cares?

52:39

No. I responded and he kinda

52:41

struck up friendship. What's that?

52:43

friendship. Yeah. And it

52:46

is it he's the only person who's that

52:48

famous that I've ever spent time with. And it really

52:50

it's it's he can eat. I bet

52:52

now he can't. Now the murders in the building's big.

52:55

I bet you it's actually gotten worse from him again.

52:58

But he very famously stopped doing stadium

53:01

comedy because he said, it's not it's

53:03

not a show anymore. It's -- Mhmm. --

53:06

it's like the zoo. It's like, you know, he

53:08

talks about it and born standing up in his his

53:10

book. He did I think

53:12

he really didn't like that level of fame. But once

53:14

you get there, you can't there's no turning back.

53:17

Here's Mark, by the way, practicing.

53:20

And apparently, this

53:23

is his sparring partner made his debut

53:25

in the knocked a gun, and maybe that's why

53:27

he he

53:27

bought it out. That way,

53:28

he could just buy a ticket to it. A knock Okay.

53:30

Just buy one ticket. I can't see those.

53:33

But, you know, can I just say sorry. Go

53:35

ahead. If you're Mark Zuckerberg, you don't wanna be

53:37

harassed, the whole fight. You know,

53:40

maybe you need that flying wedge of Chronie's,

53:42

and I think that Yeah. He should

53:44

just you wanna go watch sports, go

53:46

go suck it up, and go watch it in person. to

53:49

people. III mean, it's not like they don't have

53:51

booths. Right? Like they have sky

53:53

booths. Sex leach or sex leach

53:56

or private booths. Yeah. Yeah. I

53:58

mean, the champagne is just a solve

54:00

problem. Right? Like, this

54:02

is the like, a famously solved

54:04

problem. Right? Of of sports stadiums

54:07

having VIPs who wanna sit

54:09

in a fancy booth. Wear a mustache

54:11

of it, you know, a hat. just a

54:13

little disguise. You can bring it up the

54:15

private room. Yeah. It's Yeah.

54:18

That exact III don't know. I mean, it says

54:20

it says this c thing, but I

54:23

mean, who says, yes, that's a good idea.

54:26

To to the end, you're getting bad advice

54:28

if you think that's a good idea to do this. Maybe not

54:30

for Mark. But remember, this is also the guy who,

54:32

on the fourth of July last year, posted

54:34

a picture of him wearing sunscreen

54:38

holding the American flag on a motorized

54:41

foot, you know, Leo, that III

54:43

have less of a problem with. If you wanna go make

54:45

an idiot of yourself in the ocean, By all

54:47

means, people do it every day. I

54:49

remember when he posted this that we had somebody

54:51

on and said, you know, you

54:53

don't if you're Mark Zuckerberg, you have a

54:56

fey lengths of PR professionals guiding

54:59

you, protecting your image at all times.

55:02

How did that didn't they get

55:04

a bad saying, you know, like like,

55:06

no one goes, hey, you might look a little out of touch

55:08

if you end up, you know, buying out the whole UFC

55:11

game. And I I love the fact that all the UFC

55:13

fighters called, you know, BS on it because

55:15

deservedly so. No.

55:18

That's That okay. No problem with that. If you wanna

55:20

go pull the flag, make any of it

55:22

out yourself in the ocean. Like, have

55:24

that there's a big ocean, but there's

55:26

only limited seats to the UFC. people

55:29

were offended by this though. I do remember

55:31

they were very offended. I'm not sure why. People are

55:33

offended. People are offended by everything.

55:35

So Yeah. It's

55:37

true. It's

55:41

not it's not dignified. It's not

55:43

dignified. Maybe that's that's all that's wrong

55:45

with that. Look,

55:47

is is it is it something I would do?

55:50

No. No. Whatever. Whatever.

55:52

We're taking Caesar. robots. a sports

55:54

event on that on that one. You know what's

55:56

fun? We are now, thanks

55:59

to the ongoing action between

56:01

Twitter and Elon Musk in the Delaware

56:03

Court of chancery, privy

56:05

to the fascinating texts

56:09

going back and forth between

56:12

Elon and other wealthy

56:14

individuals during

56:17

his attempt to purchase

56:19

Twitter. And

56:20

I like the I think it was the Atlantic's take

56:22

on it, but it just shows you these guys

56:24

are is stupid. as

56:28

you might even imagine, they Elon

56:30

Musk's texts shatter

56:32

the myth Of the text

56:35

tech genius. This is Charlie Warzel, writing

56:37

the world's richest man has some embarrassing

56:40

friends. number

56:42

of whom have been on this show, including Jason

56:44

Callahanis, volunteered

56:50

to run Twitter for

56:53

Elon. And then at at one point

56:57

said, I you know, let me ask my you

56:59

know, I'll ask around. remember he did this. He asked

57:01

around and said, hey, may wanna invest, you know,

57:03

put some money into Musk's acquisition.

57:06

I can I can help.

57:08

Marshall

57:10

writes few and Musk's phone appeared as

57:12

excitable as the Angel Investor, Jason Kalek,

57:14

and his who peppered his friend

57:16

with flattery and random ideas for

57:18

the service. In the span of thirty

57:21

minutes, Calican has suggested

57:23

a five point plan for Twitter that would

57:25

introduce a membership tier, creator revenue splits

57:27

algorithmic transparency and changes to the

57:29

company's operation after pledging his

57:31

loyalty. You have my sword,

57:34

he texted Musk. Callahan

57:36

has pushed new ideas for weeks.

57:40

For weeks, imagine we

57:42

asked Justin Bieber not

57:44

beaver beaver to come back and let him

57:46

DM his fans. He could sell

57:48

a million and merchandiser tickets instantly would

57:50

be insane.

57:54

Finally, Musk says, sends

57:56

a message back.

57:59

Morgan

57:59

Stanley and Jared think you are using

58:02

our friendship in a good way. This makes it

58:04

seem like I'm desperate. Please

58:06

stop. Which

58:09

poor old Jason. only

58:11

ever want to support you. Elon,

58:14

love you, man. And he said he'd

58:16

jump on a grande for him. Yeah. So

58:20

I wrote about this today. I just put the link in the IRC.

58:23

I wrote about this today for my column on medium.

58:25

And I think that, you know, the

58:27

way to understand how this

58:29

works is that it's,

58:33

you know, to be an innovator is

58:35

not to have unique genius, it's to have

58:37

good timing. that

58:39

that, you know, if you scroll down a little there,

58:41

Leo, you'll see these two diagrams I

58:44

have about what it takes to invent the helicopter.

58:46

So, you know, for five hundred years, people

58:49

invented helicopters. Right? They they were

58:51

like, oh, I've seen a screw press and I've seen a maple

58:53

key. I've invented the helicopter, but

58:55

it wasn't until someone else had invented

58:57

you know, over now, internal combustion

59:01

-- Yeah. -- that you could you could get a helicopter.

