AI transcript
0:00:02 (upbeat music)
0:00:04 When Shopify was founded in 2006,
0:00:07 it was just a year prior to the launch of iPhone.
0:00:10 Needless to say, smartphones were a far cry
0:00:11 from the ubiquitous accessory
0:00:13 that now billions carry around.
0:00:17 And fast forward to today, nearly 20 years later,
0:00:19 where Shopify merchants can accept cryptocurrency,
0:00:21 create immersive shopping experiences
0:00:22 with augmented reality,
0:00:26 and even use AI to generate product descriptions.
0:00:29 So how does a nearly $100 billion company
0:00:30 stay ahead of these trends?
0:00:33 And what will shopping look like in the generations to come?
0:00:36 Plus, why would A16Z never have invested in Shopify
0:00:38 in its early day?
0:00:40 And why has the CEO of this company
0:00:42 chosen to continue writing code?
0:00:44 These are just some of the questions covered
0:00:45 in today’s episode,
0:00:48 the conversation the Shopify co-founder and CEO
0:00:52 to bias Luki and A16Z co-founder, Ben Horowitz.
0:00:54 This episode was also originally published
0:00:57 on our sister podcast, Web3 with A16Z.
0:01:00 So if you are excited about the next generation
0:01:03 of the internet, find Web3 with A16Z
0:01:04 wherever you get your podcasts.
0:01:06 All right, let’s get started.
0:01:14 – Welcome to Web3 with A16Z,
0:01:17 a show about building the next generation of the internet
0:01:20 from the team at A16Z Crypto.
0:01:22 Today’s episode features Tobias Lutke,
0:01:26 CEO and co-founder of the e-commerce platform Shopify.
0:01:29 Speaking with A16Z co-founder, Ben Horowitz,
0:01:32 at our second annual founder summit in November,
0:01:35 they discuss what it takes to build a breakout startup
0:01:39 in a crowded category, the changing face of retail,
0:01:41 how to affect change in the workplace,
0:01:44 and how to handle individual emotions and corporate culture,
0:01:47 including dealing with calls for activism,
0:01:50 as well as the value of embracing negativity.
0:01:52 They also touch on the moral imperative
0:01:54 behind creating quality software,
0:01:58 the symbiosis between AI and crypto, and more.
0:02:01 As a reminder, none of the following should be taken
0:02:04 as business, legal, tax, or investment advice.
0:02:07 Please see A16Z.com/disclosures
0:02:08 for more important information,
0:02:11 including a link to a list of our investments.
0:02:19 – All right, so for those of you who don’t know Toby,
0:02:20 which you probably should know Toby,
0:02:22 you can look him up as well,
0:02:27 but he started a snowboarding company
0:02:29 in 2004 called Snowdevil,
0:02:33 and he parlayed that into Shopify,
0:02:37 which now serves millions of merchants
0:02:40 in 175 countries around the world.
0:02:42 How many countries are there, like in total?
0:02:44 – I think that’s most of them, there’s 200?
0:02:45 – Yeah, 200, yeah, but–
0:02:48 – There’s changes around the margins.
0:02:51 – You’ll get them all at some point.
0:02:54 And then he was also interesting for this crowd.
0:02:58 He was one of the first adopters of Coinbase Commerce.
0:03:02 So Toby’s been kind of big on crypto for a while now.
0:03:04 But let’s get started.
0:03:08 So you started as a snowboarding company,
0:03:12 and I always find that funny because of all the CEOs
0:03:16 that I know you’re probably the least snowboard kickback,
0:03:19 like let’s hit the powder dude.
0:03:21 So how did that happen?
0:03:23 Like why did you start out that way?
0:03:28 – Yeah, you know, when you talk to governments,
0:03:32 they always, Shopify is doing a lot about entrepreneurship,
0:03:35 which I was a lot of entrepreneurship,
0:03:38 and it’s such a brilliant institution.
0:03:39 I know everyone here is a card-carrying member
0:03:42 of the entrepreneurship is good club,
0:03:45 but politicians usually, they’re sort of vague on it,
0:03:50 but they say like they want to pass government programs
0:03:52 and policies to cause more entrepreneurship,
0:03:55 which I always think is the wrong way around.
0:04:00 But there’s actually, so the funny story is that,
0:04:03 so I was in Canada at this point, I came from Germany,
0:04:06 living in Canada 2004, three, four,
0:04:10 and I found out I cannot actually get a job.
0:04:13 I was trying to be hired as a computer programmer locally,
0:04:15 but I found out I did not have a work permit.
0:04:16 I didn’t know what a work permit was.
0:04:20 So turns out that you are in Canada,
0:04:22 allowed to start a company even without a work permit.
0:04:23 So there’s one government program
0:04:27 that actually has led to what company is being created.
0:04:29 – Well, that’s like the best government program ever.
0:04:32 Just like the payoff on Shopify paid for itself
0:04:33 a million times.
0:04:35 – And it turns out Canada is very cold.
0:04:37 So I was doing snowboarding, so I knew something about it,
0:04:40 and I started an online store for snowboarding.
0:04:42 It’s kind of a pretty simple story like this.
0:04:44 – That’s amazing.
0:04:47 So basically the key to your whole career
0:04:49 was not being able to work.
0:04:49 – Yes.
0:04:51 (laughing)
0:04:53 In fact, you know, what was in my mind too is like,
0:04:56 I was like the solution with programming at this time.
0:04:57 It was like sort of a Java word,
0:04:59 kind of was kind of very oppressive.
0:05:04 And I wanted to reclaim it as a hobby.
0:05:06 So I was like, cool, if I sell snowboards online,
0:05:08 I can make money while snowboarding,
0:05:09 and I get to do more fun stuff.
0:05:14 And it’s been extremely unsuccessful as a strategy.
0:05:17 I basically never had time to snowboard again
0:05:17 after I started.
0:05:20 (laughing)
0:05:22 – The end of your snowboarding career
0:05:24 was the start of your snowboarding business.
0:05:25 That’s pretty remarkable.
0:05:27 – Maybe I would project more back at this point
0:05:29 if things would have gone in the counterfactual way.
0:05:31 – Yeah, now the other thing that really strikes me
0:05:34 about Shopify is that, you know,
0:05:36 and I hate to say this as a venture capitalist
0:05:38 who some people think is a smart person,
0:05:42 but we probably would have never invested in Shopify
0:05:43 when you started it
0:05:47 because it was probably the most over-competed
0:05:48 category in the world.
0:05:52 Like everybody had some kind of e-commerce platform.
0:05:56 I mean, there was just so, so many of them,
0:06:00 Magento, this, that, the other.
0:06:05 So did you go, oh, no, no, no, like,
0:06:07 I’m gonna build this thing.
0:06:09 It’s gonna dominate everything.
0:06:10 Then I’m gonna have a network effect
0:06:12 and I’m gonna build, you know,
0:06:14 like one of the biggest things that anybody’s ever seen.
0:06:19 Like, how did you like get all the way to where you are?
0:06:20 Like, what was that path?
0:06:24 – Yeah, no, there’s no plan like this.
0:06:29 Just like, I think, for every entrepreneur,
0:06:31 it’s a useful thing to think about
0:06:33 what is your energy source, right?
0:06:35 Like what is actually driving you?
0:06:37 And sometimes it’s greed.
0:06:40 And we, you know, like the sort of negative emotions
0:06:42 channeled into building actually
0:06:44 the most powerful energy sources.
0:06:47 I had this, you were right.
0:06:48 Like, I mean, Netscape, IPO,
0:06:52 but like, you have to explain every platform
0:06:54 in terms of an old platform and, you know,
0:06:57 online hosted CS catalog with buy buttons
0:07:00 was like kind of a way you would mark
0:07:03 to describe even what the internet would be used for, right?
0:07:04 Like, so e-commerce was like right there
0:07:07 from the beginning, I thought there would be lots of software
0:07:09 that I could use and I didn’t find anything.
0:07:12 I was actually really upset with the quality of the software.
0:07:15 And so like, I channel that into building something
0:07:17 and that was my energy source.
0:07:19 And I just love the building.
0:07:20 – That’s actually, by the way,
0:07:23 one of the best entrepreneurial ideas there is,
0:07:25 which is this stuff sucks so bad.
0:07:27 Yeah, I’m so mad about having to use it
0:07:29 that I’m going to build my own.
0:07:32 – Totally, I honestly, it fuels me to do it this day.
0:07:34 I’m just, I still think it’s ridiculous
0:07:37 what people are subjected to in terms of quality of software.
0:07:39 It’s like, oh, it’s weird.
0:07:41 I probably did some like neurological,
0:07:43 like kind of rearrangement there,
0:07:44 but like to me, this is almost a moral issue
0:07:48 of like when people like confused with software
0:07:50 and they think about themselves as like being inadequate.
0:07:53 It’s like, how dare software make humans feel that way?
0:07:54 Right, like it’s just not okay.
0:07:57 So you want to make it a more approachable,
0:07:59 but like, I think the thing that people might have missed
0:08:02 in e-commerce or like, it’s not that I had a great insight.
0:08:05 I just sort of organically got there is bad.
0:08:06 Yes, there was a lot of e-commerce software,
0:08:09 but it was all built for already rich retailers
0:08:11 who needed to move online.
0:08:13 Then that was a phrase people used.
0:08:17 And there was no one started software
0:08:20 for starting more entrepreneurial process online.
0:08:23 Like actually the digital native, what we call it now also,
0:08:26 like if you like that term,
0:08:29 because there’s a totally different set of requirements
0:08:30 for the people who are just starting out.
0:08:32 And so that’s what we did.
0:08:33 So this is the thing, right?
0:08:38 Like it’s like, when you talk,
0:08:40 that’s like, I don’t know, 60 words per minute or something,
0:08:42 and you type fast, it’s like twice that.
0:08:45 But like, if you can do customer consultation
0:08:49 in your own brain, the brain runs at like terabytes per second
0:08:50 by bandwidth, right?
0:08:53 Like so, if you know the use case,
0:08:56 if you’re building for yourself or your past experience
0:08:58 or some kind of, if you have a really good model
0:09:01 of your customers or use cases in your mind,
0:09:03 you can do customer development at high bandwidth
0:09:06 with low latency and you can just build something
0:09:08 that you know needs to exist.
0:09:10 – Yeah, no, that’s incredible.
0:09:13 So one of the things, you’re also kind of been
0:09:19 a kind of a special kind of, or a different kind of CEO
0:09:24 than some of the ones that people read about
0:09:27 such as Adela or Bob Iger and that you have a reputation
0:09:31 for being extremely hands-on, including writing code.
0:09:34 You still have the fury that started the company.
0:09:38 How do you think about that in terms of,
0:09:40 you’re not kind of paying attention
0:09:42 to all parts of the company equally.
0:09:46 You’re clearly doing certain things and not others.
0:09:49 How do you think about that as an effective
0:09:51 kind of management style?
0:09:56 – Yeah, I mean, I imagine the people in my company
0:09:58 would disagree with your assessment
0:10:00 about some effective management style.
0:10:02 So well, it’s worked pretty well.
0:10:07 It’s working, but it certainly doesn’t proxy
0:10:11 to what people sort of, like it’s not a CEO job
0:10:14 out of central casting that I’m trying to perform, right?
0:10:17 I tried that on and it didn’t work for me at all.
0:10:20 I actually find, like I value being in all the details
0:10:23 and I ask it from all the people who are around me
0:10:25 and report to me, like almost everyone,
0:10:29 especially after COVID, I turned over everyone,
0:10:33 but one of my executives during the first 10 months of COVID.
0:10:35 And now almost everyone reports to me,
0:10:38 it’s actually a AX founder, maybe someone be acquired
0:10:41 or someone who started a company before coming to Shopify.
0:10:45 So because there’s this, I just want,
0:10:48 I find it actually really inefficient,
0:10:50 not understanding the details of what we’re doing, right?
0:10:51 Because then–
0:10:52 – Well, then you’re guessing.
0:10:53 – Then you’re guessing, right?
0:10:57 And you’re like, well, when something goes wrong,
0:10:59 now you not only need to fix it,
0:11:02 you actually have the first cram for,
0:11:04 like you have to learn like three months
0:11:05 before you can actually do a fix.
0:11:06 If you understand what’s going on
0:11:09 or what ought to be existing, it’s much faster.
0:11:13 And you want to train people on the company mission, right?
0:11:16 Like because there’s a thing that you want to exist,
0:11:18 like even if you don’t know exactly what it is,
0:11:21 like you know which direction you want people to go, right?
0:11:24 And getting everyone aligned to go in this direction
0:11:27 is only possible if you can paint picture
0:11:30 of like how does this area actually sum together
0:11:33 all the way to helping the mission overall.
0:11:38 And so, and this is sort of a sort of unique thing
0:11:39 about our times right now.
0:11:41 It’s like the infrastructure keeps changing,
0:11:44 like here’s crypto, here’s AI, here’s,
0:11:46 I mean, I started again almost,
0:11:48 like next year is 20 years ago, right?
0:11:51 So there was no mobile phone, like that offer quality
0:11:53 that we have now.
0:11:54 – Right, that’s right.
0:11:55 You started before the mobile phone.
0:11:58 – Right, so it’s like, that was one of the things
0:11:59 that just sort of happened.
0:12:03 And, you know, SaaS software was new at this time.
0:12:05 So, you know, the platform’s capabilities change
0:12:08 and you want to adopt them to like your mission
0:12:10 to what is now possible.
0:12:13 And so you have to be able to understand, well,
0:12:15 but in a kind of like ideally in your head,
0:12:21 in the world where, I take a very long-term view
0:12:24 and my biggest fear that I have
0:12:27 is that Shopify winds up being a fantastic solution
0:12:30 to a problem that people no longer have, right?
0:12:32 Or no longer solve this way.
0:12:33 So I’m trying to like need to keep it current
0:12:35 and being able to run the counterfactual about
0:12:39 this is not possible, this is what we’re trying to accomplish.
0:12:41 But if this would have been possible from the beginning,
0:12:43 we would have built this entirely different thing.
0:12:45 So now our job’s to get from here to there
0:12:48 because, again, we want to keep it current and idea.
0:12:49 – Right, you know, that’s such a great insight
0:12:53 because they think that, you know, when I work with CEOs
0:12:55 and you think about the job,
0:12:58 so much of the job is making high quality decisions
0:13:00 and setting the direction of the company.
0:13:03 And in some respects, neither feels like work.
0:13:07 You know, because you’re working on something else,
0:13:10 like how does AI work, what’s possible,
0:13:13 in order to make decisions and set the direction.
0:13:16 But what you’re doing seems actually removed
0:13:20 from anything that the company needs to do right now.
0:13:22 And I think so many people kind of neglect
0:13:24 that kind of thing for that reason.
0:13:27 The other thing that I always find is,
0:13:29 at any given point in time,
0:13:32 some things in the company are very important
0:13:33 and very high leverage.
0:13:35 And some things are just not important
0:13:37 for you to put your time into.
0:13:39 And people often get lost in that.
0:13:43 They think, oh, I’ve got to talk to this person
0:13:46 and then that person and then the other person and so forth.
0:13:48 And it’s like, no, no, you don’t.
0:13:49 You’ve got to make good decisions.
0:13:50 You have to set the direction.
0:13:52 And if you don’t do that, nobody’s doing that.
0:13:54 And then the company’s going to fail.
0:13:57 So that’s such an amazing way that you’ve gone about that.
0:13:59 – If you’re in a small team,
0:14:01 like for a new thing that your company needs to do
0:14:04 and you’re in a small team of people who are like,
0:14:08 know or like, figure out the shape of it early,
0:14:09 if it’s actually as important,
0:14:11 it will itself have a gravity
0:14:14 to change the way the company thinks about quality of software
0:14:16 shipping, the way the architect thinks.
0:14:22 And it’s also after this comes up as a new thing,
0:14:26 you then have the ability to like a type of legitimacy.
0:14:27 As a founder,
0:14:30 you have a huge amount of legitimacy in a company already.
0:14:33 But like if you also are there with the new project,
0:14:35 that’s important, that needs to be a resource and say,
0:14:38 this is what we’ll do.
0:14:40 You don’t have to have all these conversations.
0:14:43 People love clarity of like, this is what we do.
0:14:45 Like it has a gravity all to itself.
0:14:48 That’s unlike anything you can accomplish
0:14:50 by one-on-one consultation,
0:14:53 trying to talk everyone through the entire decision matrix.
0:14:56 And so what is Toby doing over there?
0:14:57 – Oh, that must be important.
0:14:59 – Yeah, that’s exactly right.
0:15:01 We’re doing like a job site click now.
