0:04

Conversations with Tyler is produced

0:06

by the Mercatus Center at George Mason

0:08

University, bridging the gap

0:10

between academic ideas and real-world problems.

0:13

Learn more at mercatus.org. For

0:16

a full transcript of every

0:19

conversation, enhanced with helpful links,

0:21

visit conversationswithtyler.com. Hello

0:27

everyone and welcome back to Conversations

0:29

with Tyler. Today I'm chatting

0:31

with Toby Lutke, who is the

0:33

co-founder and CEO of Shopify, the

0:35

famous Canadian e-commerce firm. Toby, welcome.

0:37

So good to be here. So

0:39

good to see you, Tyler. I

0:42

have so many questions. Do you still

0:44

do stand-up meetings? Whenever

0:46

we have hack days, I do them. I

0:49

know what I feel they need to make

0:51

a comeback, like the Friday afternoon stand-up meeting

0:53

is just one of the greatest accelerants for

0:55

businesses. I'm highly encouraging everyone to do them.

0:57

But digital version doesn't work so well, does

0:59

it? It's some slack bot or

1:01

so telling everyone to stand up and type

1:03

something. It just doesn't do the same thing.

1:06

It's one of the problems with remote companies.

1:08

And it may be harder to monitor. What

1:10

other systems do you have for getting good

1:12

information out of meetings? Okay, so I grew

1:14

up as an outsider, as in kind of

1:16

have an outsider mindset, which is kind of

1:18

hard to argue from first principles at this

1:20

point. But in a small town that was

1:22

super not interested in computing in Germany. So

1:24

I learned all sorts of skills for observing

1:26

people doing interesting things from afar. My

1:29

lifeline back in those days was getting a copy

1:31

of the back archives of the Linux kernel

1:34

mailing list and so on. This is how I learned C

1:36

programming. I grew up as an observer

1:38

on information. That's what I do just for

1:40

fun a lot. Also that means I grew

1:42

up on the internet, which is good and

1:44

bad. So one of my ways to get

1:46

a lot of out of meetings often involves

1:49

saying high conviction, incorrect things and just waiting

1:51

to figure out like seeing everyone wanting to

1:53

correct me. It's a more efficient way. So

1:55

it's like Twitter. If you want to learn

1:57

something, you just say something wrong. Say something wrong on the internet.

1:59

Yes. Books in real life too. It's not

2:01

people's favorite thing I do, but we

2:04

have a wiki inside, and there's a blueprint

2:06

for me. And it says that I do

2:08

this. So some people have reasonable warning these

2:11

days. The systems for making meetings better, should

2:13

we also use them in our social lives?

2:17

Family dinners. You

2:19

get together with friends. I don't

2:21

know. I think haunting what makes companies work

2:23

too far into private lives can also backfire.

2:25

Although, I did. With my

2:28

kids, I always booked one-on-ones. They call them dinner appointments. And

2:30

I tell them, I'm going to treat you as an adult.

2:32

And we're going to have an outcomeization until you say you're

2:34

a kid again. And I'm going to answer every one of

2:36

your questions to the best of my abilities. And they

2:39

choose some restaurant, we go there. And that's what we've

2:41

done for a while. And of course, that's cell phones

2:43

down. I think that's an important thing to do. What

2:45

we do with meetings is like, inside

2:48

of a company, we do a bunch of

2:50

things. We periodically delete all meetings, recurring meetings,

2:52

because it seems very hard for people to

2:55

subtract. It's very easy to add. That only

2:57

leads one way. So if you have ceremonies

2:59

like this, I

3:01

would be interesting what would happen if everyone would have

3:04

to zero-sum their follow list. Yeah,

3:06

zero-budget their follow list. I

3:09

think that would significantly change people's experience with

3:11

social media. I wonder if that's a good

3:13

idea. But also with your

3:15

friendships, it's striking to me when I'm

3:17

in Italy, I very often see what

3:19

I call street conferences. That is, people

3:21

talking to each other often heatedly, and

3:23

they're standing. When I'm in Germany, people

3:25

are talking to each other heatedly. They

3:27

tend to be sitting. You have a

3:29

similar impression? What

3:32

a fascinating observation. The idea, amtisch, it's something

3:34

very German about it. You can pound your

3:36

fist on the table. It's a stamtisch, really.

3:40

I think in both instances, people do

3:42

something, which I think has a lot

3:44

of relevance to the real-life versus social

3:46

media conundrum that people are wrestling with,

3:48

which is that I think there's a

3:51

significant human need, unacknowledged, for

3:53

venting. I think venting is

3:55

one of those extremely

3:58

important outlets that. It's

4:01

an original safe space in a way where at

4:03

some point people say, okay, well, you're clearly off.

4:05

At least people say this to me. I

4:08

can go pretty far in venting. Germans

4:12

can be going very far. Stereotypes

4:14

are funny because they often are too.

4:16

Venting catastrophizing these kind of things and

4:18

then having your friends reel you in.

4:20

But the issue I think we've seen

4:22

over, especially the beginning of this decade,

4:24

is as people ported their venting online

4:26

and then got their one take, retreated

4:28

forever because it just kept it with

4:30

imagination. I feel people misunderstood what was

4:32

actually going on for a little bit

4:34

and I hope everyone has acclimatized to

4:37

this reality now. How do

4:39

we create safe outlets for venting in

4:41

companies or institutions? What is it that one

4:43

does? Because you don't want it

4:45

to turn into negative contagion, right? I

4:48

actually think even just talking about venting being a

4:51

thing that's actually good and prefixing

4:53

then someone just wants to actually just say,

4:55

get a bunch of things, like field test

4:57

some takes is a good idea. I

5:00

think that can disable the power of

5:02

it spiraling everyone right after. I think

5:04

that's useful. I

5:06

don't know. I

5:08

didn't go to it because it wasn't my neck of a word. I've

5:11

seen parties organized or evenings organized now where

5:13

everyone gets a note with an outrageous position

5:15

that they are supposed to represent for the

5:18

rest of the evening and then they're supposed

5:20

to tear it up and toss it away

5:22

just to allow people to have plausible reliability

5:25

of whatever they want to talk about because

5:27

they can just fall on, oh, I was

5:29

told to represent this. I

5:31

think that seems like, I don't

5:34

know if it's a good idea, but I

5:36

love that someone is trying this because that

5:38

seems like social license to actually talk about

5:40

stuff that otherwise can't be talked about and

5:42

seeing where it leads. Often

5:44

it leads to exactly the wrong place and then

5:46

you get that out of your system and you

5:48

don't need to share this as a tweet afterwards.

5:51

Are German meetings different? Yes.

5:55

How? So, I'm German, I grew up

5:57

for 20 years in Germany, I then moved to Canada.