59:03

And, you know, Kevin Kelly calls us the adjacent

59:06

possible. And and this is why,

59:08

like, when it's railroading time you get railroads and

59:10

why, like, six people invented the radio within,

59:12

you know, a year of each other and so on.

59:14

an idea whose time has come. That's what

59:16

that means. Right? Yeah. Yeah. And so these

59:18

guys, you know, they had a they had a good

59:20

idea. They were not unique in having that good

59:23

idea. But what they did

59:25

have was access to the capital markets after

59:27

after edging out other people, after getting a little

59:29

bit of advantage that they could use

59:31

to you know, buy other other

59:33

people's good ideas to suppress

59:36

good ideas before they could take hold independently

59:38

by by buying our rivals or by

59:40

using predatory pricing. all of that

59:42

other stuff. And so what you end up with is people

59:44

who are just sort of mediocre donkeys,

59:47

no better than you or me,

59:49

I'm not claiming to be better than any of these

59:51

people, but I they're, you know, they're not better than

59:53

me either, and no one should put me in

59:55

charge of the lives of a hundred million Twitter users

59:57

or three billion Facebook users. years.

59:59

And and, you know, like the I think

1:00:01

we often focus too much on

1:00:04

whether these people have the right stuff for that

1:00:06

job and not enough on whether that job should

1:00:08

test. We have and I've talked about this

1:00:11

before. There's this great man hypothesis. And

1:00:14

because Elon is a billionaire or

1:00:16

Mark Cuban is a billionaire, we

1:00:19

ascribe to them, you know, they become

1:00:22

the Alexander the greats. They become the great

1:00:24

man. When,

1:00:25

in fact, maybe they didn't earn that

1:00:27

sobricay. Maybe they maybe they're just

1:00:29

the right person at the right time. That's what you're saying.

1:00:32

It's it's the providential doctrine. Right?

1:00:34

If you are rich, you must be great. And if you are

1:00:36

if you are not rich. You aren't great. Right. And

1:00:38

the way to tell whether there's someone as great as whether they're

1:00:41

rich. I think we've come to a time when

1:00:43

we've stopped worshiping billionaires or we're

1:00:45

starting to stop working billionaires. I hope. You

1:00:47

say that. I I wrote something

1:00:49

unflattering about Palantir yesterday,

1:00:51

and it turns out that Palantir is the latest

1:00:53

meme stock And there's a whole bunch

1:00:55

of weird Peter Teals. He's a short

1:00:57

seller, bud. They're in my mentions.

1:00:59

Yeah. Who is paying you to write this? Oh,

1:01:01

my and so on and so on.

1:01:04

Palantir should absolutely run the NHS.

1:01:06

I've actually I did a thread, you

1:01:08

know, post a link in there. I did thread. Palantir tier

1:01:11

should run the National Health Service of

1:01:13

Britain. Yeah. So that's the thing is Palantir is

1:01:15

so Palantir wanna they didn't

1:01:17

win. Palantir got a no bid twenty

1:01:19

six million pound contract to do work for

1:01:21

the NHS, and they're trying to leverage it to

1:01:24

this big three hundred and sixty odd million,

1:01:26

but twenty three and twenty million pound

1:01:28

contract for the NHS. And

1:01:31

they're pretty clear that no one is gonna no

1:01:34

one is gonna green light that because they're

1:01:36

palantir. And so their new strategy

1:01:38

is they're buying all the companies that have

1:01:41

NHS contracts. So this is one

1:01:43

of this is kinda maybe a little like Amazon's

1:01:46

planned to buy -- Yeah. -- one medical and

1:01:49

well, that's really interesting. So p so for

1:01:51

people who don't know, I'm sure most people know Palantir.

1:01:53

is is basically a surveillance

1:01:56

system. Much like the Palantir in

1:01:58

the Lord of the Rings, a

1:02:01

an eye of Sauron that collects data

1:02:04

and then sells it to who governments, law

1:02:06

enforcement. Are they

1:02:08

a data broker? Is that I mean, are they or is

1:02:10

it that to Denmark? They're an analytics

1:02:13

platform. They're I mean, they do a lot of, like,

1:02:15

turnkey, you know,

1:02:17

human rights abuses as a service. So,

1:02:20

like, if you wanna figure out how to do

1:02:22

how to do algorithmic racism with

1:02:24

refugees -- Yeah. --

1:02:27

you know, you can you can buy their service and feed

1:02:29

it into them and their their, you know, phrenology

1:02:31

robot will tell you that all the bad refugees

1:02:33

are brown. It strikes me that

1:02:36

that letting some a company like that

1:02:38

own or not own, obviously, that

1:02:40

we could own the NHS, but participate in

1:02:42

any way with a service

1:02:44

that has the health records of millions

1:02:46

of Britons seems like a bad

1:02:48

idea. Terrible. Not at least

1:02:50

because there's actually a really good

1:02:53

Sorry, go ahead. Now

1:02:55

you go ahead. I was just gonna say there's

1:02:57

a there's an amazing proposal in the offing,

1:02:59

Ben Goldacre, who's an evidence based medicine

1:03:01

specialist, and has done a bunch of important

1:03:04

interventions at the kind of

1:03:06

national level in the British healthcare

1:03:08

system like the register of all trials and stuff

1:03:10

that have absolutely revolutionized evidence

1:03:12

based medicine in the UK, did this

1:03:14

thing called the gold acre report, and

1:03:17

it's how you would build a research platform

1:03:19

that allow you to stacked insights from

1:03:21

the collected health records of

1:03:24

NHS service users without violating

1:03:26

their privacy. And he's like, First, you build

1:03:28

an open platform that anyone can

1:03:30

audit, anyone can use, and anyone can implement,

1:03:33

and then you host that platform and

1:03:35

close it off from everyone else, with

1:03:37

the NHS data in it and you tell

1:03:39

researchers like Send me a query

1:03:41

and I will run it on the platform and send

1:03:44

you back the data but you can't ever

1:03:46

see that data. It's owned by the public, managed by

1:03:48

the public. No. No. No. It's really good. No. No. This is

1:03:50

really good idea. This is this is

1:03:52

a this is a good like,

1:03:54

an auditable evidence to drive

1:03:56

it. Public service. This reminds me

1:03:59

of IMDB where you get everybody to create

1:04:01

great database and But it's owned by the public.

1:04:03

Oh, okay. NHS was on it. Alright. So now it's yours

1:04:05

as a on it. service. NHS owns it. Okay. Yeah. And

1:04:07

they wouldn't they wouldn't get, like, Arthur Anderson built

1:04:09

it for them or someone, you know, PricewaterhouseCooper built

1:04:11

it for them, they would have to make it open. They'd have to put the

1:04:13

code on GitHub. Right. Anyone can see this code.