0:15:04 And like, then I don’t have to figure out like,
0:15:06 yes, there’s like 15 teams that are already doing AI stuff
0:15:09 and they ought to be in some kind of constellation.
0:15:12 But now this AI assistant thing is going to run through
0:15:16 and the company, everyone wants to be a part of it.
0:15:18 And this is like, all these problems kind of disappear.
0:15:19 It’s like, it’s maybe not the most effective-
0:15:20 – Like a magic trick.
0:15:21 – Yeah, yeah, yeah.
0:15:23 Yeah, actually, I think by the way,
0:15:25 like everyone in this,
0:15:28 like everyone here has clearly read your books.
0:15:32 And if you haven’t done do that, like I think-
0:15:33 – Thank you.
0:15:35 – The best books on leadership, honestly.
0:15:38 Thank you so much for sharing all this.
0:15:40 I think in some of what you described,
0:15:43 but there’s only very few ways to do a change management.
0:15:45 One way of doing it is to something super, super,
0:15:49 super surprising and then explain everyone why you did it.
0:15:51 It’s the fastest way to get a large company.
0:15:53 In my shop, there’s about 10,000 people,
0:15:55 much, much, much faster way to do a change management
0:15:56 than hand-to-hand combat.
0:15:58 – It’s like shock therapy.
0:16:00 It’s like, what the fuck was that?
0:16:02 Oh, well, this was that, that was okay, got it.
0:16:06 Very, very, very high retention on that.
0:16:07 So let’s talk about, you know,
0:16:10 one of the things that has come up, you know,
0:16:13 a lot over the years that you had probably, you know,
0:16:16 maybe the best statement on, I thought,
0:16:20 is politics, you know, politics, activism,
0:16:22 how that’s kind of come into companies
0:16:25 has taken over a lot of the activity of companies.
0:16:29 And how do you, I guess, how do you think about that?
0:16:33 And then why’d you say what you said?
0:16:35 And then how’s that going?
0:16:39 And then how are you able to sustain it particularly
0:16:42 as things get even hotter now with, you know,
0:16:44 what’s going on in the Middle East and so forth?
0:16:47 You know, it’s gotten even more intense than it was,
0:16:49 you know, amazingly in 2020.
0:16:51 – That’s a complex topic.
0:16:56 I can, here’s, I mean, there’s a lot of conflicting ideas
0:16:58 about what people want companies to be.
0:17:02 And, you know, I think, especially in 2010
0:17:06 and maybe now again, a lot of people would like companies
0:17:08 to play the role of some kind of phantom limb
0:17:13 of what I think the whole left over by the stories
0:17:16 that people are surrounded with, like maybe with governments
0:17:17 or like, like, and so on.
0:17:20 I suddenly, people want, like I have gotten tons
0:17:23 and tons of petitions for Shopify to take stances on issues.
0:17:27 And I think everyone’s feeling the same thing.
0:17:30 This was weird to me, but like initially,
0:17:32 but like, I found myself generally agreeing
0:17:34 with what people identify as the problems
0:17:36 when this happens.
0:17:38 But I found myself, they really, if ever,
0:17:41 really agreeing with the proposed solutions
0:17:41 to those problems.
0:17:44 And I think it’s kind of important to say like,
0:17:46 hey, unless we are all aligned on this,
0:17:48 we’re probably not doing this.
0:17:51 Because, you know, at the end of the day,
0:17:53 like companies ought to be thought of
0:17:54 as a bit of a simpler thing.
0:17:57 It’s, you know, real love entrepreneurship.
0:17:59 Entrepreneurship is, you know,
0:18:01 liberal values kind of institution of, you know,
0:18:05 you’re like reaching for independence,
0:18:07 reaching for like, to better yourself,
0:18:09 like building something like enlightened,
0:18:10 the ideas of enlightened self-interest.
0:18:13 There is a bit of a political coding in this,
0:18:14 people at least think so.
0:18:16 And someone else might take a totally different opinion
0:18:17 about all these kinds of things.
0:18:18 So that’s cool.
0:18:20 The cool thing is companies themselves explorations
0:18:22 into a set of ideas.
0:18:24 Sometimes encoded in a mission sometimes by the founders.
0:18:27 And if we make every company the same,
0:18:29 then we lose that.
0:18:31 That’s actually like a weird form of diversity itself,
0:18:32 if you think about it.
0:18:36 So anyway, I went ahead and just like said,
0:18:40 like look, we cause more entrepreneurship to exist.
0:18:43 And sometimes there are social causes
0:18:45 that are aligned with this.
0:18:48 And then maybe we become active because of our mission.
0:18:50 But outside of that, you know,
0:18:53 everyone makes money and you can do whatever you want
0:18:54 in your after hours.
0:18:58 And that’s just a simpler way of thinking about the business.
0:19:02 Of course, that’s not universally loved,
0:19:05 but like after being there, they clear about it
0:19:05 and explaining it.
0:19:09 They explain, you know, we give everyone when we hire people
0:19:11 a letter of like, here are the reasons,
0:19:13 like before you sign.
0:19:15 Here’s the reason why you might not want to work
0:19:18 for Shopify because like here’s a set of ideas
0:19:19 that you disagree with.
0:19:22 And then that clarity is wonderful.
0:19:24 We keep everything so nebulous.
0:19:28 I think the world of marketing and PR
0:19:31 has just done a number on us all to be so non-committal
0:19:32 to everything.
0:19:33 – Yeah.
0:19:35 – And the problem is if you’re non-committal,
0:19:37 you leave a lot of vacuum.
0:19:38 – Right.
0:19:40 – And that vacuum is going to be filled by someone.
0:19:43 And it’s usually the people who want to fill it
0:19:46 with bad faith takes, who fill it.
0:19:50 So if you just clear about it, it works sometimes,
0:19:54 like this is something I might get into hot water with,
0:19:57 but like sometimes it’s actually good to heighten
0:19:58 the misalignment for the kind of people
0:20:00 that you don’t think should be working in a company.
0:20:04 Like, hey, we work hard and it’s kind of an-
0:20:06 – Yeah, we have poor work-life balance.
0:20:07 – Yeah, yeah.
0:20:08 – Sorry.
0:20:09 – That’s a good-
0:20:11 – You are free to work elsewhere.
0:20:15 – If it’s true, put it up, like light it up on a sign
0:20:17 because otherwise you’re going to end up
0:20:20 with the weirdest divisiveness in the company
0:20:22 because of basically false advertising.
0:20:25 So I think it’s better to just be clear.
0:20:29 So I mean, what is it like?
0:20:31 There’s a lot goes into this.
0:20:33 Like again, culture is unstable.
0:20:36 We know from the internet that every community
0:20:40 that anyone likes Reddit, like Hacker News,
0:20:42 if anyone actually likes this page,
0:20:47 it’s a product of unbelievably dedicated moderators
0:20:51 who are keeping this discourse well.
0:20:54 I think that’s an idea that I think companies
0:20:55 have started adopting too.
0:20:57 Like once you’re a couple of thousand people,
0:20:59 your Slack channel is the internet basically
0:21:02 and you need to apply similar rules.
0:21:04 You need to be able to take people to a sign saying,
0:21:05 hey, here’s the thing you’re causing.
0:21:07 When you’re sending a message to a channel
0:21:09 with 3,000 people in it,
0:21:11 that is basically the same as you spending a day
0:21:15 typing individual text message to any person in the channel.
0:21:16 And would you do that?
0:21:18 If no, probably don’t put in this kind of stuff.
0:21:22 Yeah, that’s really such a great insight that you had that.
0:21:24 Hey, I actually kind of agree with the intentions
0:21:26 that people have on most of this,
0:21:30 just not with the ideas on how to achieve those intentions
0:21:35 because that’s where politics gets really dangerous
0:21:38 in a company and I have a funny story about this.
0:21:41 So my great uncle Harold, my grandmother didn’t speak
0:21:43 for 20 years, brother and sister.
0:21:45 And I asked my father a few years ago,
0:21:46 I said, like, what happened there?
0:21:47 Why didn’t they speak?
0:21:50 He’s like, oh, well, ’cause your great uncle Harold
0:21:51 was a Trotskyist.
0:21:53 And I was like, well, what was grandma?
0:21:55 She was a Stalinist.
0:21:58 They’re two slightly different kinds of communists
0:21:59 and they didn’t speak for 20 years.
0:22:01 And that’s kind of what happens inside a company.
0:22:04 It’s not like you’re a good person, you’re a bad person.
0:22:06 This is a moral issue.
0:22:10 It’s like, well, yeah, we all want way fewer people
0:22:12 to get killed in this current conflict,
0:22:14 but how you achieve that,
0:22:16 that gets very complicated very fast.
0:22:19 And so if you let people attack each other,
0:22:22 shut down conversations, intimidate each other
0:22:26 inside your company around that, it’s just all bad.
0:22:27 – Yeah, but fighting gets the fiercest
0:22:29 when mistakes are the lowest.
0:22:31 You know, like when people almost agree,
0:22:34 when they become mortal enemies.
0:22:35 – Right.
0:22:36 – It’s a very funny effect.
0:22:38 I mean, you can.
0:22:39 – Right, in a way, the closer they are,
0:22:42 the more, yeah, yeah, yeah, exactly.
0:22:44 It’s not the way you would think.
0:22:45 I think at the end of the day,
0:22:48 basically what you have to avoid is divisiveness.
0:22:52 I see we have a no asshole rule in a company,
0:22:56 like most companies, I think, many companies choose.
0:22:58 I think it’s actually a valid opinion
0:23:00 to choose the opposite and say,
0:23:03 we are perfectly fine with brilliant jerks.
0:23:05 Like that’s actually a stable equilibrium
0:23:06 you can choose as well,
0:23:09 but like you have to be all in on that.
0:23:12 And you know, I think it’s actually useful
0:23:16 to reframe like active divisiveness
0:23:19 as like hovering a part of the same company
0:23:21 as a act of Nassau.
0:23:22 And I think that just,
0:23:25 that actually kind of transcends a lot of these conversations
0:23:29 because it gets away from the issue.
0:23:30 It’s never about the issue.
0:23:33 It’s actually about the behavior by people around the issue.
0:23:34 That is a problem in a company, right?
0:23:39 So just say, if you manage to make it a norm,
0:23:41 to say, this is just simply not what happens
0:23:46 on company funded internet servers such as Jack,
0:23:47 everything gets simpler.
0:23:51 – Take that outside, outside the virtual.
0:23:53 – Maybe one last point.
0:23:55 One book that really helped me figuring it out
0:23:59 was Thomas Sowell’s “Conflict of Visions,”
0:24:00 which is an underappreciated book.
0:24:02 I think it is very, very good.
0:24:04 Even if you read the first three chapters,
0:24:07 suddenly you realize why everyone’s actually kind of right,
0:24:12 except comes from a different sort of philosophical prior.
0:24:14 – Yeah, yeah, no question.
0:24:17 Actually, so let’s, you know,
0:24:22 you were into crypto very, very early on and why,
0:24:27 and then, you know, how do you think, you know,
0:24:29 given kind of how things have unfolded,
0:24:31 how the technology has developed,
0:24:34 how the kind of regulatory uncertainty has played,
0:24:36 like how are you thinking about it now,
0:24:39 kind of in the context of Shopify in the world?
0:24:41 – Yeah, a couple of angles to that.
0:24:45 First of all, I make a habit,
0:24:49 so growing up in a very small town in Germany,
0:24:53 not being, like I didn’t know anyone
0:24:54 who was into computers where I was,
0:24:59 like until late in my teens and I moved away, basically.
0:25:03 I think I built some skills around like outside observation,
0:25:06 like just sort of monitoring,
0:25:08 like I tried to figure out everything that’s going on
0:25:10 and, you know, like Silicon Valley,
0:25:12 then I figured out what that was,
0:25:16 but like I downloaded like the Linux kernel source code
0:25:19 and mailing this back in most of their use net,
0:25:21 like so I could study it all.
0:25:24 So I’m like, it’s weird because I’m clearly an insider now,
0:25:27 but like I’ve sort of lived and built skills
0:25:29 around outside observation.
0:25:31 So I make it a habit to like find the areas
0:25:36 where there’s the most passion and most higher stakes
0:25:39 and the most talent and people are just having a brainstorm
0:25:40 about what the future’s like.
0:25:44 I find that this is, to me, like watching television.
0:25:45 It’s so great.
0:25:50 Like so crypto is such a wellspring of brilliant insight
0:25:55 into like just like, you know, amazing technologies
0:25:57 and so on because of that I studied.
0:26:02 I find like consumer internet, like mobile apps in China
0:26:06 around 2010, 2015 so times like this, these pockets.
0:26:08 So that was sort of my angle.
0:26:11 I did make a Ruby implementation in 2012
0:26:14 of a blockchain paper of Satoshi’s paper.
0:26:15 So like it’s been on my mind
0:26:18 and just because I also love decentralization of,
0:26:20 I mean, just like giving people power,
0:26:23 but entrepreneurship, Shopify is just like give people.
0:26:25 – It’s actually the true power
0:26:27 of the people is decentralization.
0:26:27 – Exactly.
0:26:30 – And the false one is communism by the way.
0:26:33 – So there’s a hundreds of ways have that drawn me
0:26:36 into this which are my personal ones.
0:26:39 What I was hoping for is like something more practical
0:26:42 because my favorite thing is if people that I,
0:26:45 things that I find super interesting end up like coming
0:26:47 sort of in the vicinity of commerce
0:26:49 at which point I get to play with,
0:26:53 like I can do it for a birthday rather than in the evening.
0:26:58 So, so, you know, I love from technology perspective,
0:27:02 I love it from a sort of philosophical perspective.
0:27:06 You know, obviously at some point people like want
0:27:11 to use it for clearing, like buying products and so on.
0:27:13 We’ve supported this in any way we call it
0:27:15 for as long as you could have.
0:27:19 The demand has been low for using it for physical products
0:27:22 because you know, there’s (indistinct)
0:27:24 perspective, there’s some advantage credit cards,
0:27:27 but like it is crazy to me that at this point
0:27:29 that the internet just doesn’t have native currency.
0:27:30 We will get there.
0:27:33 I think people in this room hopefully will play
0:27:36 a leading role in this.
0:27:41 I, you know, again, from a personal philosophical
0:27:46 perspective, you know, I was on the internet in the 90s
0:27:51 and of course used Netscape and it came along mosaic
0:27:54 actually before that has just this incredible experience
0:27:57 of like in the future, like everyone will be able
0:28:02 to contribute and this is going to distribute power
0:28:05 in a way that is like unanticipated and not understood.
0:28:09 Shopify itself is a little bit like stuck
0:28:11 in this sort of 90s view of the internet.
0:28:14 Like it’s, I know it’s hosted software.
0:28:17 So it doesn’t conform directly to everyone has
0:28:19 their own server, but like you have your own website.
0:28:20 It’s like something you can own.
0:28:22 Like you can own a domain.
0:28:26 You can own, you know, the design of your system.
0:28:28 You can retrain the pens that is important.
0:28:32 And then I think this is like people have like my one
0:28:36 criticism to the previous sort of wave of crypto has been
0:28:40 that this idea of decentralization is too technical.
0:28:43 Like it’s like, I know people in the community here,
0:28:46 but like no one else cares about that definition of like
0:28:49 what we need to figure out like crypto needs friends
0:28:53 and the best angle to find friends is to find the fellow
0:28:58 travelers who are trying to move power to individuals again.
0:29:02 So that people can just like are not relying on like a
0:29:05 central marketplace for instance.
0:29:08 So somewhere that just has its own gatekeeper rules
0:29:10 or where the fees move to exactly what your margin is
0:29:14 and eradicate your ability to actually have a sustainable
0:29:15 business.
0:29:17 And so that’s really, really attractive to that.
0:29:20 – Yeah, so you, it almost needs a campaign,
0:29:22 like kick the ass of the tech overlords.
0:29:25 Like we’re going to give the power,
0:29:27 we’re going to give the technology back to the people
0:29:28 we’re going to take it from.
0:29:29 – Well, the, yeah.
0:29:32 – The Internet and Meta and Amazon and et cetera.
0:29:34 – Yeah, yeah.
0:29:37 I think what started with everyone start your own web
0:29:40 server, everyone write your own HTML.
0:29:44 It’s actually a life and well if you just change the framing
0:29:46 away from the exact technical implementation.
0:29:47 – Right.
0:29:50 – And so that’s a better, I think.
0:29:51 – Yeah.
0:29:51 – A larger group.
0:29:54 You can build a, you can build a movement around that.
0:29:56 You can’t build a movement around it.
0:30:01 It must be using the yakme.js, right?