6:00

I'm from Koblenz, right? I celebrated 2,000

6:02

years when I was a teenager there. Julius

6:04

Caesar might have come through, and that was

6:06

probably the most exciting thing that ever happened

6:09

there. I think it goes back into

6:11

the stereotypes culture. I'm German. I started a

6:13

company with another German both in Canada for

6:15

Americans. So like, Shopify is sort of

6:17

interesting in the year of different cultures. Well,

6:19

I mean, straight off the bat, it's

6:22

Germans just are

6:25

blunt. Like, there's just

6:28

no shit-centric configuration that

6:30

needs to be constructed to say

6:32

if something is bad. I think

6:34

Germans have a more innately

6:36

higher quality bar and less

6:39

tolerance for underperformance on that quality

6:41

bar. It's like products are either

6:43

like below it or,

6:45

you know, good, great work class above

6:48

it. But like below that, nothing

6:50

else registers as anything of value, which is totally different

6:52

in North America. And I think that's good and bad

6:54

that comes from this. I think sort of a cultural

6:57

appreciation for good products, craftsmanship done

6:59

right and so on is something,

7:01

you know, like that's more associated

7:03

with Europe. But the

7:05

quick iteration, be embarrassed by

7:07

first version and then build from there is something

7:09

that, you know, North America does better. And

7:12

so it's interesting. I've seen this

7:14

in meetings where people fell

7:16

on cultural lines of, you know, is this

7:18

should we ship this? Should we not ship

7:20

this? Is it valuable to build this way?

7:23

Or should we, you know, spend a couple

7:25

of like spent months and months

7:27

and months trying to figure out exactly what to build

7:29

and then build that and try to get it as

7:31

close and perfect as possible? Or should we just like

7:33

it very, very quickly? So yes, they go differently.

7:36

Are Canadians different in meetings

7:38

than US Americans?

7:40

Yeah, yes, that's true.

7:43

It's more on the side of American

7:45

on definitely on a minimum quality bar.

7:48

I think Canadians are often more about long

7:50

term like just like like I've seen Canadians

7:53

more often think about what's the next step after

7:55

this step, but also just lower ambition. That's probably

7:57

not the most popular thing to say around here.

8:00

but Canada has sometimes,

8:02

Canada's problem often culturally is

8:04

a go for bronze mentality,

8:06

which apparently is not uncommon

8:08

for smaller countries attached to

8:11

significantly more cultural or just bigger countries.

8:13

I found it's very easy to work

8:15

around, because I think a lot of

8:17

our success has been due to just

8:20

me and my co-founder basically allowing everyone

8:22

to go for work class. And

8:24

everyone's like, oh, OK, well, if we are allowed

8:26

to do this, then let's go. And I think

8:29

that makes a big difference. But ratcheting up ambition

8:31

for a project is something that one has to

8:33

do in a company like in Canada. But

8:36

is there something scarce that is

8:38

needed to inject that into Canada

8:40

and Canadians? Or is it simply

8:42

a matter of someone showing up and doing it, and

8:44

then it just all falls out and happens? So I

8:47

don't know. Inasmuch as Shopify may be seen as something

8:49

that succeeded, that alone didn't do it. It would have

8:51

been very, very nice if that would happen. Now,

8:54

there's another chord a founder's coming through, like some

8:57

of them have been part of Shopify or come

8:59

back from a valley.

9:01

There's some great companies in Agri, like Neo,

9:03

that are more ambitious. But I think it's

9:05

a decision. I think it's a

9:07

bit of a decision. The

9:10

time it worked perfectly was when Canada was

9:13

hosting the Winter Olympics, which is now a

9:15

little bit of ancient history. But there was

9:17

actually a program, Cara White, that's called Own

9:19

the Podium, which, because that

9:21

makes sense. It's home. We

9:24

have more winter than most times, so therefore, let's

9:26

do well. And then they did. By

9:29

far the best performance of the Canada Olympic

9:31

team of all times. And

9:33

I do wonder if it's actually, I

9:36

think to systematize it and make it

9:39

stick, changing a culture is

9:41

very, very difficult. But instances of

9:43

just giving everyone permission

9:45

to go for it have also

9:48

been super successful. Why

9:50

were you miserable in German school? I

9:54

think because German school at that time

9:57

was in love with

9:59

Syria. Realizing out answers and trying to

10:01

fill you up with as many answers as

10:03

possible and hope that you would be able

10:05

to apply them to problems you encounter later.

10:07

I don't know if there's a good

10:10

theory for that being a sensible

10:12

approach that would stand up to

10:14

reason. It certainly didn't work for me.

10:17

I kind of am literally upset. I need to

10:20

have every problem before I can learn

10:22

the answer to it. That was in

10:24

stark contrast. What was being taught in

10:26

the 90s and so in doing this

10:28

times was in stark contrast with what

10:30

was interesting for working with

10:32

computers. That was really just the most fun thing

10:34

for me to do during this time. It seemed

10:36

very valuable even then. This

10:39

probably sounds too abstract. Latin

10:42

as first extra language is just not highly

10:44

utilitarian. That is not the... knowing Latin is

10:46

very rarely the correct answer to questions you

10:48

might encounter later in life. I'm not saying

10:50

it's not valuable in some way, but maybe

10:52

like start with English. That would be a

10:54

good start. But there's plenty

10:56

of technical talent in Germany

10:58

and plenty of young people speak English

11:00

quite well. Why aren't there

11:03

more German tech giants? I

11:07

mean the

11:10

hot tech here on this is like

11:12

there are. They're just they're called Shopify

11:14

and Pantier and others. But in

11:16

Germany, Germany is not a tiny nation. The EU

11:19

is of course a large market. Enough

11:21

of you speak English to have a common language.

11:25

I would like I would love to know. Honestly, I

11:27

think about this a lot, but I don't know if

11:29

I have the best person to analyze it. Because

11:33

it's like what I'm

11:35

hearing and it makes sense. It's just

11:37

that tech is

11:40

something that Americans do from perspective of the

11:42

terms. And it's like it's it's I think

11:44

I really don't think the general

11:46

population is believes that is truly like

11:49

tech is adding a lot

11:51

to life. This may be a reflection of

11:53

areas I go to visit again. I mostly

11:55

visit family and friends in smaller like not

11:58

the tech centers of Germany having these

12:00

conversations, there is a very much

12:02

a pessimism about the future that

12:04

I think means you cannot

12:06

build tech companies because you kind of have to be

12:09

optimistic about the future to like otherwise, why

12:11

would you want to contribute to

12:13

progress and making it come to

12:15

be faster? So I think

12:18

that's one thing. I think

12:20

it's very hard to hire staff. Like

12:22

in North America, I found that people

12:24

take big chances. If they believe and

12:26

have conviction in a company doing something,

12:29

they would leave an excellent

12:31

career to give it a go. And it just

12:33

like also seems to not be true in Germany.

12:35

So like access to excellent talent is

12:38

just simply harder by some men

12:40

making culturally decisions differently, partly because I think

12:43

startups are a low status pursuit from the

12:45

best I can tell. Let's

12:47

say we compare Germany to the

12:49

Netherlands, which is culturally pretty similar,

12:51

very close to Koblenz. They

12:53

have ASML, Adyen, Netherlands

12:56

is a smaller country. Why

12:59

have they done relatively better? Or you could

13:01

cite Sweden, again, culturally not so distant from

13:03

Germany. You're asking very good questions that I

13:05

much rather would ask you, you know? I

13:09

don't know. I wish I know. I

13:11

spent, I started a small company in

13:14

Germany, didn't do anything. So it's not like people

13:16

didn't do this. I came to Canada again this

13:18

time it worked. And then I

13:20

was headstone for a very long time building my

13:22

thing because it just like was consuming. So like

13:24

I didn't pay too much attention to you. I

13:27

wasn't even very deliberate about where I started a company. I

13:30

started in Ottawa and because that's me and my

13:32

wife were doing the time she was studying there.