1:04:15

Anyone can audit this code. but the service

1:04:18

itself, the only person people

1:04:20

with a login for it would be the people who

1:04:22

ran research for the NHS. This is the

1:04:24

only way you can keep it anonymous. in

1:04:26

effect? Well, and and and and

1:04:30

but also productive. Right? So if you're a private

1:04:32

researcher at a university or pharma company,

1:04:34

anywhere else, and you're like, I wanna know,

1:04:37

what happens when these two interventions

1:04:39

are paired? What does the data tell us? You gotta give someone

1:04:41

this drug when they're doing this this

1:04:43

PT physiotherapy -- It's huge value. -- something

1:04:45

else. Huge value. So you you can then

1:04:47

get back useful data,

1:04:50

useful conclusions but you never

1:04:52

handle the data. So the goal data reports

1:04:54

really good and it's such a it's the

1:04:56

opposite of the Palantir approach, which

1:04:58

is like we will you

1:05:01

give us your data, we'll apply our

1:05:03

magic, proprietary stuff that no one

1:05:05

is allowed to know about, and then we'll tell

1:05:07

you what's in your data. We'll tell you what you

1:05:09

know, we'll tell the LAPD where

1:05:11

to go and do stop and

1:05:13

frisks, which is, you know, like,

1:05:15

by an incredible coincidence that none of us

1:05:17

could have predicted neighborhoods where a lot of brown

1:05:19

people live. Yeah. Is

1:05:22

this what Amazon's doing, Alex?

1:05:24

Is this why Amazon's acquiring, you

1:05:26

know, one medical and others? I

1:05:28

mean, I just wanna make a comment on the Palantir.

1:05:30

I think that we need to understand that this is

1:05:32

a consulting company, not a data

1:05:35

company. And, you know, they're

1:05:37

they're more in line with the Deloitte than they are

1:05:39

in line with any, like, cloud services

1:05:41

provider. It's the analytics that they they

1:05:44

sell. Well, they sell the consulting.

1:05:46

You know, they talk about the analytics. They're

1:05:48

they're sort of dressed up consulting company.

1:05:50

And my hat take is that we're

1:05:52

gonna end up seeing that. You know, they might be

1:05:55

a meme stock today, but

1:05:57

we will see the company end up in over

1:05:59

time, you know, being worth

1:06:01

you know, what they should be worth, which is a lot less.

1:06:03

And I think their worth is largely inflated through

1:06:05

through to their government connections that they

1:06:07

have actually in the Elon texts. There was something really

1:06:09

interesting about Joe Lonsdale, Micky,

1:06:12

who is the Palantir cofounder, you know,

1:06:14

hinking out with, you know, a hundred

1:06:16

Republican congressman. We all know about

1:06:18

Peter Teel. So, you know, those political

1:06:21

connections help a lot. But over time,

1:06:23

I think we're gonna see them for what they are, which is

1:06:25

that they we should address them up in this, like,

1:06:27

you know, putting their special sauce

1:06:29

on the analytics. Like Oh, that's interesting. I

1:06:31

mean, at the end of the day. That's actually Yeah. I'm

1:06:33

not saying it's good. I I'm not saying that

1:06:35

it's useful insights, but I'm saying that, like,

1:06:37

I think that there are people who will sell

1:06:40

you confirmation bias as a service. Yeah. Right?

1:06:42

That's right. I have a thing. I wanted I have I

1:06:44

I need some policy

1:06:46

based evidence, please. Right? So I know what

1:06:48

I wanna do. Please produce the insights

1:06:51

that will let me prove to my board or

1:06:53

to other people or the market or whatever that

1:06:55

what I wanna use. Yeah. Yeah. Yeah.

1:06:58

Well, that's interesting. I hadn't really thought about it.

1:07:00

I was giving Palantir all of this superpower

1:07:02

to see into our lives. And it's

1:07:04

maybe the it's like Cambridge Analytica. They're

1:07:07

Yeah. It's it's exactly the analog that I

1:07:09

was named. Yeah. Yeah. Yeah. And I also think

1:07:11

yeah. Anyway, I'll I'll leave it there.

1:07:14

Why are they buying up well taken? Why

1:07:16

are they buying up these

1:07:18

suppliers to the NHS then? What's

1:07:22

their goal? pound tier? Republic they're

1:07:24

yeah. They're they're a looter. Right? They just

1:07:26

wanna they wanna suck up bunch of public

1:07:28

private partnership money. Oh, okay.

1:07:32

because that's again that's why Willie Sutton

1:07:34

and Robert Banks. Right? because that's where the private public

1:07:36

-- That's where the money money is. -- if you if you

1:07:38

don't have the software that's gonna take you there,

1:07:40

you need the partnerships because

1:07:42

you're a consulting company. That's boy,

1:07:45

thank you for that. You know, I've I've misinterpreted

1:07:47

really kind of given them a lot more credit than

1:07:50

they deserve all this time.

1:07:52

I just wanna shout out the people

1:07:54

in the chat saying I'm here to defend Big Tech.

1:07:56

But let's look at it reasonably.

1:07:59

Right? And the more reasonable our conversation,

1:08:01

the more progress we're gonna make. I think what I

1:08:03

really admire actually about Corey's

1:08:06

book is that it isn't a it it

1:08:08

doesn't paint a target on big techs back.

1:08:10

It paints a target on companies that have gotten

1:08:13

too big. basically, with

1:08:15

predatory techno predatory methods.

1:08:18

Yes? Is that fair? Yeah. I would

1:08:20

say that that that that what it does is it

1:08:22

identifies the

1:08:25

source of the pain that artists are feeling.

1:08:27

Right? So we spent forty years

1:08:29

giving artists longer duration of

1:08:31

copyright easier in force copyright,

1:08:34

stiffer penalties for copyright, and an

1:08:36

increase scope of copyright, the

1:08:38

entertainment industry has grown, and

1:08:40

it's gotten bigger, the share of income

1:08:43

accruing to the workers who produce the materials

1:08:45

that produce those profits has

1:08:48

shrunk over the same time. And

1:08:50

the reason is that these firms have

1:08:53

between them created checkpoints where if

1:08:55

you wanna reach your audience, you

1:08:57

have to negotiate with them

1:09:00

And and typically, if there's another firm

1:09:02

you can negotiate, it's another firm that has nearly

1:09:04

identical or actually identical terms.

1:09:06

and they've all converged on a set of negotiating

1:09:09

terms that say whatever copyright you've

1:09:11

been given, you have to hand over to us.