0:30:05 Like it’s like, that’s not how you start a movement.
0:30:06 – Excellent point.
0:30:11 One last question then we’ll go to audience questions.
0:30:16 So you’ve been doing a lot of kind of experimentation
0:30:19 now with AI and then interestingly, you know,
0:30:22 kind of asked with crypto, there’s this huge push
0:30:23 to regulate AI.
0:30:28 You know, what do you think that means for kind of the
0:30:33 industry and the world and it’s a good idea, a bad idea.
0:30:37 – I think regulating AI at this point is a bad idea.
0:30:42 I think regulation around crypto, I might not be quite so.
0:30:46 Like, I think a regulatory clarity is what we should want.
0:30:49 I think you can get that by having no regulation
0:30:51 because it’s just a regulated field, but it should be
0:30:55 extremely permissive and for experimentation.
0:30:58 I think we found out some of the things that the financial
0:31:03 system had are things that are deeply desirable in crypto
0:31:04 as well, one by another.
0:31:06 I’m sure there’s like, everyone can take a different
0:31:09 position on this, but like that’s there.
0:31:14 In an AI case, man, like I just like, you gotta,
0:31:18 like you can regulate institutions or like ground rules
0:31:21 for markets, but regulating technology is like,
0:31:24 that seems like a crazy position to take.
0:31:27 Like suddenly floating points are like illegal.
0:31:29 That’s after you do too many of them.
0:31:34 – It is so weird, you know, I had a conversation
0:31:37 with a policymaker and I was like, so if this was nuclear,
0:31:40 you want regulate nukes, you’d regulate physics.
0:31:43 That’s your idea like physics is not illegal.
0:31:48 – Yeah, throw away those textbooks immediately, please.
0:31:53 – So again, I think maybe even some of the things
0:31:56 that people have identified as potential problems
0:31:59 in the future, I guess we could agree with,
0:32:04 but like again, the proposed solutions to this are crazy
0:32:06 because we get into these insane places where like,
0:32:09 cool, maybe analog computing is actually better
0:32:10 for training neural nets.
0:32:13 So that is not floating point operations anymore.
0:32:15 I was like, are you doing that now
0:32:18 because of regulations instead of because of pragmatism?
0:32:20 Or, you know, it just, I don’t know.
0:32:22 It just seems too early.
0:32:27 And it, I think that it’s a very, very bad idea right now
0:32:30 for any governments to set ceilings and rules
0:32:34 that are especially onerous for startups,
0:32:35 because I think all you’re gonna get out of that
0:32:39 is just like, you’re gonna get the place
0:32:42 where it’s gonna be all behind four-pay APIs
0:32:44 that are gonna be all controlled by the few.
0:32:49 I mean, that’s strictly better than it doesn’t exist, granted.
0:32:52 But like, clearly we need open source.
0:32:54 Look at everything that happened just now
0:32:58 after AI became available first
0:33:02 and then thanks to Metta and releasing amazing Lama models.
0:33:04 We know so much more about what Nuance do.
0:33:08 We know so much more about how to accelerate this.
0:33:11 You need to rescue these things sometimes
0:33:12 from the ivory towers
0:33:15 and you have to give them to the tinkerers.
0:33:18 And the people who remember how to memory align things
0:33:22 and what, you know, how to get the most out of a hardware.
0:33:23 Like that’s a different crowd than the people
0:33:27 who were writing Python neural topologies.
0:33:30 And it’s, we need everyone on this.
0:33:32 And then we’re gonna get the best outcome
0:33:34 because it is foundational for our future.
0:33:36 – Yeah. – Also, we need to figure out
0:33:38 how to make crypto closer aligned with AI
0:33:43 because it’s a natural way for how to pay for micropayments
0:33:45 and tokens and so on.
0:33:47 And this is really, really important now.
0:33:48 – Well, and solve so many.
0:33:50 I mean, it’s interesting.
0:33:52 Crypto solves so many AI problems
0:33:54 and AI solves so many crypto problems.
0:33:55 – Yeah. – They’re–
0:33:57 – These should be symbiotic
0:34:00 and just the sine waves of excitement of these two areas
0:34:02 just happen to be dissonant right now
0:34:04 and we need to make them harmonious partly
0:34:07 because of the bad big companies.
0:34:08 (laughs)
0:34:09 Sorry.
0:34:10 Well, that’s great.
0:34:14 So we are ready to take questions.
0:34:16 Any questions you may have?
0:34:18 Yes, sir, back there.
0:34:22 – If crypto plays out in exactly the same–
0:34:24 (mumbles)
0:34:30 – It’s like, I think very naturally the browser
0:34:32 will just like be able to move value around
0:34:34 and you will want to move value
0:34:36 in a way that can be addressed by the browser.
0:34:40 This is like, you can get pretty close to that
0:34:42 but you would have to squint
0:34:44 and you have to deal with like poor UX
0:34:46 and you have to deal with hard on ramps.
0:34:50 I think, I don’t think anything kind of magical happens
0:34:52 other than that people will have,
0:34:55 they need more utilitarian ability to like use money.
0:34:57 Again, I think if you run the counterfactual
0:34:59 and say open AI just for whatever reason
0:35:02 you couldn’t work for Stripe and just decided to,
0:35:05 I don’t know, like use some level two blockchain
0:35:10 and to pay for tokens as a initial MVP.
0:35:15 People would now move incredible amounts of money around
0:35:18 and build startups that might have these kind of ideas
0:35:19 that Sam talked about on stage,
0:35:24 about people sharing value of creating GPTs
0:35:27 and like all of the, you know, the blockchain,
0:35:30 like if you look at Open AI Dev Days,
0:35:31 especially the last 20 minutes
0:35:36 and you open at the same time VS Code to Ethereum contract,
0:35:39 you can basically write it out while he’s talking.
0:35:43 It’s like, so, we have this incredible infrastructure
0:35:45 but we can’t use it part because of fees,
0:35:46 part because of speeds,
0:35:49 because like you gotta solve the infrastructure problems
0:35:50 which I know we are doing.
0:35:52 Like this is people are building through this particular winter
0:35:53 in a way that is like,
0:35:57 seems utterly ideal from my perspective.
0:35:58 So that’s great.
0:36:03 There is, we need to bring killer use cases.
0:36:06 We need to find, like we need to have people
0:36:07 want to use this.
0:36:13 Unfortunately, there is like the focus on these kind of things,
0:36:16 especially once you need culture and social things,
0:36:18 is annoyingly not linear.
0:36:20 And that’s really trips everyone up.
0:36:21 It’s like, it’s nothing, nothing, nothing, nothing,
0:36:22 everything.
0:36:25 And because we have to get,
0:36:28 we have to like toil in obscurity
0:36:31 to build up the infrastructure
0:36:33 until it can support the kind of way
0:36:35 that people want to use it.
0:36:37 And then we need to catch lightning in a bottle.
0:36:39 Yes.
0:36:40 – Thanks, Sabi.
0:36:42 I had a question.
0:36:44 So Shopify works with a lot of merchants and brands
0:36:47 that might not be as tech savvy
0:36:49 or might not understand everything that’s going on
0:36:51 in like AI or crypto.
0:36:54 How does Shopify think about messaging to those merchants
0:36:56 and saying, hey, this is the future.
0:36:58 These are things that you should be using.
0:37:01 – Yeah, I think people actually hire us
0:37:04 for having them be ideal for whatever the internet is
0:37:07 and for whatever their buyers,
0:37:09 like they want to continue buying.
0:37:11 Like again, we start before mobile phones.
0:37:14 And so like just making sure that everyone’s store
0:37:16 is available to some mobile phones
0:37:18 back in those days, responsive, whatever,
0:37:22 just had it tremendously because their colleagues
0:37:24 who didn’t have that like just like started doing poor
0:37:29 and poor as money moved to the small form factors.
0:37:32 And they started out competing their peers in the industry.
0:37:35 So like as this is a super old and probably dumb example,
0:37:38 but like what I’m saying is like the demand has to come up.
0:37:42 Like the merchants offering like you can’t solve this problem
0:37:44 just simply by bringing the supply side up,
0:37:45 the demand side needs to go up.
0:37:50 The buyers should want to spend on the blockchain.
0:37:54 Now, that is the huge beauty.
0:37:58 You have 3% to work with here.
0:38:00 Like the credit card fees are high, right?
0:38:02 Like there’s a lot of interesting incentive systems
0:38:03 that you can do.
0:38:06 But what I found was the most amazing thing about sort of
0:38:08 when I came back to crypto,
0:38:10 paid more attention, smart contracts, if you’re,
0:38:14 is that, and I actually think even the practitioners
0:38:19 in the crypto space are under appreciating this
0:38:22 as a talent is there’s never been a field
0:38:25 of such applied widely distributed
0:38:28 and high quality thought applied game theory, right?
0:38:31 Like it’s the amount of the constructions
0:38:34 and the incentive systems that are being built
0:38:36 in this community are like tremendous.
0:38:41 Like the artists that have done really, really well
0:38:45 through NFTs like a such a good example.
0:38:48 This is the first, and now this is like,
0:38:52 man, this was the first business model of the arts
0:38:55 that we’ve gotten since patronage times, right?
0:38:57 Like there’s a huge breakthrough
0:38:59 and we’ll make a comeback.
0:39:02 This thought, this quality of thought has to be applied
0:39:06 I think to all sorts of other areas and in different ways.
0:39:09 And we can use the infrastructure that’s been built
0:39:12 to do something way better than cash back credit cards
0:39:15 like the first extra 3%.
0:39:17 Now it needs to also be meet the world where it is.
0:39:21 It requires some kind of escrow systems for charge back.
0:39:22 There is real distribution.
0:39:24 There needs to be like it,
0:39:26 not everything has to be trustless.
0:39:30 We should bring trust progress into the systems.
0:39:33 Like it’s called a wallet, which I think is a good name.
0:39:36 But in my wallet is, I don’t think I’ve cashed my wallet.
0:39:40 It’s all cards, it’s like licenses.
0:39:41 It’s like driver license and so on, right?
0:39:45 Like it’s attestations that I’m, who I say I am.
0:39:49 And bringing in trust brokers into the system
0:39:51 through the primitives we have now and saying,
0:39:54 yeah, that person is who we say they are.
0:39:56 Yes, that person can be shipped to.
0:40:01 Like if you write this thing on a package destination,
0:40:05 FedEx knows how to unclog this into an actual destination address.
0:40:08 We have to build this with the industry.
0:40:11 And then at some point,
0:40:15 we’re going to just have a much better environment for commerce.
0:40:18 And much like it’s not slightly better than using credit.
0:40:20 That it can actually be 10 times better.
0:40:22 People move around more since COVID, right?
0:40:25 Like how often have you gotten a package
0:40:27 to some other address that you’re no longer there?
0:40:30 Like this, it would be wonderful to make addresses pointers
0:40:32 that can be updated in the background and so on, so on, so on.
0:40:38 So like work with the actual infrastructure of the world
0:40:41 and modernize it on these new systems.
0:40:45 And so that all that want to accrue to people who want to take up one,
0:40:47 like who actually live in a real world
0:40:52 and not just like the hindsight synonymous, you know, abstraction.
0:40:55 I think that’s absolutely doable.
0:40:58 But man, do we have to get fees down?
0:41:00 And this is like it’s never going to happen
0:41:03 if we go back to a $50 transaction.
0:41:04 Yes.
0:41:07 We’ve got time for one more question.
0:41:10 Yes, right in the front row, since I saw you first.
0:41:13 So Shopify has a really successful partner program
0:41:16 where engineers are building themes and plugins and apps.
0:41:19 Could you talk a bit more about some of the challenges
0:41:20 starting that system up?
0:41:23 And then also, as you see more and more people speak the language
0:41:24 of NFTs and the EVM,
0:41:27 how that can evolve as we get more interoperability.
0:41:30 This is definitely, like, you’re sort of hinting at
0:41:34 how cool would it have been to have these pieces of infrastructure primitives
0:41:37 and we could have governed the entire partner system on a token economy.
0:41:38 Like, that’s totally true.
0:41:43 I think this is, I really hope someone is going to explore
0:41:46 how to build systems like the Shopify ecosystem
0:41:53 because there’s a lot of sharing and a lot of multi-volume transactions
0:41:56 which we are all doing with ledgers in the background.
0:41:59 Putting it up, like putting up any two-sided marketplace is tricky.
0:42:02 Here you have to overcome the chicken-necked problems.
0:42:05 It’s tricky, but if you manage to do it,
0:42:08 this is also really, really self-sustaining afterwards.
0:42:10 We have one of the greatest things.
0:42:14 We have about five different customers
0:42:19 with merchants who start on Shopify who now IPO’ed and are public companies.
0:42:23 We now have one who is building an app,
0:42:26 like on the Shopify app with their own private company,
0:42:32 Public Company Clavio is now in their Bay Bay successor, which is awesome.
0:42:36 Again, this is a wonderful thing of what you can do.
0:42:40 Companies are estimated to produce about six times more value
0:42:47 than they capture to the world, like enlightened self-interest novice.
0:42:53 The constrained vision in so-called terms is real.
0:42:57 And there are so many ways to build,
0:43:01 like so many businesses lend themselves to being turned into platforms later.
0:43:03 The problem is everyone tries to build a platform first,
0:43:05 and that’s really, really, really hard.
0:43:10 It’s much better to extract a platform out of a working piece of software
0:43:13 that people are already using, and so that helped us.
0:43:17 People were already trading themes as zip files,
0:43:19 and we were like, “Okay, well, we can help with that,”
0:43:23 and people are already building things for themselves
0:43:26 on the APIs for integrating their own ERP systems,
0:43:31 and then building this to the point where you could list this ERP connector
0:43:37 and build it like our SDKs made it so that it can be in so many stores.
0:43:41 But, de-forward, we hinted people and helped them into the right direction,
0:43:45 and all these kind of things ended up being incredibly valuable for the community.
0:43:50 I mean, it’s definitely one of the coolest aspects to build,
0:43:53 but again, at the end of the day, what all of this is,
0:43:59 is you build really good software that solves real problems very, very quickly,
0:44:03 and then you build a bit of a game-theoretical system on top,
0:44:12 in which people being selfish about wanting to build something accrue a huge amount of value
0:44:19 to the community, to Shopify, to the merchants, and so on.
0:44:26 I mean, I feel like the most sophisticated thought along those lines on Shopify
0:44:30 is a fairly amateur sort of baby’s first incentive system
0:44:37 compared to some of the things that were built on Ethereum and the blockchains
0:44:42 in terms of sophistication, except everything in the world of crypto
0:44:45 was so pointed at financialization and finances,
0:44:49 but I think people might have missed what could be done
0:44:52 if you think more about products rather than financial products,
0:44:55 and I think that’s hopefully going to be the next wave of crypto,
0:45:00 it’s going to be more products and solving real problems for people.
0:45:01 I’m that amazing insight.
0:45:04 I would love to please join me in thanking Koby.
0:45:05 Thank you.
0:45:08 [Music]
0:45:10 [Music]
0:45:12 [Music]
0:45:14 [Music]
0:45:16 [Music]
0:45:18 [Music]
0:45:20 [Music]
0:45:22 [Music]
0:45:24 [Music]
0:45:26 you
0:00:04 When Shopify was founded in 2006,
0:00:07 it was just a year prior to the launch of iPhone.
0:00:10 Needless to say, smartphones were a far cry
0:00:11 from the ubiquitous accessory
0:00:13 that now billions carry around.
0:00:17 And fast forward to today, nearly 20 years later,
0:00:19 where Shopify merchants can accept cryptocurrency,
0:00:21 create immersive shopping experiences
0:00:22 with augmented reality,
0:00:26 and even use AI to generate product descriptions.
0:00:29 So how does a nearly $100 billion company
0:00:30 stay ahead of these trends?
0:00:33 And what will shopping look like in the generations to come?
0:00:36 Plus, why would A16Z never have invested in Shopify
0:00:38 in its early day?
0:00:40 And why has the CEO of this company
0:00:42 chosen to continue writing code?
0:00:44 These are just some of the questions covered
0:00:45 in today’s episode,
0:00:48 the conversation the Shopify co-founder and CEO
0:00:52 to bias Luki and A16Z co-founder, Ben Horowitz.
0:00:54 This episode was also originally published
0:00:57 on our sister podcast, Web3 with A16Z.
0:01:00 So if you are excited about the next generation
0:01:03 of the internet, find Web3 with A16Z
0:01:04 wherever you get your podcasts.
0:01:06 All right, let’s get started.
0:01:14 – Welcome to Web3 with A16Z,
0:01:17 a show about building the next generation of the internet
0:01:20 from the team at A16Z Crypto.