13:35

And then we could find great talent there

13:37

that was overlooked,

13:39

it seemed, and gave everyone projects

13:42

to be ambitious with and it worked. And you know,

13:44

I think that if you create in

13:46

a geography a consensus that your company

13:48

that really, really is worth working for

13:50

because it's interesting work, great

13:52

work and might actually lead to something, then

13:55

you can build it. And I think this is

13:57

something that I just, I don't quite understand why

13:59

this is not possible to do in so

14:02

many places in Germany. Because again, Germany

14:04

does have this wonderful appreciation

14:06

of craftsmanship, which I think is

14:08

actually underrepresented in software. I think

14:10

it's only recently, usually by Europeans,

14:12

being brought up. Patrick Hollisens talks

14:14

about it more and more, and

14:16

certainly I do too. Making

14:19

software is a craft. I think

14:21

in this way, Germany, Czech Republic,

14:23

other places, Poland, are extremely enlightened

14:25

in making this part of

14:27

an apprenticeship system. And I apprenticed as a

14:29

computer programmer, and thought it was exactly the

14:31

right way to learn these things. That means

14:33

there's, I believe, a

14:35

lot of talent that then makes decisions other

14:37

than putting it together to build ambitious startups.

14:40

Something needs to be uncorked by the people

14:42

who have more insight than I have. I

14:44

think part of a hypothesis is that the

14:47

Netherlands and also Sweden are

14:49

somewhat happier countries than Germany.

14:51

People smile more, at least superficially,

14:53

the more optimistic, the more- I

14:56

think it's optimism, yeah. It's

14:58

striking to me that Germans, contrary to stereotype, I

15:00

think they have a quite good sense of humor.

15:03

But a lot of it is irony or somewhat black, and

15:07

maybe that's bad for tech. And

15:09

I wonder, people in the Bay Area, do they

15:11

have a great sense of humor? I'm not sure

15:13

they do. Maybe there's some correlations

15:15

across those variables. I think they actually banned

15:17

humor for a little while from the Bay

15:19

Area. I think it might make a comeback

15:21

now. It seems like an easy

15:23

out, but it's actually potentially- The

15:25

optimism angle is a lot bearing for this. You've got

15:28

to believe that the future is going to be better

15:30

than today to want to make the future come sooner,

15:33

which is in your tiny, tiny, tiny little way. I'm

15:35

not talking about every company's changing the world. If

15:38

you want it to work, it's causing progress,

15:40

both add

15:42

to the vector of progress, but also maybe

15:45

just change some trajectory and some space a

15:47

little bit. I think if Shopify wouldn't have

15:49

happened something, this might

15:51

be a highly distributed, many, many pieces of

15:53

software or something else would be there.

15:55

It would not look like Shopify. The

15:57

world of computing is extremely powerful.

16:00

I mean, just like the other part of the world.

16:02

So you want to be able to add something. So

16:06

this is also why ignorance is usually useful, because

16:08

you should be ignorant to the low odds in

16:11

the beginning. I think one of the reasons why

16:13

at least some founders often are young,

16:15

these kind of things are

16:17

important. Another aspect of the European

16:20

Union, it's just like people

16:22

also study very long. I know this

16:24

has gotten sort of updated in the time since I was there, but

16:26

man, in my trips

16:29

home, I had a lot of 32-year-old

16:31

student friends. And that's just like, cool.

16:34

There's a significant amount of Nobel Prizes awarded to

16:36

people for their work in their 20s. And

16:39

we should just have a clear cultural

16:41

understanding that was used for years to

16:43

be out and building things. And

16:45

I think it's nothing single causal. And I

16:47

think there's a lot of contributing facts. I

16:51

would have trouble, I think, weighting them. But like,

16:54

optimism is a one, lack of optimism

16:56

is a one I would put on the top

16:58

of the list. What is

17:00

a German language word that you still

17:02

use when you think, because there's no

17:04

close English language equivalent? I

17:08

don't have... Heimwe,

17:11

right? That's a possible contender.

17:14

Heimwe. Seinzuch. Like, for Schlimbessung is

17:16

such a good word. Okay. And

17:19

the reason for this is by trying to improve something, you

17:21

made it worse. And it's

17:23

like, I mean, again, maybe it's also like

17:25

born out of pessimism about the future. But

17:27

like, it's just so wonderful because you see

17:29

it often. You know, some, you know, Chesterton's

17:31

fans, people don't often know what

17:33

parts of world, large system are important

17:36

parts that have a lot of cultural

17:38

or technical understanding coded in them. And

17:40

which ones are just there because we

17:42

were in a hurry building the system.

17:44

And sometimes you find out which is

17:47

which very quickly afterwards. A

17:49

vote is useful in these kind of

17:51

situations. That is a good word. How

17:53

about, ausseinandersetung? The process of

17:55

coming to terms with something rather

17:58

than just putting it out there. But they are. beautiful.

18:00

I mean, even gestalt is a word that isn't,

18:02

I mean, that might actually have been integrated now

18:04

in English, but like, it has no equivalent, like,

18:07

of that. But the reverse is also interesting. You

18:09

know, entitlement is not a word

18:11

that Germans have, right? Like, which I

18:13

find really, really interesting. I make, I

18:16

sometimes in Shopify have to explain to people when

18:18

I say bandons, like, hey, gratitude and entitlement are

18:21

two sides of the spectrum. Like,

18:23

it's your choice where you are here after

18:25

you have a pleasant experience, or maybe

18:27

with some downsides along the way. Very

18:30

important conversation to have with interns sometimes, if I

18:32

don't know, provided food is cold or something like

18:34

this. I get in these situations, I sit down,

18:36

okay, well, I'm doing this in German now, because

18:38

this is the moment where I have to roll

18:40

out this thing. And then I'm just like struggling,

18:42

because there's no term. And I'm like,

18:45

that's an interesting fact that we don't have a

18:47

word for this. So this happens as well. Now, of

18:49

course, someone's going to send me like

18:52

a string like this, which actually perfectly represents

18:54

it and does it better. But anyway,

18:56

nothing I could easily recall. Do you

18:59

still read books in German? Occasionally,

19:01

I have like, I must read

19:03

one book in German

19:05

a year as a sort of self policy,

19:08

which I have violated last

19:10

year, which I was not super happy

19:12

about. But like, one thing I find

19:14

disappointing is then, you know, like,

19:17

obviously, I take the opportunities when I want to

19:19

go to something that's originally written

19:21

in German that I kind of try

19:23

to read, read this in a, in

19:25

original. So like Wittgenstein's, Trakatus is an

19:27

example of this. It's like the

19:29

English translation is so much better,

19:32

because the translator asked for so many clarifications

19:34

by Wittgenstein that it just ends up being

19:36

like readable. So I found a book, which

19:38

was, was actually beautifully said, I think the

19:40

MIT published it of like, the original German,

19:43

the English translation and a translation back to

19:45

German from the English translation, all

19:47

in three columns. And that was perfect. So sometimes

19:50

the English translations also just

19:53

often get updated again, like for Kant or

19:55

something like this. And I do I'm

19:57

actually a fan of rewriting

19:59

books. like every, I

20:02

don't know, 25, 50 years for the next two

20:04

generations down the line because it's, they just

20:06

get hard to access. So often

20:08

you get that by reading the English translation. So

20:10

I'm trying to, I feel like I lose opportunities

20:12

to do this. In

20:15

your opinion, where exactly is the

20:17

dividing line between North Germany and

20:19

South Germany? So

20:21

people in Freiburg, they'll say like, oh, it's

20:23

Mannheim, but that's insane, right? Where

20:27

is it for you? You know, you asked me to poke

20:30

Horn's Nest here, which I, like,

20:32

I'm going to have too many unhappy people,

20:35

but actually committing at all to this question.