1:09:13

And so giving creators more copyright

1:09:15

won't help in the same way that giving

1:09:17

your bullied kid more lunch money won't help.

1:09:19

Not even if the bullies are like running a national

1:09:22

campaign saying, won't someone think of America's

1:09:24

hungry children, give them more lunch money.

1:09:26

Right? They're they're just gonna take whatever lunch money

1:09:28

you give them. And that the and that

1:09:30

you know, that's the first half of the book. It's just

1:09:32

showing how these monopsinistic.

1:09:35

It's when there's a small number of buyers who

1:09:37

can control their sellers. with this an

1:09:39

opportunistic dynamic works and how it enables

1:09:42

tactics that range from simply

1:09:45

unethical to just illegal, but no

1:09:47

one can take them on even though they are illegal. We

1:09:49

we document multiple hundreds of millions of

1:09:51

dollars in Wage Theft from Audible

1:09:53

creators by Amazon. But

1:09:56

the second half of the book are interventions that

1:09:58

aren't just more copyright that

1:10:01

actually do widen out these choke points and,

1:10:03

you know, we talked about one on triangulation that I

1:10:05

really like, which is that if

1:10:07

you audit your royalty statements, which, generally,

1:10:09

you're contractually allowed

1:10:12

to do if you audit your royalty statements. In

1:10:15

order to get the money that you find

1:10:17

is owed to you, the firm

1:10:19

will generally say either you have

1:10:21

to sue us or if you want a voluntary

1:10:24

settlement you have to submit to non disclosure.

1:10:26

And so we cite research from

1:10:29

firm that specializes in auditing, recording

1:10:31

industry contracts. They've done tens of thousands.

1:10:34

In all but one instance, over decades,

1:10:37

every accounting area they located was

1:10:39

in favor of the label, not the artist. This

1:10:41

is an amazing coincidence. As always say, it's this

1:10:44

the most incredible localized probability storm

1:10:46

you can imagine. There's no other possible

1:10:48

explanation for wildest pounding errors, but

1:10:50

would just favor the the company making the royalty

1:10:53

statement. But if you go and you find missing money

1:10:55

and we have a source that found a six figure

1:10:57

error in their favor when they audited their royalty

1:10:59

statement, You

1:11:01

have to agree not to tell anyone else who's

1:11:03

being ripped off in the same way where

1:11:05

they should go and look for the money that's being stolen

1:11:07

from them. and there's an actual fix for

1:11:10

this which is relatively straightforward. Because

1:11:12

the industry is so concentrated, all

1:11:15

of its contracts are consummated in New

1:11:17

York, California, and because of Amazon Washington

1:11:19

State, contract being a matter of

1:11:21

state regulation you could introduce

1:11:24

laws at the state level that say, as a matter

1:11:26

of public policy, nondisclosure is

1:11:28

not enforceable when it relates to material

1:11:30

emissions or errors in royalties statements

1:11:33

that negatively affect people who are owed royalties.

1:11:35

And then at the stroke of a pen, every

1:11:38

artist in the world because all of their contracts

1:11:40

are governed by Washington, New York or

1:11:42

or or California law, every

1:11:44

creator in the world would suddenly have money

1:11:46

following on the more money than all of the copyright

1:11:49

extensions the last forty years have ever provided.

1:11:51

And we just fill the back half of the book with this. We've

1:11:53

got, like, you know, twelve or fourteen

1:11:55

of these interventions that actually,

1:11:58

rather than giving artists the right to feel aggrieved

1:12:00

that their copyright is being violated, will

1:12:03

let them, like, buy groceries and put braces

1:12:05

on their kid's teeth. Yeah.

1:12:09

It's nice because you focus on artists. Obviously,

1:12:11

it affects everybody from Amazon warehouse

1:12:14

workers.

1:12:15

to burger flippers.

1:12:17

But it, you

1:12:19

know, we artists, and I put myself

1:12:21

in this category, we creators

1:12:24

wanna create and they take

1:12:26

advantage of that fact. Right? Say,

1:12:28

fine. Go ahead. You keep you

1:12:30

keep creating. We'll take care of the rest. Thanks.

1:12:33

Let's take a little break. Wanna

1:12:35

talk more. We've got this

1:12:38

new journalism competition and and

1:12:40

preservation act, which thanks

1:12:43

to the Ted cruises moving

1:12:45

forward. Lots lots

1:12:47

more to talk about, including Elon Musk's robot.

1:12:49

We didn't we didn't even get to that yet. But

1:12:51

first, a word from our sponsor Corey Doctor is

1:12:53

here, Alex Canther, which is from Big Tech

1:12:56

Big Tech Technology podcast. Our

1:12:58

show today brought to you by eight sleep.

1:13:00

If I am well rested tonight, you can thank

1:13:03

my Eight Sleep Pod

1:13:05

Pro Cover two. Now they've gotten

1:13:07

even better. Eight Sleep Pod

1:13:09

cover. If you've got an Eight Sleep

1:13:11

mattress or a pod cover, you already know.

1:13:14

That's how I found out Kevin Rose said,

1:13:16

oh, you gotta do this. Amy Webb got it,

1:13:18

then she said, oh, you gotta do this. So finally about

1:13:20

a year ago, we got it. Eight Sleep

1:13:23

is the only sleep technology that dynamically cools

1:13:26

and heats each side of your bed.

1:13:29

And the reason it's sides is so your spouse

1:13:31

can have her temperature as well to maintain

1:13:33

the optimal sleeping temperature. for

1:13:36

what your body needs. Thanks to Eight Sleep. I'm

1:13:38

getting more than half an hour,

1:13:40

more deep sleep every night, up over

1:13:42

up to an hour now. It

1:13:45

combines dynamic cooling and heating with

1:13:47

biometric tracking. So it's watching you

1:13:49

as you as you sleep.

1:13:51

So it turns out your body Kind

1:13:54

of wants it to get cool. As

1:13:56

you go into deeper sleep, I have my

1:13:58

Eight Sleep set to be warm when

1:13:59

I get in bed because it's nice and cozy. and

1:14:02

then cool off. I go into deeper

1:14:04

and deeper sleep from REM to

1:14:06

deep sleep and then back up, and then it

1:14:08

warms up again in the morning. but it it'll

1:14:10

adjust automatically or you can adjust it manually

1:14:13

as cool as fifty five degrees Fahrenheit,

1:14:16

which was awesome. during the hot

1:14:18

heat wave we had this summer. And

1:14:21

in the winter, as hot as hundred ten

1:14:23

degrees Fahrenheit, it monitors not only your

1:14:25

temperature, your movements, but also

1:14:27

the room temperature. Clinical

1:14:28

data shows eight sleep users experience

1:14:31

up to nineteen percent increase in recovery.