0:01:22 Today’s episode features Tobias Lutke,
0:01:26 CEO and co-founder of the e-commerce platform Shopify.
0:01:29 Speaking with A16Z co-founder, Ben Horowitz,
0:01:32 at our second annual founder summit in November,
0:01:35 they discuss what it takes to build a breakout startup
0:01:39 in a crowded category, the changing face of retail,
0:01:41 how to affect change in the workplace,
0:01:44 and how to handle individual emotions and corporate culture,
0:01:47 including dealing with calls for activism,
0:01:50 as well as the value of embracing negativity.
0:01:52 They also touch on the moral imperative
0:01:54 behind creating quality software,
0:01:58 the symbiosis between AI and crypto, and more.
0:02:01 As a reminder, none of the following should be taken
0:02:04 as business, legal, tax, or investment advice.
0:02:07 Please see A16Z.com/disclosures
0:02:08 for more important information,
0:02:11 including a link to a list of our investments.
0:02:19 – All right, so for those of you who don’t know Toby,
0:02:20 which you probably should know Toby,
0:02:22 you can look him up as well,
0:02:27 but he started a snowboarding company
0:02:29 in 2004 called Snowdevil,
0:02:33 and he parlayed that into Shopify,
0:02:37 which now serves millions of merchants
0:02:40 in 175 countries around the world.
0:02:42 How many countries are there, like in total?
0:02:44 – I think that’s most of them, there’s 200?
0:02:45 – Yeah, 200, yeah, but–
0:02:48 – There’s changes around the margins.
0:02:51 – You’ll get them all at some point.
0:02:54 And then he was also interesting for this crowd.
0:02:58 He was one of the first adopters of Coinbase Commerce.
0:03:02 So Toby’s been kind of big on crypto for a while now.
0:03:04 But let’s get started.
0:03:08 So you started as a snowboarding company,
0:03:12 and I always find that funny because of all the CEOs
0:03:16 that I know you’re probably the least snowboard kickback,
0:03:19 like let’s hit the powder dude.
0:03:21 So how did that happen?
0:03:23 Like why did you start out that way?
0:03:28 – Yeah, you know, when you talk to governments,
0:03:32 they always, Shopify is doing a lot about entrepreneurship,
0:03:35 which I was a lot of entrepreneurship,
0:03:38 and it’s such a brilliant institution.
0:03:39 I know everyone here is a card-carrying member
0:03:42 of the entrepreneurship is good club,
0:03:45 but politicians usually, they’re sort of vague on it,
0:03:50 but they say like they want to pass government programs
0:03:52 and policies to cause more entrepreneurship,
0:03:55 which I always think is the wrong way around.
0:04:00 But there’s actually, so the funny story is that,
0:04:03 so I was in Canada at this point, I came from Germany,
0:04:06 living in Canada 2004, three, four,
0:04:10 and I found out I cannot actually get a job.
0:04:13 I was trying to be hired as a computer programmer locally,
0:04:15 but I found out I did not have a work permit.
0:04:16 I didn’t know what a work permit was.
0:04:20 So turns out that you are in Canada,
0:04:22 allowed to start a company even without a work permit.
0:04:23 So there’s one government program
0:04:27 that actually has led to what company is being created.
0:04:29 – Well, that’s like the best government program ever.
0:04:32 Just like the payoff on Shopify paid for itself
0:04:33 a million times.
0:04:35 – And it turns out Canada is very cold.
0:04:37 So I was doing snowboarding, so I knew something about it,
0:04:40 and I started an online store for snowboarding.
0:04:42 It’s kind of a pretty simple story like this.
0:04:44 – That’s amazing.
0:04:47 So basically the key to your whole career
0:04:49 was not being able to work.
0:04:49 – Yes.
0:04:51 (laughing)
0:04:53 In fact, you know, what was in my mind too is like,
0:04:56 I was like the solution with programming at this time.
0:04:57 It was like sort of a Java word,
0:04:59 kind of was kind of very oppressive.
0:05:04 And I wanted to reclaim it as a hobby.
0:05:06 So I was like, cool, if I sell snowboards online,
0:05:08 I can make money while snowboarding,
0:05:09 and I get to do more fun stuff.
0:05:14 And it’s been extremely unsuccessful as a strategy.
0:05:17 I basically never had time to snowboard again
0:05:17 after I started.
0:05:20 (laughing)
0:05:22 – The end of your snowboarding career
0:05:24 was the start of your snowboarding business.
0:05:25 That’s pretty remarkable.
0:05:27 – Maybe I would project more back at this point
0:05:29 if things would have gone in the counterfactual way.
0:05:31 – Yeah, now the other thing that really strikes me
0:05:34 about Shopify is that, you know,
0:05:36 and I hate to say this as a venture capitalist
0:05:38 who some people think is a smart person,
0:05:42 but we probably would have never invested in Shopify
0:05:43 when you started it
0:05:47 because it was probably the most over-competed
0:05:48 category in the world.
0:05:52 Like everybody had some kind of e-commerce platform.
0:05:56 I mean, there was just so, so many of them,
0:06:00 Magento, this, that, the other.
0:06:05 So did you go, oh, no, no, no, like,
0:06:07 I’m gonna build this thing.
0:06:09 It’s gonna dominate everything.
0:06:10 Then I’m gonna have a network effect
0:06:12 and I’m gonna build, you know,
0:06:14 like one of the biggest things that anybody’s ever seen.
0:06:19 Like, how did you like get all the way to where you are?
0:06:20 Like, what was that path?
0:06:24 – Yeah, no, there’s no plan like this.
0:06:29 Just like, I think, for every entrepreneur,
0:06:31 it’s a useful thing to think about
0:06:33 what is your energy source, right?
0:06:35 Like what is actually driving you?
0:06:37 And sometimes it’s greed.
0:06:40 And we, you know, like the sort of negative emotions
0:06:42 channeled into building actually
0:06:44 the most powerful energy sources.
0:06:47 I had this, you were right.
0:06:48 Like, I mean, Netscape, IPO,
0:06:52 but like, you have to explain every platform
0:06:54 in terms of an old platform and, you know,
0:06:57 online hosted CS catalog with buy buttons
0:07:00 was like kind of a way you would mark
0:07:03 to describe even what the internet would be used for, right?
0:07:04 Like, so e-commerce was like right there
0:07:07 from the beginning, I thought there would be lots of software
0:07:09 that I could use and I didn’t find anything.
0:07:12 I was actually really upset with the quality of the software.
0:07:15 And so like, I channel that into building something
0:07:17 and that was my energy source.
0:07:19 And I just love the building.
0:07:20 – That’s actually, by the way,
0:07:23 one of the best entrepreneurial ideas there is,
0:07:25 which is this stuff sucks so bad.
0:07:27 Yeah, I’m so mad about having to use it
0:07:29 that I’m going to build my own.
0:07:32 – Totally, I honestly, it fuels me to do it this day.
0:07:34 I’m just, I still think it’s ridiculous
0:07:37 what people are subjected to in terms of quality of software.
0:07:39 It’s like, oh, it’s weird.
0:07:41 I probably did some like neurological,
0:07:43 like kind of rearrangement there,
0:07:44 but like to me, this is almost a moral issue
0:07:48 of like when people like confused with software
0:07:50 and they think about themselves as like being inadequate.
0:07:53 It’s like, how dare software make humans feel that way?
0:07:54 Right, like it’s just not okay.
0:07:57 So you want to make it a more approachable,
0:07:59 but like, I think the thing that people might have missed
0:08:02 in e-commerce or like, it’s not that I had a great insight.
0:08:05 I just sort of organically got there is bad.
0:08:06 Yes, there was a lot of e-commerce software,
0:08:09 but it was all built for already rich retailers
0:08:11 who needed to move online.
0:08:13 Then that was a phrase people used.
0:08:17 And there was no one started software
0:08:20 for starting more entrepreneurial process online.
0:08:23 Like actually the digital native, what we call it now also,
0:08:26 like if you like that term,
0:08:29 because there’s a totally different set of requirements
0:08:30 for the people who are just starting out.
0:08:32 And so that’s what we did.
0:08:33 So this is the thing, right?
0:08:38 Like it’s like, when you talk,
0:08:40 that’s like, I don’t know, 60 words per minute or something,
0:08:42 and you type fast, it’s like twice that.
0:08:45 But like, if you can do customer consultation
0:08:49 in your own brain, the brain runs at like terabytes per second
0:08:50 by bandwidth, right?
0:08:53 Like so, if you know the use case,
0:08:56 if you’re building for yourself or your past experience
0:08:58 or some kind of, if you have a really good model
0:09:01 of your customers or use cases in your mind,
0:09:03 you can do customer development at high bandwidth
0:09:06 with low latency and you can just build something
0:09:08 that you know needs to exist.
0:09:10 – Yeah, no, that’s incredible.
0:09:13 So one of the things, you’re also kind of been
0:09:19 a kind of a special kind of, or a different kind of CEO
0:09:24 than some of the ones that people read about
0:09:27 such as Adela or Bob Iger and that you have a reputation
0:09:31 for being extremely hands-on, including writing code.
0:09:34 You still have the fury that started the company.
0:09:38 How do you think about that in terms of,
0:09:40 you’re not kind of paying attention
0:09:42 to all parts of the company equally.
0:09:46 You’re clearly doing certain things and not others.
0:09:49 How do you think about that as an effective
0:09:51 kind of management style?
0:09:56 – Yeah, I mean, I imagine the people in my company
0:09:58 would disagree with your assessment
0:10:00 about some effective management style.
0:10:02 So well, it’s worked pretty well.
0:10:07 It’s working, but it certainly doesn’t proxy
0:10:11 to what people sort of, like it’s not a CEO job
0:10:14 out of central casting that I’m trying to perform, right?
0:10:17 I tried that on and it didn’t work for me at all.
0:10:20 I actually find, like I value being in all the details
0:10:23 and I ask it from all the people who are around me
0:10:25 and report to me, like almost everyone,
0:10:29 especially after COVID, I turned over everyone,
0:10:33 but one of my executives during the first 10 months of COVID.
0:10:35 And now almost everyone reports to me,
0:10:38 it’s actually a AX founder, maybe someone be acquired
0:10:41 or someone who started a company before coming to Shopify.
0:10:45 So because there’s this, I just want,
0:10:48 I find it actually really inefficient,
0:10:50 not understanding the details of what we’re doing, right?
0:10:51 Because then–
0:10:52 – Well, then you’re guessing.
0:10:53 – Then you’re guessing, right?
0:10:57 And you’re like, well, when something goes wrong,
0:10:59 now you not only need to fix it,
0:11:02 you actually have the first cram for,
0:11:04 like you have to learn like three months
0:11:05 before you can actually do a fix.
0:11:06 If you understand what’s going on
0:11:09 or what ought to be existing, it’s much faster.
0:11:13 And you want to train people on the company mission, right?
0:11:16 Like because there’s a thing that you want to exist,
0:11:18 like even if you don’t know exactly what it is,
0:11:21 like you know which direction you want people to go, right?
0:11:24 And getting everyone aligned to go in this direction
0:11:27 is only possible if you can paint picture
0:11:30 of like how does this area actually sum together
0:11:33 all the way to helping the mission overall.
0:11:38 And so, and this is sort of a sort of unique thing
0:11:39 about our times right now.
0:11:41 It’s like the infrastructure keeps changing,
0:11:44 like here’s crypto, here’s AI, here’s,
0:11:46 I mean, I started again almost,
0:11:48 like next year is 20 years ago, right?
0:11:51 So there was no mobile phone, like that offer quality
0:11:53 that we have now.
0:11:54 – Right, that’s right.
0:11:55 You started before the mobile phone.
0:11:58 – Right, so it’s like, that was one of the things
0:11:59 that just sort of happened.
0:12:03 And, you know, SaaS software was new at this time.
0:12:05 So, you know, the platform’s capabilities change
0:12:08 and you want to adopt them to like your mission
0:12:10 to what is now possible.
0:12:13 And so you have to be able to understand, well,
0:12:15 but in a kind of like ideally in your head,
0:12:21 in the world where, I take a very long-term view
0:12:24 and my biggest fear that I have
0:12:27 is that Shopify winds up being a fantastic solution
0:12:30 to a problem that people no longer have, right?
0:12:32 Or no longer solve this way.
0:12:33 So I’m trying to like need to keep it current
0:12:35 and being able to run the counterfactual about
0:12:39 this is not possible, this is what we’re trying to accomplish.
0:12:41 But if this would have been possible from the beginning,
0:12:43 we would have built this entirely different thing.
0:12:45 So now our job’s to get from here to there
0:12:48 because, again, we want to keep it current and idea.
0:12:49 – Right, you know, that’s such a great insight
0:12:53 because they think that, you know, when I work with CEOs
0:12:55 and you think about the job,
0:12:58 so much of the job is making high quality decisions
0:13:00 and setting the direction of the company.
0:13:03 And in some respects, neither feels like work.
0:13:07 You know, because you’re working on something else,
0:13:10 like how does AI work, what’s possible,
0:13:13 in order to make decisions and set the direction.
0:13:16 But what you’re doing seems actually removed
0:13:20 from anything that the company needs to do right now.
0:13:22 And I think so many people kind of neglect
0:13:24 that kind of thing for that reason.
0:13:27 The other thing that I always find is,
0:13:29 at any given point in time,
0:13:32 some things in the company are very important
0:13:33 and very high leverage.
0:13:35 And some things are just not important
0:13:37 for you to put your time into.
0:13:39 And people often get lost in that.
0:13:43 They think, oh, I’ve got to talk to this person
0:13:46 and then that person and then the other person and so forth.
0:13:48 And it’s like, no, no, you don’t.
0:13:49 You’ve got to make good decisions.
0:13:50 You have to set the direction.
0:13:52 And if you don’t do that, nobody’s doing that.
0:13:54 And then the company’s going to fail.
0:13:57 So that’s such an amazing way that you’ve gone about that.
0:13:59 – If you’re in a small team,
0:14:01 like for a new thing that your company needs to do
0:14:04 and you’re in a small team of people who are like,
0:14:08 know or like, figure out the shape of it early,
0:14:09 if it’s actually as important,
0:14:11 it will itself have a gravity
0:14:14 to change the way the company thinks about quality of software
0:14:16 shipping, the way the architect thinks.
0:14:22 And it’s also after this comes up as a new thing,
0:14:26 you then have the ability to like a type of legitimacy.
0:14:27 As a founder,
0:14:30 you have a huge amount of legitimacy in a company already.
0:14:33 But like if you also are there with the new project,
0:14:35 that’s important, that needs to be a resource and say,
0:14:38 this is what we’ll do.
0:14:40 You don’t have to have all these conversations.
0:14:43 People love clarity of like, this is what we do.
0:14:45 Like it has a gravity all to itself.
0:14:48 That’s unlike anything you can accomplish
0:14:50 by one-on-one consultation,
0:14:53 trying to talk everyone through the entire decision matrix.
0:14:56 And so what is Toby doing over there?
0:14:57 – Oh, that must be important.
0:14:59 – Yeah, that’s exactly right.
0:15:01 We’re doing like a job site click now.
0:15:04 And like, then I don’t have to figure out like,
0:15:06 yes, there’s like 15 teams that are already doing AI stuff
0:15:09 and they ought to be in some kind of constellation.
0:15:12 But now this AI assistant thing is going to run through
0:15:16 and the company, everyone wants to be a part of it.
0:15:18 And this is like, all these problems kind of disappear.
0:15:19 It’s like, it’s maybe not the most effective-
0:15:20 – Like a magic trick.
0:15:21 – Yeah, yeah, yeah.
0:15:23 Yeah, actually, I think by the way,
0:15:25 like everyone in this,
0:15:28 like everyone here has clearly read your books.
0:15:32 And if you haven’t done do that, like I think-
0:15:33 – Thank you.
0:15:35 – The best books on leadership, honestly.
0:15:38 Thank you so much for sharing all this.
0:15:40 I think in some of what you described,
0:15:43 but there’s only very few ways to do a change management.
0:15:45 One way of doing it is to something super, super,
0:15:49 super surprising and then explain everyone why you did it.
0:15:51 It’s the fastest way to get a large company.
0:15:53 In my shop, there’s about 10,000 people,
0:15:55 much, much, much faster way to do a change management
0:15:56 than hand-to-hand combat.
0:15:58 – It’s like shock therapy.
0:16:00 It’s like, what the fuck was that?
0:16:02 Oh, well, this was that, that was okay, got it.
0:16:06 Very, very, very high retention on that.