20:37

So I'm going to take a pass. I

20:40

would say Limburg is still South Germany because

20:42

historically it's been Catholic, but somewhere

20:45

not too far north of Limburg,

20:48

North Germany would start. It's very, it's very hard

20:50

to draw a straight line, though. You end up

20:52

with a very, very, very jagged line. I think

20:54

if you, if you're trying to do the best

20:56

possible job there, what's happened in Germany for a

20:58

long time, like to cause culture to be, to

21:00

be, to be classified to wherever it is now.

21:04

Now, Canada, in the data I see

21:06

right now, Canada seems to be having

21:09

a per capita GDP recession. And

21:11

I'm not sure how to interpret that. The

21:13

US has been growing at a decent clip.

21:16

Europe, more or less steady, growing at

21:18

a very slow pace. Why in per

21:21

capita terms does Canada seem to be

21:23

moving backwards? Is that a composition effect?

21:25

Or how do you read that? Yeah.

21:28

I mean, this bores me a great

21:30

deal. I mean, okay. So comparing to the

21:32

United States is a bad

21:35

idea in general. I mean, actually it's the

21:37

best possible idea if you're going for optimism,

21:39

but it's not the best idea if you

21:41

are looking for staying sane, right? Like America,

21:43

like, like America

21:45

is exceptional. It's an unbelievable economic might.

21:48

It's an unbelievable country in so many

21:50

ways. You hear in every country, well,

21:52

if you compare to United States in

21:54

this one thing, things are bad. I'm

21:57

like, well, I mean, I'm not I

22:00

mean, I think the comparison of Germany to Netherlands

22:02

is a lot, makes a lot more sense. I

22:04

think that's where you can make real, you know,

22:06

figure out what might be actionable. America is just

22:08

like really, really, really, really different. So first of

22:10

all that, every once in a while,

22:12

you know, there's an economy that really can hang with

22:14

America. I think 2000 years of 1455,

22:17

like it was looking really, really well. I

22:19

think what happened there, again,

22:21

nothing has a single, singularly causal,

22:24

but like it's, it's

22:26

the productivity numbers are just really low. And

22:29

I think the employment

22:31

in the public sector, having grown

22:33

the degree it has, it's just

22:35

like, again, I don't

22:37

know if it's, if it's

22:39

causally related or correlation with the same thing,

22:41

but like it's, it just

22:44

is pretty clear to me that if the

22:46

ratio of referees to, to, to, to builders

22:48

is all critics to builders is going out

22:50

of work that like things grind

22:52

to a halt. We

22:54

saw this even Shopify adoption, like it's,

22:56

it just took a lot longer for

22:59

Canada to, to, to want digital products.

23:01

And Shopify is always selling and

23:03

finding its best customers in

23:05

the United States for first 10 years easily.

23:07

And then you have already, like if

23:09

you believe me, that Shopify is

23:12

pretty good. Now there's a compounding advantage for people

23:14

who have adopted it earlier. And I think that

23:16

is sort of like a tiny zoom of a

23:19

much bigger fractal that I think is that play there.

23:22

How has Canada changed the most since you

23:24

moved there? It

23:26

feels like it's the optimism angle. And I

23:28

think this is like a thing that worries

23:30

me the most. It just, I think Canada

23:32

had a massively underappreciated at the time project

23:34

in multiculturalism that worked. And

23:37

it started under Pierre Trudeau really

23:39

just putting country together with great

23:42

leadership, great vision. And Canada had a string of

23:44

leaders that were like kind of almost too good

23:47

for a small country. You know, there

23:49

was like this, hey, they're getting a

23:51

long world, they're friendly, they're like, it's

23:53

a high trust environment. There's the best

23:55

days are going to be ahead of

23:57

us. And I think it's so hard

23:59

to. point at exactly what change did.

24:01

The cultural narrative has just simply shifted

24:04

now. I do think

24:06

people are a little bit more

24:08

circumspect and looking at the country.

24:10

I think the problems are more

24:13

clear. It's like there's

24:16

a lot hanging on real estate, and real

24:18

estate is not by itself. It's

24:22

valuable to a second order because of all

24:24

other sort of things being valuable. I

24:26

think the worst thing is, from my

24:29

perspective, is that Canada

24:31

seems to be OK just exporting

24:33

the raw materials for everything. And

24:35

that started as beaver pelts being

24:37

sent to London for turning into

24:39

high margin hats. And Canada

24:42

has no refineries. It all goes to Houston.

24:44

It produces a good deal of energy. Waterloo

24:47

is basically a raw material export, as

24:50

well as one of the greatest schools

24:52

in the planet. Waterloo students are lot

24:54

bearing for second value companies. There's a

24:56

lot of readiness to export the CO

24:58

cultural conversation about maybe we should

25:02

have pre-built things here, as well. It

25:05

seems like a country which has

25:09

so little self-confidence.

25:12

I see this as running Shopify. A

25:15

huge amount of our employees we hired last year,

25:17

I think it was 60% or so of our

25:19

engineers with boomerangs coming back from storied

25:22

American companies, most of them were

25:24

Canadian. Because like

25:27

they often say this because we wanted

25:31

to work for Apple because my parents said, man,

25:34

you're really doing well in tech. You might actually

25:36

get a real job in that Apple.

25:40

And if you hear this a lot, you do. And then you

25:42

go there. And you know what? You actually like to do that

25:44

Shopify. And then you come back. And I think this kind of

25:46

thing needs to happen. We need

25:48

to have some more of these stories out there.

25:50

Canada is a pretty good country. I think it's

25:52

a major, major asset to the United States as

25:54

a great friend. And I think a stronger Canada

25:56

is better for absolutely everyone. Do

25:59

you agree with the stereo? type that

26:01

Canada is especially weak when it comes

26:03

to branding. So there's Shopify, there's Molson,

26:05

you could say there's hockey, NHL,

26:09

but not that many Canadian brands.