1:14:33

up to thirty two percent improvement in sleep quality,

1:14:36

up to thirty four percent more deep

1:14:38

sleep. More deep sleep makes a

1:14:40

huge difference. I can you know, I just

1:14:42

feel it the next day. I I feel

1:14:44

like I I am, you know, in in

1:14:47

alive and awake and and everything feels

1:14:49

better my mental clarity is better.

1:14:53

Eight Sleep just launched. In fact, we gotta get one this

1:14:55

new next generation pod, the pod three.

1:14:58

which has more sensors, double the amount

1:15:00

of sensors so they've got more accurate sleep and

1:15:02

health tracking, giving you the

1:15:04

absolute best sleep experience

1:15:06

on Earth. I'm I'm still loving.

1:15:08

are a pod two pro cover. I'm very

1:15:11

happy with it. The it's

1:15:12

not magic, but it it definitely feels

1:15:15

like it. I love the night

1:15:17

sleep. I get and you might need this.

1:15:19

More than thirty percent of Americans struggle with

1:15:21

sleep. One of the primary

1:15:23

causes of poor sleep feeling hot at

1:15:25

night. feeling

1:15:27

hot at night. You don't you don't wanna be

1:15:29

hot at night. Go to h sleep dot com slash

1:15:31

twitch. I'm gonna get you a hundred fifty dollars

1:15:34

off at checkout. Sleep cozy

1:15:36

this fall. Eight Sleep currently ships within

1:15:38

the US, Canada, UK, and and the

1:15:40

EU and Australia. Some countries in the

1:15:42

EU. Eight sleep dot com slash

1:15:45

twit. And

1:15:45

if you go there, you'll save a hundred fifty dollars

1:15:48

a a checkout on your

1:15:50

pod. Get the pod.

1:15:52

Trust me. Trust

1:15:53

me. You'll sleep better, and everybody needs

1:15:55

a better night sleep. Thank

1:15:56

you, Eight Sleep. For your

1:15:59

support, Keep me up

1:16:01

at night the idea of having a humanoid robot

1:16:03

walking around through my house. Fortunately,

1:16:06

it probably isn't gonna be Elon

1:16:08

Musk's.

1:16:10

robot.

1:16:11

Elon

1:16:13

and Tesla have an AI event

1:16:15

every year. Last year,

1:16:17

you may remember Musk

1:16:20

talked about his humanoid robot and

1:16:23

brought out a dancer in

1:16:25

a costume to

1:16:28

to show it off. Let me see if I can find a video

1:16:30

of

1:16:31

Optimus is what he's calling it.

1:16:33

I

1:16:33

think because he's a fan of the Transformers,

1:16:36

I don't know. certainly

1:16:38

made progress over the Leotard -- Yeah.

1:16:40

-- person. Yeah. Definitely better than the

1:16:42

Leotard person. It

1:16:45

barely could walk by itself. Here

1:16:47

it comes. I'll turn off the sound because it's

1:16:49

just too annoying.

1:16:52

Elon wants to put these in his

1:16:54

factories, but I think he also feel he says

1:16:56

it could be transformative for civilization.

1:17:00

He's still a little bit behind Boston Dynamics,

1:17:03

not exactly doing back flips or opening

1:17:05

doors for people, But

1:17:08

do you really That's the that's

1:17:10

the answer from last year. It

1:17:12

is not doing that. it was

1:17:14

not doing that. Elon also I would

1:17:16

take the dancer. I would take the dancer.

1:17:18

This is this is Elon's video

1:17:21

of the robot very shapely

1:17:24

delivering. Oh, maybe you can water your

1:17:26

plants. Oh, oh, that's good.

1:17:30

I thought what's her name? Chelsea

1:17:32

Steiner at the Mary Sue had a

1:17:34

good summation here.

1:17:36

Like all Musk promises, this one is vague,

1:17:38

got impressive, and riddled with issues. It's

1:17:40

also wildly unrealistic to imagine that the

1:17:42

robot will be capable enough to replace the labor

1:17:44

force as we know it. here's when Optimus

1:17:47

can do half the moves of this superstar. Here's

1:17:49

the here's the Boston Dynamics robot

1:17:51

doing back flips, dancing, Even

1:17:53

this though, they're carefully curated videos

1:17:56

because, you know, half the time it falls over.

1:18:00

Yeah. I know. But who wants this anyway?

1:18:02

Elan's factory already is loaded with robots,

1:18:04

but they're the traditional giant

1:18:06

German and Japanese robots that could pick

1:18:08

up a car, turn it around, and put it

1:18:11

back on the on the line and so forth.

1:18:13

But can they automate

1:18:15

the racism of his fact? No. No. That's

1:18:17

that's only a human can really Right now, their

1:18:19

their wage bill for doing the racism

1:18:21

is very high, and they're gonna need to automate

1:18:23

that some other And I have to bring in pay.

1:18:27

Facebook's bot let Tay do

1:18:29

it. Yeah. Yeah. Maybe it was Microsoft's.

1:18:31

Oh, but but here's another here's

1:18:33

another unpopular take since I guess that's

1:18:35

my role on the show today. I think

1:18:37

we we should be hesitant before we

1:18:40

discount, you know, this type of thing. And

1:18:42

it might not be the must've got. Okay?

1:18:44

which by all accounts is not a very impressive

1:18:46

machine. But we are seeing

1:18:48

real interesting advances in

1:18:50

robotics right now, humanoid style

1:18:52

robotics And like we've seen

1:18:55

with AI, these things, you know, we we

1:18:57

see these breakthroughs of research. We get we get

1:18:59

lots of duds. And then all of a sudden,

1:19:01

we type in a sentence and the robot is drawing picture

1:19:03

for us or we talk to it, and we'd have

1:19:05

a Google engineered fill engineered

1:19:07

field that had sent you. And I think that

1:19:09

the interesting report that no one's paying

1:19:12

attention to. Everyone pays attention to what Musk

1:19:14

does. But an interesting report that no one's paying

1:19:16

attention to along this line is that Amazon

1:19:18

has been trained has been training robots

1:19:21

to pick items out of

1:19:24

boxes and then, you know, end up, you know,

1:19:26

stowing them. So Most

1:19:28

of Amazon's workforce in the in the warehouses

1:19:30

are people that pick stuff out of boxes and

1:19:32

they put them into shelves, pick them into from the

1:19:34

shelves, and put them into crates. to be shipped

1:19:37

off, and they're making real progress there.

1:19:39

And so I think this idea of, like, you know,

1:19:41

Elon's robot is easy to laugh up,

1:19:43

but the advances that we're seeing you

1:19:45

know, in this type of robotics are very,

1:19:47

very real. And the fact that Musk is

1:19:50

talking about disappearing in in the warehouse,

1:19:52

so it doesn't come out of nowhere. And there is definitely

1:19:55

some progress here.