0:16:07 So let’s talk about, you know,
0:16:10 one of the things that has come up, you know,
0:16:13 a lot over the years that you had probably, you know,
0:16:16 maybe the best statement on, I thought,
0:16:20 is politics, you know, politics, activism,
0:16:22 how that’s kind of come into companies
0:16:25 has taken over a lot of the activity of companies.
0:16:29 And how do you, I guess, how do you think about that?
0:16:33 And then why’d you say what you said?
0:16:35 And then how’s that going?
0:16:39 And then how are you able to sustain it particularly
0:16:42 as things get even hotter now with, you know,
0:16:44 what’s going on in the Middle East and so forth?
0:16:47 You know, it’s gotten even more intense than it was,
0:16:49 you know, amazingly in 2020.
0:16:51 – That’s a complex topic.
0:16:56 I can, here’s, I mean, there’s a lot of conflicting ideas
0:16:58 about what people want companies to be.
0:17:02 And, you know, I think, especially in 2010
0:17:06 and maybe now again, a lot of people would like companies
0:17:08 to play the role of some kind of phantom limb
0:17:13 of what I think the whole left over by the stories
0:17:16 that people are surrounded with, like maybe with governments
0:17:17 or like, like, and so on.
0:17:20 I suddenly, people want, like I have gotten tons
0:17:23 and tons of petitions for Shopify to take stances on issues.
0:17:27 And I think everyone’s feeling the same thing.
0:17:30 This was weird to me, but like initially,
0:17:32 but like, I found myself generally agreeing
0:17:34 with what people identify as the problems
0:17:36 when this happens.
0:17:38 But I found myself, they really, if ever,
0:17:41 really agreeing with the proposed solutions
0:17:41 to those problems.
0:17:44 And I think it’s kind of important to say like,
0:17:46 hey, unless we are all aligned on this,
0:17:48 we’re probably not doing this.
0:17:51 Because, you know, at the end of the day,
0:17:53 like companies ought to be thought of
0:17:54 as a bit of a simpler thing.
0:17:57 It’s, you know, real love entrepreneurship.
0:17:59 Entrepreneurship is, you know,
0:18:01 liberal values kind of institution of, you know,
0:18:05 you’re like reaching for independence,
0:18:07 reaching for like, to better yourself,
0:18:09 like building something like enlightened,
0:18:10 the ideas of enlightened self-interest.
0:18:13 There is a bit of a political coding in this,
0:18:14 people at least think so.
0:18:16 And someone else might take a totally different opinion
0:18:17 about all these kinds of things.
0:18:18 So that’s cool.
0:18:20 The cool thing is companies themselves explorations
0:18:22 into a set of ideas.
0:18:24 Sometimes encoded in a mission sometimes by the founders.
0:18:27 And if we make every company the same,
0:18:29 then we lose that.
0:18:31 That’s actually like a weird form of diversity itself,
0:18:32 if you think about it.
0:18:36 So anyway, I went ahead and just like said,
0:18:40 like look, we cause more entrepreneurship to exist.
0:18:43 And sometimes there are social causes
0:18:45 that are aligned with this.
0:18:48 And then maybe we become active because of our mission.
0:18:50 But outside of that, you know,
0:18:53 everyone makes money and you can do whatever you want
0:18:54 in your after hours.
0:18:58 And that’s just a simpler way of thinking about the business.
0:19:02 Of course, that’s not universally loved,
0:19:05 but like after being there, they clear about it
0:19:05 and explaining it.
0:19:09 They explain, you know, we give everyone when we hire people
0:19:11 a letter of like, here are the reasons,
0:19:13 like before you sign.
0:19:15 Here’s the reason why you might not want to work
0:19:18 for Shopify because like here’s a set of ideas
0:19:19 that you disagree with.
0:19:22 And then that clarity is wonderful.
0:19:24 We keep everything so nebulous.
0:19:28 I think the world of marketing and PR
0:19:31 has just done a number on us all to be so non-committal
0:19:32 to everything.
0:19:33 – Yeah.
0:19:35 – And the problem is if you’re non-committal,
0:19:37 you leave a lot of vacuum.
0:19:38 – Right.
0:19:40 – And that vacuum is going to be filled by someone.
0:19:43 And it’s usually the people who want to fill it
0:19:46 with bad faith takes, who fill it.
0:19:50 So if you just clear about it, it works sometimes,
0:19:54 like this is something I might get into hot water with,
0:19:57 but like sometimes it’s actually good to heighten
0:19:58 the misalignment for the kind of people
0:20:00 that you don’t think should be working in a company.
0:20:04 Like, hey, we work hard and it’s kind of an-
0:20:06 – Yeah, we have poor work-life balance.
0:20:07 – Yeah, yeah.
0:20:08 – Sorry.
0:20:09 – That’s a good-
0:20:11 – You are free to work elsewhere.
0:20:15 – If it’s true, put it up, like light it up on a sign
0:20:17 because otherwise you’re going to end up
0:20:20 with the weirdest divisiveness in the company
0:20:22 because of basically false advertising.
0:20:25 So I think it’s better to just be clear.
0:20:29 So I mean, what is it like?
0:20:31 There’s a lot goes into this.
0:20:33 Like again, culture is unstable.
0:20:36 We know from the internet that every community
0:20:40 that anyone likes Reddit, like Hacker News,
0:20:42 if anyone actually likes this page,
0:20:47 it’s a product of unbelievably dedicated moderators
0:20:51 who are keeping this discourse well.
0:20:54 I think that’s an idea that I think companies
0:20:55 have started adopting too.
0:20:57 Like once you’re a couple of thousand people,
0:20:59 your Slack channel is the internet basically
0:21:02 and you need to apply similar rules.
0:21:04 You need to be able to take people to a sign saying,
0:21:05 hey, here’s the thing you’re causing.
0:21:07 When you’re sending a message to a channel
0:21:09 with 3,000 people in it,
0:21:11 that is basically the same as you spending a day
0:21:15 typing individual text message to any person in the channel.
0:21:16 And would you do that?
0:21:18 If no, probably don’t put in this kind of stuff.
0:21:22 Yeah, that’s really such a great insight that you had that.
0:21:24 Hey, I actually kind of agree with the intentions
0:21:26 that people have on most of this,
0:21:30 just not with the ideas on how to achieve those intentions
0:21:35 because that’s where politics gets really dangerous
0:21:38 in a company and I have a funny story about this.
0:21:41 So my great uncle Harold, my grandmother didn’t speak
0:21:43 for 20 years, brother and sister.
0:21:45 And I asked my father a few years ago,
0:21:46 I said, like, what happened there?
0:21:47 Why didn’t they speak?
0:21:50 He’s like, oh, well, ’cause your great uncle Harold
0:21:51 was a Trotskyist.
0:21:53 And I was like, well, what was grandma?
0:21:55 She was a Stalinist.
0:21:58 They’re two slightly different kinds of communists
0:21:59 and they didn’t speak for 20 years.
0:22:01 And that’s kind of what happens inside a company.
0:22:04 It’s not like you’re a good person, you’re a bad person.
0:22:06 This is a moral issue.
0:22:10 It’s like, well, yeah, we all want way fewer people
0:22:12 to get killed in this current conflict,
0:22:14 but how you achieve that,
0:22:16 that gets very complicated very fast.
0:22:19 And so if you let people attack each other,
0:22:22 shut down conversations, intimidate each other
0:22:26 inside your company around that, it’s just all bad.
0:22:27 – Yeah, but fighting gets the fiercest
0:22:29 when mistakes are the lowest.
0:22:31 You know, like when people almost agree,
0:22:34 when they become mortal enemies.
0:22:35 – Right.
0:22:36 – It’s a very funny effect.
0:22:38 I mean, you can.
0:22:39 – Right, in a way, the closer they are,
0:22:42 the more, yeah, yeah, yeah, exactly.
0:22:44 It’s not the way you would think.
0:22:45 I think at the end of the day,
0:22:48 basically what you have to avoid is divisiveness.
0:22:52 I see we have a no asshole rule in a company,
0:22:56 like most companies, I think, many companies choose.
0:22:58 I think it’s actually a valid opinion
0:23:00 to choose the opposite and say,
0:23:03 we are perfectly fine with brilliant jerks.
0:23:05 Like that’s actually a stable equilibrium
0:23:06 you can choose as well,
0:23:09 but like you have to be all in on that.
0:23:12 And you know, I think it’s actually useful
0:23:16 to reframe like active divisiveness
0:23:19 as like hovering a part of the same company
0:23:21 as a act of Nassau.
0:23:22 And I think that just,
0:23:25 that actually kind of transcends a lot of these conversations
0:23:29 because it gets away from the issue.
0:23:30 It’s never about the issue.
0:23:33 It’s actually about the behavior by people around the issue.
0:23:34 That is a problem in a company, right?
0:23:39 So just say, if you manage to make it a norm,
0:23:41 to say, this is just simply not what happens
0:23:46 on company funded internet servers such as Jack,
0:23:47 everything gets simpler.
0:23:51 – Take that outside, outside the virtual.
0:23:53 – Maybe one last point.
0:23:55 One book that really helped me figuring it out
0:23:59 was Thomas Sowell’s “Conflict of Visions,”
0:24:00 which is an underappreciated book.
0:24:02 I think it is very, very good.
0:24:04 Even if you read the first three chapters,
0:24:07 suddenly you realize why everyone’s actually kind of right,
0:24:12 except comes from a different sort of philosophical prior.
0:24:14 – Yeah, yeah, no question.
0:24:17 Actually, so let’s, you know,
0:24:22 you were into crypto very, very early on and why,
0:24:27 and then, you know, how do you think, you know,
0:24:29 given kind of how things have unfolded,
0:24:31 how the technology has developed,
0:24:34 how the kind of regulatory uncertainty has played,
0:24:36 like how are you thinking about it now,
0:24:39 kind of in the context of Shopify in the world?
0:24:41 – Yeah, a couple of angles to that.
0:24:45 First of all, I make a habit,
0:24:49 so growing up in a very small town in Germany,
0:24:53 not being, like I didn’t know anyone
0:24:54 who was into computers where I was,
0:24:59 like until late in my teens and I moved away, basically.
0:25:03 I think I built some skills around like outside observation,
0:25:06 like just sort of monitoring,
0:25:08 like I tried to figure out everything that’s going on
0:25:10 and, you know, like Silicon Valley,
0:25:12 then I figured out what that was,
0:25:16 but like I downloaded like the Linux kernel source code
0:25:19 and mailing this back in most of their use net,
0:25:21 like so I could study it all.
0:25:24 So I’m like, it’s weird because I’m clearly an insider now,
0:25:27 but like I’ve sort of lived and built skills
0:25:29 around outside observation.
0:25:31 So I make it a habit to like find the areas
0:25:36 where there’s the most passion and most higher stakes
0:25:39 and the most talent and people are just having a brainstorm
0:25:40 about what the future’s like.
0:25:44 I find that this is, to me, like watching television.
0:25:45 It’s so great.
0:25:50 Like so crypto is such a wellspring of brilliant insight
0:25:55 into like just like, you know, amazing technologies
0:25:57 and so on because of that I studied.
0:26:02 I find like consumer internet, like mobile apps in China
0:26:06 around 2010, 2015 so times like this, these pockets.
0:26:08 So that was sort of my angle.
0:26:11 I did make a Ruby implementation in 2012
0:26:14 of a blockchain paper of Satoshi’s paper.
0:26:15 So like it’s been on my mind
0:26:18 and just because I also love decentralization of,
0:26:20 I mean, just like giving people power,
0:26:23 but entrepreneurship, Shopify is just like give people.
0:26:25 – It’s actually the true power
0:26:27 of the people is decentralization.
0:26:27 – Exactly.
0:26:30 – And the false one is communism by the way.
0:26:33 – So there’s a hundreds of ways have that drawn me
0:26:36 into this which are my personal ones.
0:26:39 What I was hoping for is like something more practical
0:26:42 because my favorite thing is if people that I,
0:26:45 things that I find super interesting end up like coming
0:26:47 sort of in the vicinity of commerce
0:26:49 at which point I get to play with,
0:26:53 like I can do it for a birthday rather than in the evening.
0:26:58 So, so, you know, I love from technology perspective,
0:27:02 I love it from a sort of philosophical perspective.
0:27:06 You know, obviously at some point people like want
0:27:11 to use it for clearing, like buying products and so on.
0:27:13 We’ve supported this in any way we call it
0:27:15 for as long as you could have.
0:27:19 The demand has been low for using it for physical products
0:27:22 because you know, there’s (indistinct)
0:27:24 perspective, there’s some advantage credit cards,
0:27:27 but like it is crazy to me that at this point
0:27:29 that the internet just doesn’t have native currency.
0:27:30 We will get there.
0:27:33 I think people in this room hopefully will play
0:27:36 a leading role in this.
0:27:41 I, you know, again, from a personal philosophical
0:27:46 perspective, you know, I was on the internet in the 90s
0:27:51 and of course used Netscape and it came along mosaic
0:27:54 actually before that has just this incredible experience
0:27:57 of like in the future, like everyone will be able
0:28:02 to contribute and this is going to distribute power
0:28:05 in a way that is like unanticipated and not understood.
0:28:09 Shopify itself is a little bit like stuck
0:28:11 in this sort of 90s view of the internet.
0:28:14 Like it’s, I know it’s hosted software.
0:28:17 So it doesn’t conform directly to everyone has
0:28:19 their own server, but like you have your own website.
0:28:20 It’s like something you can own.
0:28:22 Like you can own a domain.
0:28:26 You can own, you know, the design of your system.
0:28:28 You can retrain the pens that is important.
0:28:32 And then I think this is like people have like my one
0:28:36 criticism to the previous sort of wave of crypto has been
0:28:40 that this idea of decentralization is too technical.
0:28:43 Like it’s like, I know people in the community here,
0:28:46 but like no one else cares about that definition of like
0:28:49 what we need to figure out like crypto needs friends
0:28:53 and the best angle to find friends is to find the fellow
0:28:58 travelers who are trying to move power to individuals again.
0:29:02 So that people can just like are not relying on like a
0:29:05 central marketplace for instance.
0:29:08 So somewhere that just has its own gatekeeper rules
0:29:10 or where the fees move to exactly what your margin is
0:29:14 and eradicate your ability to actually have a sustainable
0:29:15 business.
0:29:17 And so that’s really, really attractive to that.
0:29:20 – Yeah, so you, it almost needs a campaign,
0:29:22 like kick the ass of the tech overlords.
0:29:25 Like we’re going to give the power,
0:29:27 we’re going to give the technology back to the people
0:29:28 we’re going to take it from.
0:29:29 – Well, the, yeah.
0:29:32 – The Internet and Meta and Amazon and et cetera.
0:29:34 – Yeah, yeah.
0:29:37 I think what started with everyone start your own web
0:29:40 server, everyone write your own HTML.
0:29:44 It’s actually a life and well if you just change the framing
0:29:46 away from the exact technical implementation.
0:29:47 – Right.
0:29:50 – And so that’s a better, I think.
0:29:51 – Yeah.
0:29:51 – A larger group.
0:29:54 You can build a, you can build a movement around that.
0:29:56 You can’t build a movement around it.
0:30:01 It must be using the yakme.js, right?
0:30:05 Like it’s like, that’s not how you start a movement.
0:30:06 – Excellent point.
0:30:11 One last question then we’ll go to audience questions.
0:30:16 So you’ve been doing a lot of kind of experimentation
0:30:19 now with AI and then interestingly, you know,
0:30:22 kind of asked with crypto, there’s this huge push
0:30:23 to regulate AI.
0:30:28 You know, what do you think that means for kind of the
0:30:33 industry and the world and it’s a good idea, a bad idea.
0:30:37 – I think regulating AI at this point is a bad idea.
0:30:42 I think regulation around crypto, I might not be quite so.
0:30:46 Like, I think a regulatory clarity is what we should want.
0:30:49 I think you can get that by having no regulation
0:30:51 because it’s just a regulated field, but it should be
0:30:55 extremely permissive and for experimentation.
0:30:58 I think we found out some of the things that the financial
0:31:03 system had are things that are deeply desirable in crypto
0:31:04 as well, one by another.
0:31:06 I’m sure there’s like, everyone can take a different
0:31:09 position on this, but like that’s there.
0:31:14 In an AI case, man, like I just like, you gotta,
0:31:18 like you can regulate institutions or like ground rules
0:31:21 for markets, but regulating technology is like,
0:31:24 that seems like a crazy position to take.
0:31:27 Like suddenly floating points are like illegal.
0:31:29 That’s after you do too many of them.
0:31:34 – It is so weird, you know, I had a conversation
0:31:37 with a policymaker and I was like, so if this was nuclear,
0:31:40 you want regulate nukes, you’d regulate physics.