26:11

Why is that? Or do you challenge the

26:13

premise? Because Canada exports some products. Shopify

26:18

and Lululemon export, sure, forgetting, I

26:20

mean, Molson, sure. But I don't

26:22

think it's not ability to do

26:24

branding. It's just Canada does not

26:26

appreciate commercialization of any kind. It's

26:28

like Canada wants to

26:30

invent. It's

26:33

remarkable how many papers that are foundational

26:35

to the current revolution of AI

26:38

Boom are in the University of

26:40

Toronto, Waterloo papers, Jeff Hinton and

26:42

his lab and Ben Gio and

26:44

so on. Canada

26:48

loves to have a Eureka moment. It's

26:51

seen as a low status thing to do to then go

26:53

and try to build a business around it. Which,

26:56

you know, is probably amazing from

26:58

a perspective of our neighbors. But

27:00

like, probably not so good for the,

27:02

you know, sort of wealth of a

27:04

country like it's not metabolizing any other

27:07

kind of innovation. Shopify is very much

27:10

an operation here as like, let's make

27:12

something that at least was a brand

27:14

of businesses initially and increasingly people beyond

27:16

businesses recognize it. So yeah,

27:18

like, I think that's an

27:20

important thing. Like, it's the same thing as

27:22

like, maybe you don't refine more, or we

27:24

don't make the hats for from from the beavers

27:27

or, you know, we just don't create the

27:29

final product, we send for raw materials everywhere.

27:31

And I would, if I could

27:33

change one thing, it's I would do that, I

27:35

would put like, and by the way, this is

27:37

deeply encoded in policy, right? So

27:39

there's a thing called SRED tax

27:41

credits, I'm not gonna bore you

27:43

with the details there. But like,

27:45

you can claim those to, if

27:47

you if you do, it says

27:49

research in development, try to claim

27:52

them for anything that's commercially related.

27:54

It's actually, it's remarkable. At

27:56

various points, stop just applying for them just

27:58

because Too

28:00

commercially in lots of ways so you're being

28:02

paid for doing original research which I think

28:04

is a great policy like because some like

28:07

often Original research kind of quote in your

28:09

boost But then

28:11

going and turning this into a product

28:13

you're like completely left out like if

28:16

you want to claim anything You will

28:18

have to ask every one of your

28:20

people in the staff to have meticulous

28:23

time sheets and submit an ungodly burden

28:25

of documentation which literally makes the commercialization

28:27

jobs terrible and so therefore You

28:30

know it's like the good people don't want to

28:32

book in this environment and so on so on so Do

28:35

you think in Canada there will be

28:37

an enduring backlash against immigration? I don't

28:39

mean the phony student visas Let's assume

28:41

that's taken care of but immigration as

28:43

it had been proceeding is that the

28:46

standing equilibrium? Or is that going to

28:48

dwindle and asymptote? So Canada is about

28:51

I don't know if it's accurate numbers But I think there's

28:53

a directionally right this is it kind of is about 41

28:56

million people in the last three years

28:58

three million people immigrated to Canada Which

29:00

you know is a significant percentage increase

29:03

in in size of population There

29:05

is a lot of cultural conversation about this

29:08

I think most of a conversation that I

29:10

see is not really about the veracity of

29:13

immigration in general But actually about like

29:15

the sort of fact that in

29:18

this time of adding three million immigrants We added

29:20

like almost no housing which is

29:23

like so that's you know Just that's

29:25

that's just like a not a great idea to do

29:27

that that's causing a lot of you know bad down

29:30

senior effects Immigration has my entire

29:32

20 years I've been here doing something

29:34

been very popular in Canada which I

29:36

thought was one of the most unique

29:38

parts of the country That's sort of

29:40

part of first statement. I made earlier

29:42

about an almost unacknowledged effortlessness to multiculturalism

29:45

that That worked the Canada

29:47

also implemented the thing that everyone's talking

29:49

about a skills based visa program at point

29:51

with point systems Which is well designed

29:53

and and has been doing a lot of

29:55

work for Canada in the past It's

29:58

not quite clear to me. Why? we

30:00

walked away from these priors that have

30:03

clearly identified the work, all of it. And

30:06

certainly, things

30:08

in policy land changed and opinions

30:10

changed. I don't think people

30:12

like this experiment. My significant hope is that

30:15

this is not going to be

30:17

one of us baby out with a bathwater moment

30:19

because we have a great skills-based immigration system. I

30:21

think China should just fall back on that and

30:23

run that up. Why

30:27

does Ottawa remain such an underrated

30:29

city? So Americans will take a

30:31

three-day trip to Toronto or Montreal,

30:33

but Ottawa is excellent. There's the

30:35

National Museum, very good food. Obviously,

30:37

it's nation's capital. Why

30:40

does it stay so unknown? But a wonderful setup.

30:43

And it's close, right? Thank you for the

30:45

platform. It's a wonderful city. We

30:47

ended up there because my wife was born there, but

30:49

then actually started there. Maybe

30:52

not expecting to go back. And then we stayed for 20

30:54

years and we built like Shopify built a

30:56

great company there because it's just like people

31:00

really love it there. And they were like itching for

31:02

better employees, I suppose. Yeah.

31:05

I mean, obviously, it also gets really

31:07

cold. I had these wonderful parts. My

31:10

commute to work was a canal, skating

31:12

on a canal every day. And

31:14

where else in the world can you have a commute like this? But

31:17

also, the first time I visited, it was really, really

31:19

cold in winter. And I was like, I sort of

31:21

had my questions. And then the summer you

31:24

go to a cottage and that's

31:26

just like the other side of the world thing.

31:28

And it's just like it's a wonderful quality of

31:31

life. And I think that matters. Don't you have

31:33

the world's largest outdoor ice skating rink and it's

31:35

seven kilometers or something? Yeah. That

31:38

was my commute. That was your commute, all

31:40

seven kilometers? Or how much of it? Well,

31:42

maybe two of us was a commute. It

31:44

was right office downtown. And then we were

31:46

living along the canal, like two blocks in.

31:49

And it was very cool. A very, very

31:51

nice thing. Nice way to stop a day.