1:19:56

And and I'm not I'm not

1:19:58

laughing. I think that that

1:19:59

this is some serious technology to be

1:20:02

reckoned with. you know, maybe not coming from Elon

1:20:04

Musk but here today for sure. There's

1:20:06

a there's a term out of AI research

1:20:08

which is a centaur which is when

1:20:11

you have human AI collaboration, like

1:20:13

a chess robot and a chess player playing

1:20:15

together to do things that neither than it

1:20:17

could do on their own. But there's also this

1:20:19

this sort of out of labor economics. This

1:20:21

has turned the reverse center, which

1:20:23

is when the body is the inconvenient

1:20:26

meat puppet for the AI. And

1:20:29

Amazon are kind of the masters of that. I'm gonna

1:20:31

paste a link into the chat of a a

1:20:33

thing I wrote about reverse centers and Amazon.

1:20:35

So that would be things like their drivers

1:20:38

being subjected to kind of

1:20:41

like super human conditions or

1:20:43

super human constraints by

1:20:46

the by the AI's

1:20:49

that are monitoring them in the cars or

1:20:51

packers being driven to do

1:20:54

unrealistic working tempos

1:20:57

in their warehouses. And and Amazon

1:21:00

leads the country in warehouse injuries.

1:21:02

Yeah. And the more automated Amazon

1:21:05

warehouse is, the more injuries

1:21:08

people incur, I'll I'll

1:21:10

stipulate that, like, automation is

1:21:12

a thing. and robots are are

1:21:14

cool and that Amazon

1:21:17

has every motive in the world to try to make good robots.

1:21:19

But I also wanna sound the note of caution

1:21:22

that a lot of what Amazon has booked to

1:21:24

its shareholders as profits in automation

1:21:27

have really been ways to get people to work

1:21:29

in at an unsafe tempo in

1:21:31

ways that put themselves at risk, and in the case of their

1:21:33

drivers put other people at risk other users

1:21:35

of the road at risk. I know a kid who works at

1:21:37

Whole Foods owned by Amazon. and

1:21:40

it's a very different experience. Sometimes, some hours,

1:21:43

he's working for Whole Foods, some hours, he's

1:21:45

working for Amazon. And the at

1:21:47

the minute, he's on the clock for Amazon, there

1:21:50

are very clear, almost

1:21:53

unachievable metrics for his performance,

1:21:55

and

1:21:55

they're made very clear.

1:21:57

and it is a very different

1:21:59

experience.

1:21:59

It's kinda you nailed

1:22:01

it. It's you are you are

1:22:03

at the mercy of the machine. Yeah.

1:22:06

And I need to state that I'm not celebrating this

1:22:08

stuff. But when we look at that No. No.

1:22:10

I think you're right, Alex. Don't downplay. I think

1:22:12

we and we've seen this very, very powerful

1:22:15

technology. I agree. We've seen this with the explosion,

1:22:17

thanks to stable diffusion and dolly too,

1:22:19

and mid journey. I mean, just explosion

1:22:22

in AI art. Do you

1:22:24

think though you know,

1:22:26

initially, my initial reaction was this is

1:22:28

like a Cambrian explosion. Like, suddenly,

1:22:31

this is taken off, and we've reached

1:22:33

the high stick with AI in

1:22:35

some interesting way. And

1:22:37

now more and more I'm thinking it's a

1:22:39

parlor. trick that

1:22:42

the the AI is just it's

1:22:45

not really AI almost. It's just combining

1:22:47

images together in an interesting

1:22:50

Wait. What's your take on it, Alex? I have

1:22:52

you tried Dolly before? Yeah. No. It's very

1:22:54

impressive and stable diffusion and mid journey.

1:22:56

Yeah. They're very stable diffusion. seems to be

1:22:59

the fastest moving because it's open

1:23:01

source people can run it on their own servers and there are

1:23:03

a lot people who've adopted it. I

1:23:05

follow the stable diffusion Reddit

1:23:08

subbred it, and it's pretty amazing

1:23:10

what what they're doing. And yet,

1:23:14

I'm not convinced it's I

1:23:17

don't know. Is it AI? Is it true AI?

1:23:20

Of course. Yeah. Yeah. And and the images

1:23:22

are gonna be as creative as, you know,

1:23:24

the prompts that you're gonna write them. I don't think we've

1:23:26

had anything like this to, you know, in our history

1:23:29

before. And, you know, there's there's so many

1:23:31

interesting applications. Here's one.

1:23:33

I think that so anyone who's worked

1:23:36

with the marketing department, you know,

1:23:38

knows that it's always a struggle to communicate

1:23:40

to the creative department what you need them to

1:23:42

to create or what you'd like them create because

1:23:45

words are imprecise and art, you know, pictures worth

1:23:47

a thousand words. So you can't just

1:23:50

be like, I need a picture of this and and they'll know exactly

1:23:52

what you're doing. Actually, you got you know, graphic

1:23:54

designer to product manager or marketing manager

1:23:56

is one of the most difficult pieces of communication

1:23:59

I think in the business

1:23:59

world. and

1:24:00

I've been there and it's tough. Yeah. And there are ways

1:24:02

you get around to make this more prominent or we're

1:24:04

trying to get this message across. We have creative brief.

1:24:07

Now, I think what this stuff is, the amateurs

1:24:09

can make a version of the image,

1:24:11

you know, with something like dolly and then

1:24:13

pass it to the professionals, and then you

1:24:16

know, be able to vice versa. Both

1:24:18

of these these systems seem to really work well

1:24:20

with a sketch. For here's

1:24:22

an example of using old cartoons as

1:24:25

and it images says Fred Flintstone

1:24:27

turned into this is so

1:24:29

start with the if you can't the starting point,

1:24:32

it is kind of it was inspector gadget,

1:24:34

you know. I mean, don't know. I I continue

1:24:36

to be blown away every time I do, you

1:24:38

know, one of these searches. And, I mean, Leo,

1:24:40

it's sort of it kinda shows you

1:24:42

how much advances we've made. I mean, how

1:24:44

serious of events of events we've made.

1:24:47

If you're if you're sitting there and typing

1:24:49

in a sentence prompt, you know,

1:24:51

on your browser. And then next thing you know,

1:24:54

a machine will draw it for you. Yeah. You're

1:24:56

like, man, you know. Like,

1:24:58

think about how far we've come. You know,

1:25:00

if that's your reaction, Yeah. This

1:25:02

stuff becomes unremarkable once

1:25:04

it works because

1:25:05

we expect it, and I think that's the place we

1:25:07

are today with these programs. Look at this

1:25:10

progression. from

1:25:11

what is essentially a stick figure

1:25:13

as this is y

1:25:15

flu and stable diffusion combined

1:25:17

together. as

1:25:18

it creates a more and more realistic

1:25:21

little human help. There's

1:25:22

so it is a human partnership, I guess.