0:31:43 That’s your idea like physics is not illegal.
0:31:48 – Yeah, throw away those textbooks immediately, please.
0:31:53 – So again, I think maybe even some of the things
0:31:56 that people have identified as potential problems
0:31:59 in the future, I guess we could agree with,
0:32:04 but like again, the proposed solutions to this are crazy
0:32:06 because we get into these insane places where like,
0:32:09 cool, maybe analog computing is actually better
0:32:10 for training neural nets.
0:32:13 So that is not floating point operations anymore.
0:32:15 I was like, are you doing that now
0:32:18 because of regulations instead of because of pragmatism?
0:32:20 Or, you know, it just, I don’t know.
0:32:22 It just seems too early.
0:32:27 And it, I think that it’s a very, very bad idea right now
0:32:30 for any governments to set ceilings and rules
0:32:34 that are especially onerous for startups,
0:32:35 because I think all you’re gonna get out of that
0:32:39 is just like, you’re gonna get the place
0:32:42 where it’s gonna be all behind four-pay APIs
0:32:44 that are gonna be all controlled by the few.
0:32:49 I mean, that’s strictly better than it doesn’t exist, granted.
0:32:52 But like, clearly we need open source.
0:32:54 Look at everything that happened just now
0:32:58 after AI became available first
0:33:02 and then thanks to Metta and releasing amazing Lama models.
0:33:04 We know so much more about what Nuance do.
0:33:08 We know so much more about how to accelerate this.
0:33:11 You need to rescue these things sometimes
0:33:12 from the ivory towers
0:33:15 and you have to give them to the tinkerers.
0:33:18 And the people who remember how to memory align things
0:33:22 and what, you know, how to get the most out of a hardware.
0:33:23 Like that’s a different crowd than the people
0:33:27 who were writing Python neural topologies.
0:33:30 And it’s, we need everyone on this.
0:33:32 And then we’re gonna get the best outcome
0:33:34 because it is foundational for our future.
0:33:36 – Yeah. – Also, we need to figure out
0:33:38 how to make crypto closer aligned with AI
0:33:43 because it’s a natural way for how to pay for micropayments
0:33:45 and tokens and so on.
0:33:47 And this is really, really important now.
0:33:48 – Well, and solve so many.
0:33:50 I mean, it’s interesting.
0:33:52 Crypto solves so many AI problems
0:33:54 and AI solves so many crypto problems.
0:33:55 – Yeah. – They’re–
0:33:57 – These should be symbiotic
0:34:00 and just the sine waves of excitement of these two areas
0:34:02 just happen to be dissonant right now
0:34:04 and we need to make them harmonious partly
0:34:07 because of the bad big companies.
0:34:08 (laughs)
0:34:09 Sorry.
0:34:10 Well, that’s great.
0:34:14 So we are ready to take questions.
0:34:16 Any questions you may have?
0:34:18 Yes, sir, back there.
0:34:22 – If crypto plays out in exactly the same–
0:34:24 (mumbles)
0:34:30 – It’s like, I think very naturally the browser
0:34:32 will just like be able to move value around
0:34:34 and you will want to move value
0:34:36 in a way that can be addressed by the browser.
0:34:40 This is like, you can get pretty close to that
0:34:42 but you would have to squint
0:34:44 and you have to deal with like poor UX
0:34:46 and you have to deal with hard on ramps.
0:34:50 I think, I don’t think anything kind of magical happens
0:34:52 other than that people will have,
0:34:55 they need more utilitarian ability to like use money.
0:34:57 Again, I think if you run the counterfactual
0:34:59 and say open AI just for whatever reason
0:35:02 you couldn’t work for Stripe and just decided to,
0:35:05 I don’t know, like use some level two blockchain
0:35:10 and to pay for tokens as a initial MVP.
0:35:15 People would now move incredible amounts of money around
0:35:18 and build startups that might have these kind of ideas
0:35:19 that Sam talked about on stage,
0:35:24 about people sharing value of creating GPTs
0:35:27 and like all of the, you know, the blockchain,
0:35:30 like if you look at Open AI Dev Days,
0:35:31 especially the last 20 minutes
0:35:36 and you open at the same time VS Code to Ethereum contract,
0:35:39 you can basically write it out while he’s talking.
0:35:43 It’s like, so, we have this incredible infrastructure
0:35:45 but we can’t use it part because of fees,
0:35:46 part because of speeds,
0:35:49 because like you gotta solve the infrastructure problems
0:35:50 which I know we are doing.
0:35:52 Like this is people are building through this particular winter
0:35:53 in a way that is like,
0:35:57 seems utterly ideal from my perspective.
0:35:58 So that’s great.
0:36:03 There is, we need to bring killer use cases.
0:36:06 We need to find, like we need to have people
0:36:07 want to use this.
0:36:13 Unfortunately, there is like the focus on these kind of things,
0:36:16 especially once you need culture and social things,
0:36:18 is annoyingly not linear.
0:36:20 And that’s really trips everyone up.
0:36:21 It’s like, it’s nothing, nothing, nothing, nothing,
0:36:22 everything.
0:36:25 And because we have to get,
0:36:28 we have to like toil in obscurity
0:36:31 to build up the infrastructure
0:36:33 until it can support the kind of way
0:36:35 that people want to use it.
0:36:37 And then we need to catch lightning in a bottle.
0:36:39 Yes.
0:36:40 – Thanks, Sabi.
0:36:42 I had a question.
0:36:44 So Shopify works with a lot of merchants and brands
0:36:47 that might not be as tech savvy
0:36:49 or might not understand everything that’s going on
0:36:51 in like AI or crypto.
0:36:54 How does Shopify think about messaging to those merchants
0:36:56 and saying, hey, this is the future.
0:36:58 These are things that you should be using.
0:37:01 – Yeah, I think people actually hire us
0:37:04 for having them be ideal for whatever the internet is
0:37:07 and for whatever their buyers,
0:37:09 like they want to continue buying.
0:37:11 Like again, we start before mobile phones.
0:37:14 And so like just making sure that everyone’s store
0:37:16 is available to some mobile phones
0:37:18 back in those days, responsive, whatever,
0:37:22 just had it tremendously because their colleagues
0:37:24 who didn’t have that like just like started doing poor
0:37:29 and poor as money moved to the small form factors.
0:37:32 And they started out competing their peers in the industry.
0:37:35 So like as this is a super old and probably dumb example,
0:37:38 but like what I’m saying is like the demand has to come up.
0:37:42 Like the merchants offering like you can’t solve this problem
0:37:44 just simply by bringing the supply side up,
0:37:45 the demand side needs to go up.
0:37:50 The buyers should want to spend on the blockchain.
0:37:54 Now, that is the huge beauty.
0:37:58 You have 3% to work with here.
0:38:00 Like the credit card fees are high, right?
0:38:02 Like there’s a lot of interesting incentive systems
0:38:03 that you can do.
0:38:06 But what I found was the most amazing thing about sort of
0:38:08 when I came back to crypto,
0:38:10 paid more attention, smart contracts, if you’re,
0:38:14 is that, and I actually think even the practitioners
0:38:19 in the crypto space are under appreciating this
0:38:22 as a talent is there’s never been a field
0:38:25 of such applied widely distributed
0:38:28 and high quality thought applied game theory, right?
0:38:31 Like it’s the amount of the constructions
0:38:34 and the incentive systems that are being built
0:38:36 in this community are like tremendous.
0:38:41 Like the artists that have done really, really well
0:38:45 through NFTs like a such a good example.
0:38:48 This is the first, and now this is like,
0:38:52 man, this was the first business model of the arts
0:38:55 that we’ve gotten since patronage times, right?
0:38:57 Like there’s a huge breakthrough
0:38:59 and we’ll make a comeback.
0:39:02 This thought, this quality of thought has to be applied
0:39:06 I think to all sorts of other areas and in different ways.
0:39:09 And we can use the infrastructure that’s been built
0:39:12 to do something way better than cash back credit cards
0:39:15 like the first extra 3%.
0:39:17 Now it needs to also be meet the world where it is.
0:39:21 It requires some kind of escrow systems for charge back.
0:39:22 There is real distribution.
0:39:24 There needs to be like it,
0:39:26 not everything has to be trustless.
0:39:30 We should bring trust progress into the systems.
0:39:33 Like it’s called a wallet, which I think is a good name.
0:39:36 But in my wallet is, I don’t think I’ve cashed my wallet.
0:39:40 It’s all cards, it’s like licenses.
0:39:41 It’s like driver license and so on, right?
0:39:45 Like it’s attestations that I’m, who I say I am.
0:39:49 And bringing in trust brokers into the system
0:39:51 through the primitives we have now and saying,
0:39:54 yeah, that person is who we say they are.
0:39:56 Yes, that person can be shipped to.
0:40:01 Like if you write this thing on a package destination,
0:40:05 FedEx knows how to unclog this into an actual destination address.
0:40:08 We have to build this with the industry.
0:40:11 And then at some point,
0:40:15 we’re going to just have a much better environment for commerce.
0:40:18 And much like it’s not slightly better than using credit.
0:40:20 That it can actually be 10 times better.
0:40:22 People move around more since COVID, right?
0:40:25 Like how often have you gotten a package
0:40:27 to some other address that you’re no longer there?
0:40:30 Like this, it would be wonderful to make addresses pointers
0:40:32 that can be updated in the background and so on, so on, so on.
0:40:38 So like work with the actual infrastructure of the world
0:40:41 and modernize it on these new systems.
0:40:45 And so that all that want to accrue to people who want to take up one,
0:40:47 like who actually live in a real world
0:40:52 and not just like the hindsight synonymous, you know, abstraction.
0:40:55 I think that’s absolutely doable.
0:40:58 But man, do we have to get fees down?
0:41:00 And this is like it’s never going to happen
0:41:03 if we go back to a $50 transaction.
0:41:04 Yes.
0:41:07 We’ve got time for one more question.
0:41:10 Yes, right in the front row, since I saw you first.
0:41:13 So Shopify has a really successful partner program
0:41:16 where engineers are building themes and plugins and apps.
0:41:19 Could you talk a bit more about some of the challenges
0:41:20 starting that system up?
0:41:23 And then also, as you see more and more people speak the language
0:41:24 of NFTs and the EVM,
0:41:27 how that can evolve as we get more interoperability.
0:41:30 This is definitely, like, you’re sort of hinting at
0:41:34 how cool would it have been to have these pieces of infrastructure primitives
0:41:37 and we could have governed the entire partner system on a token economy.
0:41:38 Like, that’s totally true.
0:41:43 I think this is, I really hope someone is going to explore
0:41:46 how to build systems like the Shopify ecosystem
0:41:53 because there’s a lot of sharing and a lot of multi-volume transactions
0:41:56 which we are all doing with ledgers in the background.
0:41:59 Putting it up, like putting up any two-sided marketplace is tricky.
0:42:02 Here you have to overcome the chicken-necked problems.
0:42:05 It’s tricky, but if you manage to do it,
0:42:08 this is also really, really self-sustaining afterwards.
0:42:10 We have one of the greatest things.
0:42:14 We have about five different customers
0:42:19 with merchants who start on Shopify who now IPO’ed and are public companies.
0:42:23 We now have one who is building an app,
0:42:26 like on the Shopify app with their own private company,
0:42:32 Public Company Clavio is now in their Bay Bay successor, which is awesome.
0:42:36 Again, this is a wonderful thing of what you can do.
0:42:40 Companies are estimated to produce about six times more value
0:42:47 than they capture to the world, like enlightened self-interest novice.
0:42:53 The constrained vision in so-called terms is real.
0:42:57 And there are so many ways to build,
0:43:01 like so many businesses lend themselves to being turned into platforms later.
0:43:03 The problem is everyone tries to build a platform first,
0:43:05 and that’s really, really, really hard.
0:43:10 It’s much better to extract a platform out of a working piece of software
0:43:13 that people are already using, and so that helped us.
0:43:17 People were already trading themes as zip files,
0:43:19 and we were like, “Okay, well, we can help with that,”
0:43:23 and people are already building things for themselves
0:43:26 on the APIs for integrating their own ERP systems,
0:43:31 and then building this to the point where you could list this ERP connector
0:43:37 and build it like our SDKs made it so that it can be in so many stores.
0:43:41 But, de-forward, we hinted people and helped them into the right direction,
0:43:45 and all these kind of things ended up being incredibly valuable for the community.
0:43:50 I mean, it’s definitely one of the coolest aspects to build,
0:43:53 but again, at the end of the day, what all of this is,
0:43:59 is you build really good software that solves real problems very, very quickly,
0:44:03 and then you build a bit of a game-theoretical system on top,
0:44:12 in which people being selfish about wanting to build something accrue a huge amount of value
0:44:19 to the community, to Shopify, to the merchants, and so on.
0:44:26 I mean, I feel like the most sophisticated thought along those lines on Shopify
0:44:30 is a fairly amateur sort of baby’s first incentive system
0:44:37 compared to some of the things that were built on Ethereum and the blockchains
0:44:42 in terms of sophistication, except everything in the world of crypto
0:44:45 was so pointed at financialization and finances,
0:44:49 but I think people might have missed what could be done
0:44:52 if you think more about products rather than financial products,
0:44:55 and I think that’s hopefully going to be the next wave of crypto,
0:45:00 it’s going to be more products and solving real problems for people.
0:45:01 I’m that amazing insight.
0:45:04 I would love to please join me in thanking Koby.
0:45:05 Thank you.
0:45:08 [Music]
0:45:10 [Music]
0:45:12 [Music]
0:45:14 [Music]
0:45:16 [Music]
0:45:18 [Music]
0:45:20 [Music]
0:45:22 [Music]
0:45:24 [Music]
0:45:26 you
0:00:02 (upbeat music)
0:00:04 When Shopify was founded in 2006,
0:00:07 it was just a year prior to the launch of iPhone.
0:00:10 Needless to say, smartphones were a far cry
0:00:11 from the ubiquitous accessory
0:00:13 that now billions carry around.
0:00:17 And fast forward to today, nearly 20 years later,
0:00:19 where Shopify merchants can accept cryptocurrency,
0:00:21 create immersive shopping experiences
0:00:22 with augmented reality,
0:00:26 and even use AI to generate product descriptions.
0:00:29 So how does a nearly $100 billion company
0:00:30 stay ahead of these trends?
0:00:33 And what will shopping look like in the generations to come?
0:00:36 Plus, why would A16Z never have invested in Shopify
0:00:38 in its early day?
0:00:40 And why has the CEO of this company
0:00:42 chosen to continue writing code?
0:00:44 These are just some of the questions covered
0:00:45 in today’s episode,
0:00:48 the conversation the Shopify co-founder and CEO
0:00:52 to bias Luki and A16Z co-founder, Ben Horowitz.
0:00:54 This episode was also originally published
0:00:57 on our sister podcast, Web3 with A16Z.
0:01:00 So if you are excited about the next generation
0:01:03 of the internet, find Web3 with A16Z
0:01:04 wherever you get your podcasts.
0:01:06 All right, let’s get started.
0:01:14 – Welcome to Web3 with A16Z,
0:01:17 a show about building the next generation of the internet
0:01:20 from the team at A16Z Crypto.
0:01:22 Today’s episode features Tobias Lutke,
0:01:26 CEO and co-founder of the e-commerce platform Shopify.
0:01:29 Speaking with A16Z co-founder, Ben Horowitz,
0:01:32 at our second annual founder summit in November,
0:01:35 they discuss what it takes to build a breakout startup
0:01:39 in a crowded category, the changing face of retail,
0:01:41 how to affect change in the workplace,
0:01:44 and how to handle individual emotions and corporate culture,
0:01:47 including dealing with calls for activism,
0:01:50 as well as the value of embracing negativity.
0:01:52 They also touch on the moral imperative
0:01:54 behind creating quality software,
0:01:58 the symbiosis between AI and crypto, and more.
0:02:01 As a reminder, none of the following should be taken
0:02:04 as business, legal, tax, or investment advice.
0:02:07 Please see A16Z.com/disclosures
0:02:08 for more important information,
0:02:11 including a link to a list of our investments.
0:02:19 – All right, so for those of you who don’t know Toby,
0:02:20 which you probably should know Toby,
0:02:22 you can look him up as well,
0:02:27 but he started a snowboarding company
0:02:29 in 2004 called Snowdevil,
0:02:33 and he parlayed that into Shopify,
0:02:37 which now serves millions of merchants
0:02:40 in 175 countries around the world.
0:02:42 How many countries are there, like in total?
0:02:44 – I think that’s most of them, there’s 200?
0:02:45 – Yeah, 200, yeah, but–
0:02:48 – There’s changes around the margins.