31:54

I'm sure I wrote my best code those days. reality

32:00

stores matter? That's a good

32:02

question. Virtual reality stores, I

32:06

don't have a great answer

32:09

there. I don't think they

32:11

report the exact, like the Fifth Avenue

32:13

boutiques online other than for having virtual

32:15

twins for them, for people who specifically

32:17

want to see those. I don't think

32:19

the future of e-commerce is going to

32:21

be strolling through malls or virtual malls

32:23

or these kind of things. So I

32:25

think the exact story this store kind

32:28

of composes is going

32:30

to be different. I think the innovation and virtual

32:32

reality are going to be much

32:34

more about virtual avatars or real

32:37

people having you talking to the

32:39

product. Shopif represents mostly

32:41

the catalog of product that people really want

32:43

rather than the necessities. The Fifth

32:46

Avenue boutiques would also be the ones using Shop Point

32:48

of Sale. Purchases are a lot

32:51

more deliberate around this. People often spend weeks

32:53

thinking about this. This is something they would like

32:55

to purchase and they're really looking forward to the

32:58

package arriving, hopefully very quickly. I think

33:01

there's lots and lots of touch points

33:03

there. The place that is probably the

33:05

most virtual area that we see is

33:08

already furniture. Placing

33:10

the couch in your living room is just

33:12

better than looking at it in some

33:15

store. I think we see the

33:17

early innings in this. There's a couple of

33:20

technologies that we are tracking like Gaussian Splats

33:22

and these kind of things that are just

33:24

going to make it vastly simpler for people

33:26

to make digital twins available

33:28

from whatever they managed to put together in

33:30

real life. I think a bunch of this

33:32

is coming, but I don't

33:34

know what's the date and what exactly is

33:36

the form factor. Not

33:39

talking about back office, but actual

33:41

retail. Do you see in

33:43

advance how AI is going to be changing retail and

33:46

what does that look like? I

33:48

think it will play a significant role for sure. Change

33:51

retail, I think it will also, I mean, I think

33:54

we will see significantly better parts being

33:56

made. I do think so. I have

33:58

extremely bullish view. on AI,

34:01

specifically around the utilitarian value,

34:03

I think there's enormous advantages

34:06

for company building. I think there's enormous advantage for

34:08

product creation. I think what we

34:10

engineers experience around copilots just right

34:12

now getting good, that are

34:15

helping us do the job that we

34:17

already have a significant craft in, but

34:19

do it better, is extremely

34:22

convincing. And I think people would

34:24

want a copilot or a sidekick

34:27

or something like this along more

34:29

of the things they

34:31

do, which are like at the edges of

34:33

their ability, which I think in

34:36

a retail world on the creation side of businesses

34:38

and on the creation side of products, it's just

34:40

like basically all the time. It's a stretch. It's

34:43

a way you put yourself out there, you create

34:45

the best thing. It's a deeply personal thing to

34:47

create the first version of a product that we

34:49

try to create a company around. And

34:52

so I think that's really, really

34:54

powerful, like super, like highly intelligent,

34:56

very knowledgeable, zero

34:59

judgment, all that's available, fast returning,

35:01

even text AIs are

35:03

going to be fantastic. But

35:05

increasingly, I think software is going to

35:07

go through a rethink for this decade.

35:10

It's just quite clear that

35:12

most software, like

35:14

we have learned how to build

35:16

excellent user interfaces, quite approachable. Like

35:18

we simplify an enormously

35:21

complex space to easy to reason

35:23

about point decisions in a pretty

35:27

approachable and legible interface, which also look good.

35:29

And you know, like all these kinds of

35:31

things, that's sort of top

35:33

of the hill we've been climbing for 20

35:35

years or a little bit less since whatever

35:37

moment you pick in which web 2.0 started,

35:39

which really was the beginning of engineers saying,

35:41

hey, we figure out how to build applications

35:44

off the internet. And that's the string we've

35:46

been pulling on for all these years. And

35:48

we've built all very, very valuable companies that

35:50

basically replaces the like going directly to the

35:52

database or going to the command line and

35:54

we build these interfaces. I think now the

35:58

instead of creating a

36:00

place where someone can run around and

36:02

switch a whole lot of toggles and

36:05

change preferences to suit their

36:07

particular idea. I think people

36:09

can just tell us their goal. And

36:11

then we can work together on this. I think

36:13

goal oriented software is actually what we always wanted.

36:15

Because that's actually meets people where they are. It's

36:17

like how you work with colleagues together too. And

36:20

it's not, I'm actually really excited figuring out what

36:22

this is going to look like. I love the

36:24

time so ever significant transition, I thought that 2010

36:26

to 2020 was boring.

36:29

Because we basically just scaled the stuff we figured out towards

36:31

the end of 2000 to 2010 period. And so now we're

36:36

going to get into much more interesting times again, there's a

36:38

lot to be figured out. And that's exciting and interesting.

36:40

And I think we will end up is a much,

36:43

much, much, much higher amount

36:45

that couldn't have seen from the original hits,

36:47

right. And so that's, that's always exciting to

36:49

me. And I think it's going to be

36:51

very valuable to people. Now

36:53

you work with so many retailers, do

36:56

you feel you understand retail price stickiness?

36:58

Because economists don't a lot of economic

37:00

models imply prices are sticky. But when

37:02

they move, they should move a lot.

37:05

But you look at the data we have,

37:07

it seems that big and small price

37:09

movements are about equally likely, which means

37:11

we as economists are fools. How

37:14

well do you understand all this? I

37:17

don't think I have that line. Honestly, I just

37:21

like business are just so different, they are

37:23

hard to average out. And like, there

37:25

are a lot of businesses that

37:28

do their pricing strategy is aesthetics. And

37:30

aesthetics is one of those handwaves that

37:32

humans do, to explain away

37:35

enormous amount of background processing that goes

37:37

into it in the best case scenario,

37:40

like an entire career of knowledge rolled

37:42

into an intuitive quick decision, or

37:45

completely making it up like, like, sort of both sides

37:48

of like mid mid meme here. Like,

37:51

I think economics fundamentally will have to, you know,

37:53

roll a lot of data points into an average

37:55

and then try to see which direction people do.

37:58

And there's a lot of canceling each other. about

38:00

going on in the spaces that we

38:02

are concerned about. But sort of interesting,

38:04

we went for a high inflation period

38:06

and just tracking

38:09

when prices in the system were kind

38:12

of following that was deeply

38:14

different based on what kind of products

38:17

are and how people consider purchasing, buying

38:19

these products, obviously on margins too. So

38:22

again, I feel for the economists,

38:25

because I don't think, you know, I think

38:29

physics has given us a sense

38:31

that there is a simple equation underneath everything.

38:33

And we've built an aesthetic around this. And

38:35

I think often too many other fields want

38:37

to be more like physics. And I think

38:39

actually things are wonderful when they're complex. Like

38:42

I think, I don't know if you want

38:44

to talk about company building, but

38:46

companies are complex adaptive systems,

38:48

much more than being sort

38:50

of industry-applied versions of military,

38:53

slightly more complex organizations of military service. And

38:56

the recognition, which is not that old, right?

38:59

To a certain degree, it can explain why

39:02

companies run by some of

39:04

their engineering type people have been outperforming

39:06

things, because people have an incorrect understanding

39:08

what engineering is and how

39:10

it works. Engineering is fundamentally, at

39:12

least for the last 30 years, has

39:15

mostly spent time on trying to take non-deterministic

39:18

systems to make them indeterministic, which is kind

39:20

of what we do

39:22

in the real world with policy mostly

39:25

and process. So I think what we

39:27

can do now, if you're an engineer

39:29

running a company, I think you come

39:32

pre-equipped with ideas like systems thinking instead

39:34

of World War II organizational structures. Companies

39:37

are the ultimate non-deterministic systems that you're

39:39

trying to get to build fantastic products

39:41

at great pace, inclusive of all the

39:44

creativity by various actors in the company

39:46

and trying to

39:48

build inside of it a culture, a story,

39:50

incentive systems that are just making it so

39:52

that the maximum amount of everyone's

39:54

activity actually for the submission. I

39:57

find these things just so fascinating to

39:59

think about. But because this is

40:01

sort of going back to the beginning, I saw

40:03

myself as an outsider and I've made a study

40:05

of other fields from afar. There's

40:07

so many amazing ideas in basically any

40:10

given field that you can possibly name

40:12

or imagine. Like, and often what

40:14

happens is like every field kind of reinvents the

40:16

same core ideas and gives different names to them.