1:25:25

that makes sense. The funniest thing I saw in Reddit

1:25:27

last week, I just paced the URL for it into

1:25:29

the chat is was a a

1:25:31

Drake meme whose title

1:25:33

is what makes you a human and the

1:25:36

the pushing things away is

1:25:38

to love and care about others and the the yes

1:25:40

that's right is the selecting all images with by

1:25:42

weeks. Oh,

1:25:44

isn't that interesting? yeah. That's good.

1:25:47

Talking about capture. Yeah. Yeah.

1:25:49

So it's it's I mean, I

1:25:52

I agree that that this is a a very

1:25:54

powerful technology. It's

1:25:56

it's super interesting to work with.

1:25:58

the you

1:25:59

know, I've also been where you are trying to find

1:26:01

art to use as reference to give to an Illustrator.

1:26:04

Being able to describe art is really great. I

1:26:06

mean, Google image search was was

1:26:08

a huge phase

1:26:11

change for that as well. Yeah. And,

1:26:13

you know, III was an imagineer

1:26:15

for a while. And in the imagineering

1:26:18

archive, there's a room or I don't even know if

1:26:20

it's still there, but there's a room with bankers boxes

1:26:22

filled with magazine clippings of illustrations --

1:26:25

Yeah. -- organized by And

1:26:27

if you had to draw a water

1:26:29

fountain, they just had a box

1:26:31

of clipped out illustrations of waterfounds

1:26:33

that you would ring onto archives and

1:26:36

they would send up a box of reference

1:26:38

for you. Yeah. You know, so they're they they

1:26:40

your this this is definitely something that is

1:26:42

making the lives of illustrators easier,

1:26:45

making it easier for people who aren't illustrators

1:26:47

to talk to people who aren't and say, that's what

1:26:49

I mean when I say water fountain, this picture

1:26:51

here is the the kind of water fountain. I mean, not

1:26:53

this picture over here. And and

1:26:56

certainly, like Dolly and and all the other

1:26:58

ones help there as well. But

1:27:00

I will say that I don't see a path

1:27:02

from statistical

1:27:05

inference and, you

1:27:07

know, deep learning networks to

1:27:10

GAI. I I think that

1:27:12

to say that if that's true and Inference,

1:27:15

we get GAIs, like saying if we read

1:27:17

these horses carefully enough, will have a

1:27:19

locomotive. Right. Right. You are doing

1:27:21

general artificial intelligence. is human

1:27:23

scale intelligence, Scott. But is there anybody

1:27:25

actually arguing that Corey? Oh, yeah.

1:27:28

Tons. Oh, yeah. That's the whole argument.

1:27:30

Right? The whole argument that -- Mhmm. --

1:27:32

well, the whole, like, Nick Bostrom, Elon

1:27:34

Musk, Sky Net is coming

1:27:36

out of our AI. I have looked at open AI

1:27:39

and what it can do and now I'm afraid for the

1:27:41

human race. They're they're basically saying

1:27:43

we are gonna selectively breed this horse

1:27:45

long enough that eventually we're gonna have

1:27:47

a locomotive and then it's gonna kill us all.

1:27:50

And it it is like such obvious

1:27:53

nonsense to me that I'm quite baffled by

1:27:55

it. And you know what

1:27:57

is the discontinuity? So it's

1:27:59

obviously horses and locomotives. Clear.

1:28:02

But so But this is about human cognition.

1:28:04

Is there something about human cognition

1:28:07

that is unreachable?

1:28:09

No. No. No. It's just not it's just not a human

1:28:12

cognition is not statistical inference. You

1:28:14

know, we don't ever mess it entirely. It's

1:28:17

not it's not I mean statistical inference

1:28:19

might be a component of it, it probably

1:28:22

is. But the idea that that

1:28:25

increasing innovation

1:28:28

in the realm of statistical inference eventually

1:28:30

produces human

1:28:32

cognition. It's just it's it's wrong.

1:28:35

And you know, there's this

1:28:37

corollary, which is automation

1:28:41

unemployment corollary. Right? The

1:28:43

the looming automation unemployment crisis,

1:28:45

which again, I think, is just, like,

1:28:47

doesn't isn't right. Like, its

1:28:49

its foundations are wrong. So, like,

1:28:52

you look at the stories about automation

1:28:54

unemployment, you see things like, oh, the

1:28:56

most popular job in America's truck driver

1:28:59

and driving trucks is something that we can do

1:29:01

with ML because we can give them

1:29:03

a dedicated lane on the highway and set

1:29:05

them to following each other and basically invent

1:29:07

a shitty train But you

1:29:12

know, the thing is that the Bureau of Labor Statistics

1:29:14

category for truck driver incorporates

1:29:17

anyone who operates a heavy goods vehicle.

1:29:19

So it's not the most like sixteen wheeler

1:29:21

driver is not the most popular job

1:29:24

in America. It is a relative

1:29:26

all an unimportant part of our overall economy,

1:29:28

not that those people are unimportant, but like they

1:29:30

you know, if all of the truck drivers were unemployed tomorrow,

1:29:33

the change in the unemployment figures would not

1:29:35

be very large. meanwhile, we're

1:29:37

just not getting anywhere with the self driving cars.

1:29:40

Right? Like that that there's so much

1:29:42

smoke and mirrors and self firing cars, the only

1:29:44

ones that seem to perform at all are the ones

1:29:46

that actually just have a human

1:29:48

remotely driving

1:29:49

the car, overseeing the car,

1:29:51

And if those people's attention wanders, which

1:29:53

it will inevitably,

1:29:55

then those cars become murder bots.

1:29:57

And, you know, it raises an important point, which

1:29:59

is

1:29:59

that

1:30:00

we already have a lot of human intelligence.

1:30:03

Like, we

1:30:04

have billion humans. We

1:30:06

don't have enough non human intelligence.

1:30:08

Right? Like that, the capacity to be

1:30:10

vigilant for things that happens very rarely

1:30:13

is is not a capacity that humans

1:30:15

mostly have. a few people may

1:30:17

have it, but it's not a widespread trait

1:30:19

in our population, which is why the TSA

1:30:22

is really good at spotting water bottles

1:30:24

and really bad at spotting guns. because

1:30:26

they

1:30:27

never see a gun. Right? But they see water bottles

1:30:29

all day long. Like, you cannot leave neurons

1:30:31

trained to

1:30:32

do a pattern recognition for

1:30:34

a pattern that you never encounter

1:30:36

because those neurons will be retrained to

1:30:39

make you better at the

1:30:40

pattern recognition that you do all day.

1:30:42

Right?