0:02:51 – You’ll get them all at some point.
0:02:54 And then he was also interesting for this crowd.
0:02:58 He was one of the first adopters of Coinbase Commerce.
0:03:02 So Toby’s been kind of big on crypto for a while now.
0:03:04 But let’s get started.
0:03:08 So you started as a snowboarding company,
0:03:12 and I always find that funny because of all the CEOs
0:03:16 that I know you’re probably the least snowboard kickback,
0:03:19 like let’s hit the powder dude.
0:03:21 So how did that happen?
0:03:23 Like why did you start out that way?
0:03:28 – Yeah, you know, when you talk to governments,
0:03:32 they always, Shopify is doing a lot about entrepreneurship,
0:03:35 which I was a lot of entrepreneurship,
0:03:38 and it’s such a brilliant institution.
0:03:39 I know everyone here is a card-carrying member
0:03:42 of the entrepreneurship is good club,
0:03:45 but politicians usually, they’re sort of vague on it,
0:03:50 but they say like they want to pass government programs
0:03:52 and policies to cause more entrepreneurship,
0:03:55 which I always think is the wrong way around.
0:04:00 But there’s actually, so the funny story is that,
0:04:03 so I was in Canada at this point, I came from Germany,
0:04:06 living in Canada 2004, three, four,
0:04:10 and I found out I cannot actually get a job.
0:04:13 I was trying to be hired as a computer programmer locally,
0:04:15 but I found out I did not have a work permit.
0:04:16 I didn’t know what a work permit was.
0:04:20 So turns out that you are in Canada,
0:04:22 allowed to start a company even without a work permit.
0:04:23 So there’s one government program
0:04:27 that actually has led to what company is being created.
0:04:29 – Well, that’s like the best government program ever.
0:04:32 Just like the payoff on Shopify paid for itself
0:04:33 a million times.
0:04:35 – And it turns out Canada is very cold.
0:04:37 So I was doing snowboarding, so I knew something about it,
0:04:40 and I started an online store for snowboarding.
0:04:42 It’s kind of a pretty simple story like this.
0:04:44 – That’s amazing.
0:04:47 So basically the key to your whole career
0:04:49 was not being able to work.
0:04:49 – Yes.
0:04:51 (laughing)
0:04:53 In fact, you know, what was in my mind too is like,
0:04:56 I was like the solution with programming at this time.
0:04:57 It was like sort of a Java word,
0:04:59 kind of was kind of very oppressive.
0:05:04 And I wanted to reclaim it as a hobby.
0:05:06 So I was like, cool, if I sell snowboards online,
0:05:08 I can make money while snowboarding,
0:05:09 and I get to do more fun stuff.
0:05:14 And it’s been extremely unsuccessful as a strategy.
0:05:17 I basically never had time to snowboard again
0:05:17 after I started.
0:05:20 (laughing)
0:05:22 – The end of your snowboarding career
0:05:24 was the start of your snowboarding business.
0:05:25 That’s pretty remarkable.
0:05:27 – Maybe I would project more back at this point
0:05:29 if things would have gone in the counterfactual way.
0:05:31 – Yeah, now the other thing that really strikes me
0:05:34 about Shopify is that, you know,
0:05:36 and I hate to say this as a venture capitalist
0:05:38 who some people think is a smart person,
0:05:42 but we probably would have never invested in Shopify
0:05:43 when you started it
0:05:47 because it was probably the most over-competed
0:05:48 category in the world.
0:05:52 Like everybody had some kind of e-commerce platform.
0:05:56 I mean, there was just so, so many of them,
0:06:00 Magento, this, that, the other.
0:06:05 So did you go, oh, no, no, no, like,
0:06:07 I’m gonna build this thing.
0:06:09 It’s gonna dominate everything.
0:06:10 Then I’m gonna have a network effect
0:06:12 and I’m gonna build, you know,
0:06:14 like one of the biggest things that anybody’s ever seen.
0:06:19 Like, how did you like get all the way to where you are?
0:06:20 Like, what was that path?
0:06:24 – Yeah, no, there’s no plan like this.
0:06:29 Just like, I think, for every entrepreneur,
0:06:31 it’s a useful thing to think about
0:06:33 what is your energy source, right?
0:06:35 Like what is actually driving you?
0:06:37 And sometimes it’s greed.
0:06:40 And we, you know, like the sort of negative emotions
0:06:42 channeled into building actually
0:06:44 the most powerful energy sources.
0:06:47 I had this, you were right.
0:06:48 Like, I mean, Netscape, IPO,
0:06:52 but like, you have to explain every platform
0:06:54 in terms of an old platform and, you know,
0:06:57 online hosted CS catalog with buy buttons
0:07:00 was like kind of a way you would mark
0:07:03 to describe even what the internet would be used for, right?
0:07:04 Like, so e-commerce was like right there
0:07:07 from the beginning, I thought there would be lots of software
0:07:09 that I could use and I didn’t find anything.
0:07:12 I was actually really upset with the quality of the software.
0:07:15 And so like, I channel that into building something
0:07:17 and that was my energy source.
0:07:19 And I just love the building.
0:07:20 – That’s actually, by the way,
0:07:23 one of the best entrepreneurial ideas there is,
0:07:25 which is this stuff sucks so bad.
0:07:27 Yeah, I’m so mad about having to use it
0:07:29 that I’m going to build my own.
0:07:32 – Totally, I honestly, it fuels me to do it this day.
0:07:34 I’m just, I still think it’s ridiculous
0:07:37 what people are subjected to in terms of quality of software.
0:07:39 It’s like, oh, it’s weird.
0:07:41 I probably did some like neurological,
0:07:43 like kind of rearrangement there,
0:07:44 but like to me, this is almost a moral issue
0:07:48 of like when people like confused with software
0:07:50 and they think about themselves as like being inadequate.
0:07:53 It’s like, how dare software make humans feel that way?
0:07:54 Right, like it’s just not okay.
0:07:57 So you want to make it a more approachable,
0:07:59 but like, I think the thing that people might have missed
0:08:02 in e-commerce or like, it’s not that I had a great insight.
0:08:05 I just sort of organically got there is bad.
0:08:06 Yes, there was a lot of e-commerce software,
0:08:09 but it was all built for already rich retailers
0:08:11 who needed to move online.
0:08:13 Then that was a phrase people used.
0:08:17 And there was no one started software
0:08:20 for starting more entrepreneurial process online.
0:08:23 Like actually the digital native, what we call it now also,
0:08:26 like if you like that term,
0:08:29 because there’s a totally different set of requirements
0:08:30 for the people who are just starting out.
0:08:32 And so that’s what we did.
0:08:33 So this is the thing, right?
0:08:38 Like it’s like, when you talk,
0:08:40 that’s like, I don’t know, 60 words per minute or something,
0:08:42 and you type fast, it’s like twice that.
0:08:45 But like, if you can do customer consultation
0:08:49 in your own brain, the brain runs at like terabytes per second
0:08:50 by bandwidth, right?
0:08:53 Like so, if you know the use case,
0:08:56 if you’re building for yourself or your past experience
0:08:58 or some kind of, if you have a really good model
0:09:01 of your customers or use cases in your mind,
0:09:03 you can do customer development at high bandwidth
0:09:06 with low latency and you can just build something
0:09:08 that you know needs to exist.
0:09:10 – Yeah, no, that’s incredible.
0:09:13 So one of the things, you’re also kind of been
0:09:19 a kind of a special kind of, or a different kind of CEO
0:09:24 than some of the ones that people read about
0:09:27 such as Adela or Bob Iger and that you have a reputation
0:09:31 for being extremely hands-on, including writing code.
0:09:34 You still have the fury that started the company.
0:09:38 How do you think about that in terms of,
0:09:40 you’re not kind of paying attention
0:09:42 to all parts of the company equally.
0:09:46 You’re clearly doing certain things and not others.
0:09:49 How do you think about that as an effective
0:09:51 kind of management style?
0:09:56 – Yeah, I mean, I imagine the people in my company
0:09:58 would disagree with your assessment
0:10:00 about some effective management style.
0:10:02 So well, it’s worked pretty well.
0:10:07 It’s working, but it certainly doesn’t proxy
0:10:11 to what people sort of, like it’s not a CEO job
0:10:14 out of central casting that I’m trying to perform, right?
0:10:17 I tried that on and it didn’t work for me at all.
0:10:20 I actually find, like I value being in all the details
0:10:23 and I ask it from all the people who are around me
0:10:25 and report to me, like almost everyone,
0:10:29 especially after COVID, I turned over everyone,
0:10:33 but one of my executives during the first 10 months of COVID.
0:10:35 And now almost everyone reports to me,
0:10:38 it’s actually a AX founder, maybe someone be acquired
0:10:41 or someone who started a company before coming to Shopify.
0:10:45 So because there’s this, I just want,
0:10:48 I find it actually really inefficient,
0:10:50 not understanding the details of what we’re doing, right?
0:10:51 Because then–
0:10:52 – Well, then you’re guessing.
0:10:53 – Then you’re guessing, right?
0:10:57 And you’re like, well, when something goes wrong,
0:10:59 now you not only need to fix it,
0:11:02 you actually have the first cram for,
0:11:04 like you have to learn like three months
0:11:05 before you can actually do a fix.
0:11:06 If you understand what’s going on
0:11:09 or what ought to be existing, it’s much faster.
0:11:13 And you want to train people on the company mission, right?
0:11:16 Like because there’s a thing that you want to exist,
0:11:18 like even if you don’t know exactly what it is,
0:11:21 like you know which direction you want people to go, right?
0:11:24 And getting everyone aligned to go in this direction
0:11:27 is only possible if you can paint picture
0:11:30 of like how does this area actually sum together
0:11:33 all the way to helping the mission overall.
0:11:38 And so, and this is sort of a sort of unique thing
0:11:39 about our times right now.
0:11:41 It’s like the infrastructure keeps changing,
0:11:44 like here’s crypto, here’s AI, here’s,
0:11:46 I mean, I started again almost,
0:11:48 like next year is 20 years ago, right?
0:11:51 So there was no mobile phone, like that offer quality
0:11:53 that we have now.
0:11:54 – Right, that’s right.
0:11:55 You started before the mobile phone.
0:11:58 – Right, so it’s like, that was one of the things
0:11:59 that just sort of happened.
0:12:03 And, you know, SaaS software was new at this time.
0:12:05 So, you know, the platform’s capabilities change
0:12:08 and you want to adopt them to like your mission
0:12:10 to what is now possible.
0:12:13 And so you have to be able to understand, well,
0:12:15 but in a kind of like ideally in your head,
0:12:21 in the world where, I take a very long-term view
0:12:24 and my biggest fear that I have
0:12:27 is that Shopify winds up being a fantastic solution
0:12:30 to a problem that people no longer have, right?
0:12:32 Or no longer solve this way.
0:12:33 So I’m trying to like need to keep it current
0:12:35 and being able to run the counterfactual about
0:12:39 this is not possible, this is what we’re trying to accomplish.
0:12:41 But if this would have been possible from the beginning,
0:12:43 we would have built this entirely different thing.
0:12:45 So now our job’s to get from here to there
0:12:48 because, again, we want to keep it current and idea.
0:12:49 – Right, you know, that’s such a great insight
0:12:53 because they think that, you know, when I work with CEOs
0:12:55 and you think about the job,
0:12:58 so much of the job is making high quality decisions
0:13:00 and setting the direction of the company.
0:13:03 And in some respects, neither feels like work.
0:13:07 You know, because you’re working on something else,
0:13:10 like how does AI work, what’s possible,
0:13:13 in order to make decisions and set the direction.
0:13:16 But what you’re doing seems actually removed
0:13:20 from anything that the company needs to do right now.
0:13:22 And I think so many people kind of neglect
0:13:24 that kind of thing for that reason.
0:13:27 The other thing that I always find is,
0:13:29 at any given point in time,
0:13:32 some things in the company are very important
0:13:33 and very high leverage.
0:13:35 And some things are just not important
0:13:37 for you to put your time into.
0:13:39 And people often get lost in that.
0:13:43 They think, oh, I’ve got to talk to this person
0:13:46 and then that person and then the other person and so forth.
0:13:48 And it’s like, no, no, you don’t.
0:13:49 You’ve got to make good decisions.
0:13:50 You have to set the direction.
0:13:52 And if you don’t do that, nobody’s doing that.
0:13:54 And then the company’s going to fail.
0:13:57 So that’s such an amazing way that you’ve gone about that.
0:13:59 – If you’re in a small team,
0:14:01 like for a new thing that your company needs to do
0:14:04 and you’re in a small team of people who are like,
0:14:08 know or like, figure out the shape of it early,
0:14:09 if it’s actually as important,
0:14:11 it will itself have a gravity
0:14:14 to change the way the company thinks about quality of software
0:14:16 shipping, the way the architect thinks.
0:14:22 And it’s also after this comes up as a new thing,
0:14:26 you then have the ability to like a type of legitimacy.
0:14:27 As a founder,
0:14:30 you have a huge amount of legitimacy in a company already.
0:14:33 But like if you also are there with the new project,
0:14:35 that’s important, that needs to be a resource and say,
0:14:38 this is what we’ll do.
0:14:40 You don’t have to have all these conversations.
0:14:43 People love clarity of like, this is what we do.
0:14:45 Like it has a gravity all to itself.
0:14:48 That’s unlike anything you can accomplish
0:14:50 by one-on-one consultation,
0:14:53 trying to talk everyone through the entire decision matrix.
0:14:56 And so what is Toby doing over there?
0:14:57 – Oh, that must be important.
0:14:59 – Yeah, that’s exactly right.
0:15:01 We’re doing like a job site click now.
0:15:04 And like, then I don’t have to figure out like,
0:15:06 yes, there’s like 15 teams that are already doing AI stuff
0:15:09 and they ought to be in some kind of constellation.
0:15:12 But now this AI assistant thing is going to run through
0:15:16 and the company, everyone wants to be a part of it.
0:15:18 And this is like, all these problems kind of disappear.
0:15:19 It’s like, it’s maybe not the most effective-
0:15:20 – Like a magic trick.
0:15:21 – Yeah, yeah, yeah.
0:15:23 Yeah, actually, I think by the way,
0:15:25 like everyone in this,
0:15:28 like everyone here has clearly read your books.
0:15:32 And if you haven’t done do that, like I think-
0:15:33 – Thank you.
0:15:35 – The best books on leadership, honestly.
0:15:38 Thank you so much for sharing all this.
0:15:40 I think in some of what you described,
0:15:43 but there’s only very few ways to do a change management.
0:15:45 One way of doing it is to something super, super,
0:15:49 super surprising and then explain everyone why you did it.
0:15:51 It’s the fastest way to get a large company.
0:15:53 In my shop, there’s about 10,000 people,
0:15:55 much, much, much faster way to do a change management
0:15:56 than hand-to-hand combat.
0:15:58 – It’s like shock therapy.
0:16:00 It’s like, what the fuck was that?
0:16:02 Oh, well, this was that, that was okay, got it.
0:16:06 Very, very, very high retention on that.
0:16:07 So let’s talk about, you know,
0:16:10 one of the things that has come up, you know,
0:16:13 a lot over the years that you had probably, you know,
0:16:16 maybe the best statement on, I thought,
0:16:20 is politics, you know, politics, activism,
0:16:22 how that’s kind of come into companies
0:16:25 has taken over a lot of the activity of companies.
0:16:29 And how do you, I guess, how do you think about that?
0:16:33 And then why’d you say what you said?
0:16:35 And then how’s that going?
0:16:39 And then how are you able to sustain it particularly
0:16:42 as things get even hotter now with, you know,
0:16:44 what’s going on in the Middle East and so forth?
0:16:47 You know, it’s gotten even more intense than it was,
0:16:49 you know, amazingly in 2020.
0:16:51 – That’s a complex topic.
0:16:56 I can, here’s, I mean, there’s a lot of conflicting ideas
0:16:58 about what people want companies to be.
0:17:02 And, you know, I think, especially in 2010
0:17:06 and maybe now again, a lot of people would like companies
0:17:08 to play the role of some kind of phantom limb
0:17:13 of what I think the whole left over by the stories
0:17:16 that people are surrounded with, like maybe with governments
0:17:17 or like, like, and so on.
0:17:20 I suddenly, people want, like I have gotten tons
0:17:23 and tons of petitions for Shopify to take stances on issues.
0:17:27 And I think everyone’s feeling the same thing.
0:17:30 This was weird to me, but like initially,
0:17:32 but like, I found myself generally agreeing
0:17:34 with what people identify as the problems
0:17:36 when this happens.