40:19

And making a study of doing this kind of thing

40:22

and just saying, okay, well, how do we build a

40:24

better company? I think companies are very bad, like all

40:26

of them. Like I think literally everyone, me and my

40:28

contemporaries, especially, we're gonna be terribly

40:30

embarrassed by the companies we ran in the early

40:32

2020s. And so,

40:35

because there's all these things we didn't yet have

40:37

or didn't yet understand or so, and then we

40:39

eventually will figure this out and then how could

40:41

we even build anything before we figured out this

40:43

thing? Again, I find that just

40:45

like such an interesting meta field of

40:47

research. It's almost a plight good thing.

40:50

What if we never figure it out? I mean, how

40:53

sure are you that in the future it

40:55

will be that much better? We'll have

40:57

better technology, but organization? Yeah,

41:00

we have new primitives on which we can,

41:02

like sometimes also philosophies, right? But I'm not

41:04

saying we're gonna build a company. There's no,

41:06

like everything is a set of trade offs.

41:08

I just like compared, like if the best

41:11

soccer team on planet earth gets at

41:13

like 80% of perfection, you freeze frame,

41:15

the replay, everyone, very few people use

41:18

a muscle that incorrectly while

41:20

actually approaching the wall and

41:22

then beautiful orchestration of cooperation

41:24

without zero communication. That's

41:26

like, what's a company? A

41:29

company is 5%. Like

41:31

how many memos are never read? And so,

41:34

which to me tells you, if you

41:37

just get to 6%, you're already doing better. Like

41:39

that's a pretty good way of not being as

41:41

embarrassed as everyone else. I think realistically

41:43

there's gonna be a limit because these are not the,

41:45

like soccer games the same one every time you can

41:47

actually practice for it. Like a company's, you know, every

41:49

day is a new day. It's a new puzzle box

41:52

dumped on everyone's desk. It's a different environment, but

41:55

still like I think companies are vastly better now.

41:57

Like when I started as an engineer even like.

42:00

uh, apprenticing under my Meister, he said, you have

42:02

two years after you start writing code for a

42:04

project, after which it's like someone puts a mental

42:06

note codebase and you're never going to change a

42:08

thing again. And that was just like

42:10

accepted back then. And now like we have

42:13

pieces of software that are 20 years

42:15

old and they're delight to work on because

42:17

we just built up these understandings. There's not

42:19

like a lot of these lessons in other

42:21

areas as well. What do

42:23

you think is the most common mistake your

42:25

third party retailers make in modeling the world?

42:28

Thinking they have to build products for that other people

42:31

like. I think this is the, this

42:33

is the sign killer. It's just like Shopify

42:35

is like someone's called it Kevin

42:37

Kelly's 1002 fans as they applied

42:40

at scale. Yes, everyone's different, but

42:43

actually there's clear clusters

42:46

and the people who are willing to dedicate

42:48

themselves to build a product and like, like

42:50

go through that entire rigmarole and put themselves

42:53

out there. Well, actually

42:55

inside of a cluster, more like the people everyone else

42:58

wants to be. And if they build

43:00

things that they would love to have in the world, turns out

43:02

that's they have extreme inside and

43:04

authority over this rather than running. There

43:06

was some cross pollination from the lean

43:08

startup book to the retail world. And

43:10

I think that's especially in retail. This

43:12

has been bad. I'm actually wondering

43:14

if that was such a good set of ideas. I think

43:16

that's good ideas in the book, but like it

43:19

feels a little bit ungenerant to just like I

43:22

think customers are not the people who

43:24

should say what needs to be built. I think

43:26

they need to explain their problems and it's like

43:28

the builders have to figure out how to solve

43:30

these problems better. I think that's otherwise an application

43:33

of vision. I think the best companies end up

43:35

like following a long term vision and a long

43:37

term mission. So I think that's part of it.

43:39

And then the people who have access to capital,

43:41

they under invest in growth. That always happens too,

43:43

which I know it's scary to do

43:46

internet marketing as getting harder and harder, but

43:48

like it's

43:51

not priced out yet. Did

43:53

you learn all this selling snowboards or

43:55

it took until Shopify? Okay, so

43:58

the greatest thing about running Shopify is

44:00

my customers are incredibly

44:03

inspiring individuals. It's like, in

44:06

a lot of places, it's very hard to convince

44:09

people to actually talk to their customers. It's actually

44:11

like sometimes our problem is the opposite. We have

44:13

too many people in conversations with

44:15

our customers. They are super

44:17

open to sharing what they see, and they're

44:19

delightfully discontent with what we give them. They

44:22

will tell us how to do this better

44:24

every single time. But they will also tell

44:26

us here's why, because

44:28

they are entrepreneurs. And Shopify

44:30

is honestly like a celebration

44:33

of the small bits of

44:35

capitalism. It's like we love entrepreneurship. I should

44:37

say plenty of our customers have started on

44:39

Shopify, and I know absolutely

44:41

massive billion dollar plus retailers.

44:44

So it's not like there's

44:46

a great variety spending pretty

44:49

much the entire spectrum of our retail industry

44:51

and size represented on the system now. But

44:53

a lot of the largest

44:56

people who started on the platform, now

44:58

has been around for 20 years. But

45:00

we love entrepreneurship. We love founding

45:03

the concept of companies as a self-expression.

45:05

You know, just there's glory in entrepreneurship.

45:07

And it's actually it's underappreciated. It's like

45:09

really, really, really, everyone talks about it,

45:11

and politicians always like are like pro

45:13

business formation. But like often the

45:16

behavior doesn't confirm to us.

45:18

It keeps getting harder to policy wise. I mean,

45:20

it's talk about, you know, that's certainly an aspect

45:22

in Germany as well. It gets hard and hard

45:24

to start companies in some places. Again,

45:26

the United States isn't the opposite. We are not

45:29

there. We have API's to start businesses, which is

45:31

exactly how things should be. Friction

45:34

changes the behavior a lot because

45:36

everyone's allowed to be an intelligent actor in

45:39

their local incentive system. And if you're massively

45:41

like disincentivized of starting a company, then by

45:43

just like BS, you have to deal with

45:46

and people want. So we want to be

45:48

a counterforce to that. We want to remove

45:50

friction where we can. Like again, we can't,

45:52

we can potentially advocate against bad policies, but

45:55

we can do a lot about what happens

45:57

after the policies stop mattering.

46:00

step because every single time we've made

46:02

Shopify more approachable or things that were

46:04

previously complex and gating for people's success,

46:06

every single time we made

46:09

something significantly simpler, it actually caused

46:11

more success. More people otherwise didn't

46:15

make the hurdle and ended up making

46:17

it through it. And not stopping is

46:19

actually the thing that really, really leads

46:21

to success in a reductionist way. So

46:24

we find that just to be a really, really important

46:26

discovery. Again, when we

46:28

are part of a journey, I like to

46:30

create the center systems and the business system

46:33

of Shopify in such a way that we actually are on the same side

46:36

of the table with our customers.

46:38

The best thing we can do to

46:40

grow Shopify is make our customers more

46:42

successful because we're in this together, economically

46:44

speaking. So they take an active role

46:47

in talking to us.

46:49

Every one of the product managers has hundreds

46:51

of active, what's up conversations with fast growing

46:53

businesses. And I think that's a

46:56

really, really, really fun way to build a business.