1:30:42

And so they just forget like you

1:30:44

can train them to spot guns on an x-ray,

1:30:46

but then they'll forget not because

1:30:48

they're lazy or whatever, but because they never see guns

1:30:51

on an x-ray. They see water bottles all day long.

1:30:53

That's

1:30:55

interesting. So you're not you're not doesn't

1:30:58

require a notion of a human goal or

1:31:00

some sort of magical capability that

1:31:02

Cognizant Yeah. No. The cost just also

1:31:04

the cost is. Yeah. The conversations

1:31:07

that I think there are conversations that really

1:31:09

occur on the fringes about the stuff leading

1:31:11

to general intelligence. And I

1:31:13

I think that that, you know, the mainstream

1:31:15

conversation about this stuff looks at it.

1:31:17

rationally and says there's

1:31:19

a lot of stuff that we can do, humans and artificial

1:31:22

intelligence combined. And I

1:31:24

can Jira and the people sorry. Go ahead.

1:31:26

I beg your pardon. Go ahead.

1:31:29

No. No. You finish. I'm sorry. I thought you're done.

1:31:31

Yeah. I apologize because the latency in

1:31:33

the in the Skype is is killing us as

1:31:35

usual, not Skype Zoom. Go

1:31:38

ahead, Corey. I was just gonna

1:31:41

say during the lockdown, the World Economic Forum

1:31:43

asked me to give them a talk on technological unemployment.

1:31:46

And when I sent them the text of the talk, they

1:31:48

withdrew the invitation. So I turned it into a

1:31:50

column. and it's the

1:31:52

it just put it in the URL there. But You weren't

1:31:54

saying what they wanted you to say, Corey. I think, basically,

1:31:57

I said, like, I don't think we're gonna have

1:31:59

AI driven unemployment because even

1:32:01

if we automate some stuff, like, we're gonna

1:32:03

have to, you know, relocate every

1:32:05

city twenty kilometers inland over the next three

1:32:08

hundred years. That's full employment for everyone no

1:32:10

matter how many robots we build, there's just like

1:32:12

more work than than we can imagine, and

1:32:15

they they didn't like that at all.

1:32:17

really interesting. Yeah. Didn't

1:32:19

fit their their model. But

1:32:21

they they really believe and that's why I raised

1:32:23

it. They really believe in technological

1:32:26

unemployment, generally eye breakthroughs on

1:32:28

the immediate horizon. You know,

1:32:30

they they they aren't fringe beliefs in the in

1:32:33

the In the halls of power or in the

1:32:35

halls of business or even in finance,

1:32:37

they're they are accepted as gospel.

1:32:40

You know, there are a few things like general, you know.

1:32:43

there are a few things like general AI, hold on a second.

1:32:45

General AI, fusion, quantum

1:32:47

computing,

1:32:49

the

1:32:50

it seems prudent to perhaps

1:32:53

consider their eventual

1:32:55

invention even if

1:32:57

they're not necessarily around the corner.

1:33:01

Go ahead, Alex. I mean, it's just

1:33:03

not what I hear. You know, I mean, maybe do you think it's

1:33:05

gonna happen? No. I'm sad. I do

1:33:07

will it happen eventually? Who knows? But

1:33:09

this I I just don't agree with Corey

1:33:11

about, you know, this being an accepted

1:33:14

thing. At least not in the conversations

1:33:16

I hear. you know, maybe the Oh, you're

1:33:18

saying that people don't believe, generally,

1:33:21

I don't think so. No. I think look at the response

1:33:23

to what happened with this Google engineer who said that

1:33:25

the chatbot was Google didn't

1:33:27

like it. sent to you. Google fired him.

1:33:30

Anyone who with any standing in the research

1:33:32

candidate said that's not fair. You know,

1:33:34

faulty and idiot. Yeah. And -- Yeah. -- yeah.

1:33:36

And, like, I don't know. I mean, Corey,

1:33:38

maybe the, you know, the influencer class

1:33:40

likes to talk about this at their conferences. But

1:33:43

this idea that AGI is right around the corner,

1:33:45

just to me, I've never heard that, you

1:33:48

know, from credible folks in the industry.

1:33:50

Sure. Well, I mean, science fiction

1:33:52

writers certainly think it's nonsense. Right?

1:33:54

Like it is a recurring theme at science fiction

1:33:56

conventions. Why are all these CEOs out

1:33:59

there saying this is this this is

1:34:01

this on the horizon. But they are saying Well, I think

1:34:03

this I think the CEOs aren't saying it.

1:34:05

I think the CEOs are saying that it's very

1:34:07

powerful technology. But I don't think

1:34:09

they're saying we're gonna be hand in hand with

1:34:12

artificial general intelligence, and

1:34:14

that's what we're working toward. mean mean, you

1:34:16

know, investing in that, and Wall Street can

1:34:18

believe it. That's a hypothesis

1:34:20

to the center for existential risk who are,

1:34:22

like, you know, have attracted, like, billions

1:34:24

of dollars, hundreds of millions of dollars in

1:34:26

capital. I mean Well, just aren't you already saying

1:34:28

that we shouldn't equate money with smarts?

1:34:31

No. It's true. It's true. Right? So there what

1:34:33

I'm saying is that there it may be a small number,

1:34:35

a very rich people who believe this, but there

1:34:37

are some very rich people who believe it and a bunch

1:34:39

of weird stands for them who also believe

1:34:42

it. I just

1:34:42

think if you speak with people, credible folks

1:34:44

in the industry, you know, folks who are actually doing

1:34:46

the research. But Nick Bostrom mean,

1:34:49

I I like the guy. He's a philosopher. He's

1:34:51

not here. He doesn't work in in machine learning.

1:34:53

Yep. and you speak with researchers, you speak with

1:34:55

the tech companies, you know, they might use marketing

1:34:58

terms to talk about the power of their artificial intelligence.

1:35:01

Cindar, Pichai calling it, you know, as powerful

1:35:03

as fire. You know, that sounds like marketing to

1:35:05

me. But I never hear him or

1:35:07

anybody at Google talking about you know,

1:35:09

us reaching artificial general intelligence outside

1:35:12

of one guy, you know, who actually has an

1:35:14

interesting story to tell, and I did have one

1:35:16

on my podcast. We had an interesting conversation.

1:35:19

That that being said, you

1:35:21

know, he's the extreme

1:35:23

exception

1:35:24

and not the rule.

1:35:26

you're you're so I that's good. I have to I

1:35:28

have this is the oh, yeah. Here it is.

1:35:32

Greg La Moyne. Right? Blakely

1:35:35

mine. Yeah. Yeah. Blakely mine. Yeah. Yeah. Yeah.

1:35:37

Mhmm. Nice. Yeah.

1:35:39

I think think the picture