46:58

It's a very, very, very rare

47:00

thing that your customers are often the source

47:02

of your inspiration. What's an

47:04

interesting book you've read lately? Interesting

47:07

book, I think. I've been sort of on

47:09

a fantasy kick, which is not super conducive to

47:11

that. But what do you learn about management from

47:14

reading fantasy? You read

47:16

Sanderson, right? Am I correct? That's

47:18

right. That's right. I read Sanderson.

47:20

They're long. Very rare commitments. My

47:23

son read the entire Stormlight archives for books on

47:25

March break and I've such reading speed envy since

47:27

then. It takes me a lot, a lot longer.

47:29

I mean, I think fantasy

47:31

is full. I mean, fantasy is very

47:34

often a mirror to society in some

47:36

reductionist way. So it's a sort of,

47:38

it's a simulation. It's a simpler scenario

47:40

with some variables changed. And any book

47:42

that you don't like toss across a

47:44

room is a book that usually has

47:46

realistic characters that have some depth to

47:48

them following their story, given sort of

47:50

the changes in environments. Fantastic. I mean,

47:52

obviously a lot of rings is an amazing

47:54

match. If you will, like the

47:57

way Gandalf shows up in just the right time.

48:00

exact right combination of words is certainly something

48:02

that's extremely valuable. At least it

48:04

conforms to the best version of the

48:06

business systems that we used to build.

48:08

And I think that's really valuable. That's

48:10

interesting. Seeing a state is just a

48:12

fantastically book. He just passed

48:14

away. You probably saw that on Twitter. It's

48:16

yesterday or so, a couple days ago. Yeah,

48:18

what a fantastic mind. It's one of those

48:21

books you read which feels like it should

48:23

be a book, particularly

48:25

a good book in a space of lots of books. But

48:27

it seems like totally, there seems to be nothing around it.

48:29

It's just like... I've had

48:31

people ask me, oh, recommend to me other

48:33

books like Seeing Like a State. I'm not

48:36

sure what to say. Exactly. There's nothing quite

48:38

like it. It's really wonderful. I've been fascinated

48:40

with the Burnham books. You mean Managerial Revolution?

48:42

And Machiavellians are just like, especially for when

48:44

they were written incredible books. The degree by

48:47

which we have known a lot of these

48:49

kind of things but haven't known the solution

48:51

to some of the things that Burnham discusses

48:53

are just remarkable. I think

48:56

the best book I've read, was A Conflict

48:58

of Visions recently by Thomas Sauer. I just

49:00

find that is an incredibly insightful book

49:02

by creating a prior to a lot

49:05

of political conversations, a

49:07

higher order differentiation between people. I've just

49:09

found myself to be someone who... I

49:13

fundamentally think humans are limited and that's the best

49:16

thing about humans. And

49:18

that systems that we can

49:20

build can lead us to

49:22

incredibly amazing feats of cooperation,

49:24

coordination and optimizing

49:26

everything you do to find the best set

49:29

of trade-offs feels like an extremely mature way

49:31

of seeing the world. I'm just

49:33

constantly fascinated with Sauer's writing. Final

49:36

question. What is it you hope to learn next?

49:39

I have been on a wonderful

49:41

reconnecting with engineering kick for us

49:43

a little while. I've really, really

49:45

had a great time coming back.

49:47

I think co-pilots and AI has

49:49

allowed me to like mitigate

49:52

all the downsides of just not

49:54

spending any like weekends and book

49:56

on computing projects. Again, I'm incredibly interested.

50:00

I'm interested in LLMs, Transformer, the machine

50:02

learning world. It feels

50:04

almost unending well for information

50:06

and focus. So I'm just

50:09

fairly tapped on this. It's hard

50:11

to say this as a field

50:13

because it changes every couple of days. I

50:16

am most thinking about, and I really feel like

50:18

I can contribute a little bit

50:20

to just thinking about how to have higher trust

50:22

companies at scale. I just think people make a

50:24

lot of... People

50:27

describe as I'm a small company person, I'm

50:29

a big company person. I just don't think

50:32

what's actually the differences. I think people are

50:34

using labels around something they feel and they

50:36

haven't got the right words. It is probably

50:38

many things that contribute to this. But

50:41

certainly parts of these things are a sense

50:43

of agency and a bit ability to impact.

50:45

A lot of what happens in companies is

50:48

that policies and processes, they're

50:50

well-meaning. They bring up the floor so no

50:52

one actually does something really, really wrong. But

50:55

what people don't see is they also bring down the

50:57

ceiling. And so you end up in these places where

51:00

it doesn't matter who you are, you're going to do seven out

51:02

of 10 work. And just

51:04

the whole concept of entrepreneurship is about going

51:06

for world class and people need

51:09

to leave in many places after a company gets

51:11

a certain size. I really

51:13

think this is simply a path dependent,

51:15

unacknowledged situation that just comes from the

51:17

way we have... The tools

51:20

that we had to coordinate

51:22

in people and everyone holds

51:24

their dismissal of trust as a part of

51:26

this. I think the

51:29

best areas of a company like

51:32

Shopify keep the ceiling open so

51:34

that everyone can reach as high

51:36

as they have ambition for. Sometimes

51:39

teams come together and just do absolute work class.

51:42

That means you have to be willing to accept

51:44

underperformance as well. Like a floor can't be quite

51:47

so high. So sometimes you get something that you

51:49

can't trip. Sometimes you get something

51:51

in the wrong direction. This all comes with

51:53

a territory. And businesses tend to think about

51:55

these things as disastrous events and will do

51:57

everything to not experience this and therefore stay in the

51:59

room. them a all creativity, I think

52:01

we are now gaining

52:04

tools and approaches that can

52:06

do this at scale,

52:08

just like what engineers experience with writing

52:10

code. And then you have a co-pilot

52:12

that helps you write code well, given

52:15

what you're working on and quickly

52:17

gain the insights that you need and the

52:19

task, the area understanding, and

52:21

then after you are done with it,

52:23

there's automated systems that test, there's like

52:25

automated linting, automated unit tests and so

52:27

on. It's to me,

52:29

this, I know this sounds incredibly nerdy,

52:31

but what this basically is, is trust

52:34

plus co-pilot and automated

52:36

verify. And so as a take on

52:38

the trust, but all trust that verify

52:40

thing, I think that creates

52:43

a wonderfully fun environment. It

52:45

turns working on areas into,

52:48

you know, almost video game-ish. And I think

52:50

we know how to build these systems now.

52:52

And I think we can build enormously better

52:54

companies this way. They're just more fun for

52:56

everyone. And I also lead to just better

52:58

products. And I just, this

53:01

is clearly possible because I've seen it be possible. And

53:03

we just have to come up

53:05

with a couple more ideas along those ways. And

53:07

we have to figure out the particular downsides because

53:09

again, nothing is perfect. Just different

53:12

sets of trade-offs. I think the different,

53:14

the trade-offs of building a

53:16

company this way, rather than just reducing it

53:19

to zero trust and mechanize everything is enormous

53:21

for society and like just productivity and

53:23

just like fun at work. And yeah,

53:25

so I'm excited about that. Toby

53:28

Luttke, thank you very much. Great conversation. I really

53:30

enjoyed this. Thank you, Tyler. Thanks,

53:53

John. And the show is At Cow &

53:55

Convo's. Until next time, please

53:57

keep listening and learning. you