0:17:38 But I found myself, they really, if ever,
0:17:41 really agreeing with the proposed solutions
0:17:41 to those problems.
0:17:44 And I think it’s kind of important to say like,
0:17:46 hey, unless we are all aligned on this,
0:17:48 we’re probably not doing this.
0:17:51 Because, you know, at the end of the day,
0:17:53 like companies ought to be thought of
0:17:54 as a bit of a simpler thing.
0:17:57 It’s, you know, real love entrepreneurship.
0:17:59 Entrepreneurship is, you know,
0:18:01 liberal values kind of institution of, you know,
0:18:05 you’re like reaching for independence,
0:18:07 reaching for like, to better yourself,
0:18:09 like building something like enlightened,
0:18:10 the ideas of enlightened self-interest.
0:18:13 There is a bit of a political coding in this,
0:18:14 people at least think so.
0:18:16 And someone else might take a totally different opinion
0:18:17 about all these kinds of things.
0:18:18 So that’s cool.
0:18:20 The cool thing is companies themselves explorations
0:18:22 into a set of ideas.
0:18:24 Sometimes encoded in a mission sometimes by the founders.
0:18:27 And if we make every company the same,
0:18:29 then we lose that.
0:18:31 That’s actually like a weird form of diversity itself,
0:18:32 if you think about it.
0:18:36 So anyway, I went ahead and just like said,
0:18:40 like look, we cause more entrepreneurship to exist.
0:18:43 And sometimes there are social causes
0:18:45 that are aligned with this.
0:18:48 And then maybe we become active because of our mission.
0:18:50 But outside of that, you know,
0:18:53 everyone makes money and you can do whatever you want
0:18:54 in your after hours.
0:18:58 And that’s just a simpler way of thinking about the business.
0:19:02 Of course, that’s not universally loved,
0:19:05 but like after being there, they clear about it
0:19:05 and explaining it.
0:19:09 They explain, you know, we give everyone when we hire people
0:19:11 a letter of like, here are the reasons,
0:19:13 like before you sign.
0:19:15 Here’s the reason why you might not want to work
0:19:18 for Shopify because like here’s a set of ideas
0:19:19 that you disagree with.
0:19:22 And then that clarity is wonderful.
0:19:24 We keep everything so nebulous.
0:19:28 I think the world of marketing and PR
0:19:31 has just done a number on us all to be so non-committal
0:19:32 to everything.
0:19:33 – Yeah.
0:19:35 – And the problem is if you’re non-committal,
0:19:37 you leave a lot of vacuum.
0:19:38 – Right.
0:19:40 – And that vacuum is going to be filled by someone.
0:19:43 And it’s usually the people who want to fill it
0:19:46 with bad faith takes, who fill it.
0:19:50 So if you just clear about it, it works sometimes,
0:19:54 like this is something I might get into hot water with,
0:19:57 but like sometimes it’s actually good to heighten
0:19:58 the misalignment for the kind of people
0:20:00 that you don’t think should be working in a company.
0:20:04 Like, hey, we work hard and it’s kind of an-
0:20:06 – Yeah, we have poor work-life balance.
0:20:07 – Yeah, yeah.
0:20:08 – Sorry.
0:20:09 – That’s a good-
0:20:11 – You are free to work elsewhere.
0:20:15 – If it’s true, put it up, like light it up on a sign
0:20:17 because otherwise you’re going to end up
0:20:20 with the weirdest divisiveness in the company
0:20:22 because of basically false advertising.
0:20:25 So I think it’s better to just be clear.
0:20:29 So I mean, what is it like?
0:20:31 There’s a lot goes into this.
0:20:33 Like again, culture is unstable.
0:20:36 We know from the internet that every community
0:20:40 that anyone likes Reddit, like Hacker News,
0:20:42 if anyone actually likes this page,
0:20:47 it’s a product of unbelievably dedicated moderators
0:20:51 who are keeping this discourse well.
0:20:54 I think that’s an idea that I think companies
0:20:55 have started adopting too.
0:20:57 Like once you’re a couple of thousand people,
0:20:59 your Slack channel is the internet basically
0:21:02 and you need to apply similar rules.
0:21:04 You need to be able to take people to a sign saying,
0:21:05 hey, here’s the thing you’re causing.
0:21:07 When you’re sending a message to a channel
0:21:09 with 3,000 people in it,
0:21:11 that is basically the same as you spending a day
0:21:15 typing individual text message to any person in the channel.
0:21:16 And would you do that?
0:21:18 If no, probably don’t put in this kind of stuff.
0:21:22 Yeah, that’s really such a great insight that you had that.
0:21:24 Hey, I actually kind of agree with the intentions
0:21:26 that people have on most of this,
0:21:30 just not with the ideas on how to achieve those intentions
0:21:35 because that’s where politics gets really dangerous
0:21:38 in a company and I have a funny story about this.
0:21:41 So my great uncle Harold, my grandmother didn’t speak
0:21:43 for 20 years, brother and sister.
0:21:45 And I asked my father a few years ago,
0:21:46 I said, like, what happened there?
0:21:47 Why didn’t they speak?
0:21:50 He’s like, oh, well, ’cause your great uncle Harold
0:21:51 was a Trotskyist.
0:21:53 And I was like, well, what was grandma?
0:21:55 She was a Stalinist.
0:21:58 They’re two slightly different kinds of communists
0:21:59 and they didn’t speak for 20 years.
0:22:01 And that’s kind of what happens inside a company.
0:22:04 It’s not like you’re a good person, you’re a bad person.
0:22:06 This is a moral issue.
0:22:10 It’s like, well, yeah, we all want way fewer people
0:22:12 to get killed in this current conflict,
0:22:14 but how you achieve that,
0:22:16 that gets very complicated very fast.
0:22:19 And so if you let people attack each other,
0:22:22 shut down conversations, intimidate each other
0:22:26 inside your company around that, it’s just all bad.
0:22:27 – Yeah, but fighting gets the fiercest
0:22:29 when mistakes are the lowest.
0:22:31 You know, like when people almost agree,
0:22:34 when they become mortal enemies.
0:22:35 – Right.
0:22:36 – It’s a very funny effect.
0:22:38 I mean, you can.
0:22:39 – Right, in a way, the closer they are,
0:22:42 the more, yeah, yeah, yeah, exactly.
0:22:44 It’s not the way you would think.
0:22:45 I think at the end of the day,
0:22:48 basically what you have to avoid is divisiveness.
0:22:52 I see we have a no asshole rule in a company,
0:22:56 like most companies, I think, many companies choose.
0:22:58 I think it’s actually a valid opinion
0:23:00 to choose the opposite and say,
0:23:03 we are perfectly fine with brilliant jerks.
0:23:05 Like that’s actually a stable equilibrium
0:23:06 you can choose as well,
0:23:09 but like you have to be all in on that.
0:23:12 And you know, I think it’s actually useful
0:23:16 to reframe like active divisiveness
0:23:19 as like hovering a part of the same company
0:23:21 as a act of Nassau.
0:23:22 And I think that just,
0:23:25 that actually kind of transcends a lot of these conversations
0:23:29 because it gets away from the issue.
0:23:30 It’s never about the issue.
0:23:33 It’s actually about the behavior by people around the issue.
0:23:34 That is a problem in a company, right?
0:23:39 So just say, if you manage to make it a norm,
0:23:41 to say, this is just simply not what happens
0:23:46 on company funded internet servers such as Jack,
0:23:47 everything gets simpler.
0:23:51 – Take that outside, outside the virtual.
0:23:53 – Maybe one last point.
0:23:55 One book that really helped me figuring it out
0:23:59 was Thomas Sowell’s “Conflict of Visions,”
0:24:00 which is an underappreciated book.
0:24:02 I think it is very, very good.
0:24:04 Even if you read the first three chapters,
0:24:07 suddenly you realize why everyone’s actually kind of right,
0:24:12 except comes from a different sort of philosophical prior.
0:24:14 – Yeah, yeah, no question.
0:24:17 Actually, so let’s, you know,
0:24:22 you were into crypto very, very early on and why,
0:24:27 and then, you know, how do you think, you know,
0:24:29 given kind of how things have unfolded,
0:24:31 how the technology has developed,
0:24:34 how the kind of regulatory uncertainty has played,
0:24:36 like how are you thinking about it now,
0:24:39 kind of in the context of Shopify in the world?
0:24:41 – Yeah, a couple of angles to that.
0:24:45 First of all, I make a habit,
0:24:49 so growing up in a very small town in Germany,
0:24:53 not being, like I didn’t know anyone
0:24:54 who was into computers where I was,
0:24:59 like until late in my teens and I moved away, basically.
0:25:03 I think I built some skills around like outside observation,
0:25:06 like just sort of monitoring,
0:25:08 like I tried to figure out everything that’s going on
0:25:10 and, you know, like Silicon Valley,
0:25:12 then I figured out what that was,
0:25:16 but like I downloaded like the Linux kernel source code
0:25:19 and mailing this back in most of their use net,
0:25:21 like so I could study it all.
0:25:24 So I’m like, it’s weird because I’m clearly an insider now,
0:25:27 but like I’ve sort of lived and built skills
0:25:29 around outside observation.
0:25:31 So I make it a habit to like find the areas
0:25:36 where there’s the most passion and most higher stakes
0:25:39 and the most talent and people are just having a brainstorm
0:25:40 about what the future’s like.
0:25:44 I find that this is, to me, like watching television.
0:25:45 It’s so great.
0:25:50 Like so crypto is such a wellspring of brilliant insight
0:25:55 into like just like, you know, amazing technologies
0:25:57 and so on because of that I studied.
0:26:02 I find like consumer internet, like mobile apps in China
0:26:06 around 2010, 2015 so times like this, these pockets.
0:26:08 So that was sort of my angle.
0:26:11 I did make a Ruby implementation in 2012
0:26:14 of a blockchain paper of Satoshi’s paper.
0:26:15 So like it’s been on my mind
0:26:18 and just because I also love decentralization of,
0:26:20 I mean, just like giving people power,
0:26:23 but entrepreneurship, Shopify is just like give people.
0:26:25 – It’s actually the true power
0:26:27 of the people is decentralization.
0:26:27 – Exactly.
0:26:30 – And the false one is communism by the way.
0:26:33 – So there’s a hundreds of ways have that drawn me
0:26:36 into this which are my personal ones.
0:26:39 What I was hoping for is like something more practical
0:26:42 because my favorite thing is if people that I,
0:26:45 things that I find super interesting end up like coming
0:26:47 sort of in the vicinity of commerce
0:26:49 at which point I get to play with,
0:26:53 like I can do it for a birthday rather than in the evening.
0:26:58 So, so, you know, I love from technology perspective,
0:27:02 I love it from a sort of philosophical perspective.
0:27:06 You know, obviously at some point people like want
0:27:11 to use it for clearing, like buying products and so on.
0:27:13 We’ve supported this in any way we call it
0:27:15 for as long as you could have.
0:27:19 The demand has been low for using it for physical products
0:27:22 because you know, there’s (indistinct)
0:27:24 perspective, there’s some advantage credit cards,
0:27:27 but like it is crazy to me that at this point
0:27:29 that the internet just doesn’t have native currency.
0:27:30 We will get there.
0:27:33 I think people in this room hopefully will play
0:27:36 a leading role in this.
0:27:41 I, you know, again, from a personal philosophical
0:27:46 perspective, you know, I was on the internet in the 90s
0:27:51 and of course used Netscape and it came along mosaic
0:27:54 actually before that has just this incredible experience
0:27:57 of like in the future, like everyone will be able
0:28:02 to contribute and this is going to distribute power
0:28:05 in a way that is like unanticipated and not understood.
0:28:09 Shopify itself is a little bit like stuck
0:28:11 in this sort of 90s view of the internet.
0:28:14 Like it’s, I know it’s hosted software.
0:28:17 So it doesn’t conform directly to everyone has
0:28:19 their own server, but like you have your own website.
0:28:20 It’s like something you can own.
0:28:22 Like you can own a domain.
0:28:26 You can own, you know, the design of your system.
0:28:28 You can retrain the pens that is important.
0:28:32 And then I think this is like people have like my one
0:28:36 criticism to the previous sort of wave of crypto has been
0:28:40 that this idea of decentralization is too technical.
0:28:43 Like it’s like, I know people in the community here,
0:28:46 but like no one else cares about that definition of like
0:28:49 what we need to figure out like crypto needs friends
0:28:53 and the best angle to find friends is to find the fellow
0:28:58 travelers who are trying to move power to individuals again.
0:29:02 So that people can just like are not relying on like a
0:29:05 central marketplace for instance.
0:29:08 So somewhere that just has its own gatekeeper rules
0:29:10 or where the fees move to exactly what your margin is
0:29:14 and eradicate your ability to actually have a sustainable
0:29:15 business.
0:29:17 And so that’s really, really attractive to that.
0:29:20 – Yeah, so you, it almost needs a campaign,
0:29:22 like kick the ass of the tech overlords.
0:29:25 Like we’re going to give the power,
0:29:27 we’re going to give the technology back to the people
0:29:28 we’re going to take it from.
0:29:29 – Well, the, yeah.
0:29:32 – The Internet and Meta and Amazon and et cetera.
0:29:34 – Yeah, yeah.
0:29:37 I think what started with everyone start your own web
0:29:40 server, everyone write your own HTML.
0:29:44 It’s actually a life and well if you just change the framing
0:29:46 away from the exact technical implementation.
0:29:47 – Right.
0:29:50 – And so that’s a better, I think.
0:29:51 – Yeah.
0:29:51 – A larger group.
0:29:54 You can build a, you can build a movement around that.
0:29:56 You can’t build a movement around it.
0:30:01 It must be using the yakme.js, right?
0:30:05 Like it’s like, that’s not how you start a movement.
0:30:06 – Excellent point.
0:30:11 One last question then we’ll go to audience questions.
0:30:16 So you’ve been doing a lot of kind of experimentation
0:30:19 now with AI and then interestingly, you know,
0:30:22 kind of asked with crypto, there’s this huge push
0:30:23 to regulate AI.
0:30:28 You know, what do you think that means for kind of the
0:30:33 industry and the world and it’s a good idea, a bad idea.
0:30:37 – I think regulating AI at this point is a bad idea.
0:30:42 I think regulation around crypto, I might not be quite so.
0:30:46 Like, I think a regulatory clarity is what we should want.
0:30:49 I think you can get that by having no regulation
0:30:51 because it’s just a regulated field, but it should be
0:30:55 extremely permissive and for experimentation.
0:30:58 I think we found out some of the things that the financial
0:31:03 system had are things that are deeply desirable in crypto
0:31:04 as well, one by another.
0:31:06 I’m sure there’s like, everyone can take a different
0:31:09 position on this, but like that’s there.
0:31:14 In an AI case, man, like I just like, you gotta,
0:31:18 like you can regulate institutions or like ground rules
0:31:21 for markets, but regulating technology is like,
0:31:24 that seems like a crazy position to take.
0:31:27 Like suddenly floating points are like illegal.
0:31:29 That’s after you do too many of them.
0:31:34 – It is so weird, you know, I had a conversation
0:31:37 with a policymaker and I was like, so if this was nuclear,
0:31:40 you want regulate nukes, you’d regulate physics.
0:31:43 That’s your idea like physics is not illegal.
0:31:48 – Yeah, throw away those textbooks immediately, please.
0:31:53 – So again, I think maybe even some of the things
0:31:56 that people have identified as potential problems
0:31:59 in the future, I guess we could agree with,
0:32:04 but like again, the proposed solutions to this are crazy
0:32:06 because we get into these insane places where like,
0:32:09 cool, maybe analog computing is actually better
0:32:10 for training neural nets.
0:32:13 So that is not floating point operations anymore.
0:32:15 I was like, are you doing that now
0:32:18 because of regulations instead of because of pragmatism?
0:32:20 Or, you know, it just, I don’t know.
0:32:22 It just seems too early.
0:32:27 And it, I think that it’s a very, very bad idea right now
0:32:30 for any governments to set ceilings and rules
0:32:34 that are especially onerous for startups,
0:32:35 because I think all you’re gonna get out of that
0:32:39 is just like, you’re gonna get the place
0:32:42 where it’s gonna be all behind four-pay APIs
0:32:44 that are gonna be all controlled by the few.
0:32:49 I mean, that’s strictly better than it doesn’t exist, granted.
0:32:52 But like, clearly we need open source.
0:32:54 Look at everything that happened just now
0:32:58 after AI became available first
0:33:02 and then thanks to Metta and releasing amazing Lama models.
0:33:04 We know so much more about what Nuance do.
0:33:08 We know so much more about how to accelerate this.
0:33:11 You need to rescue these things sometimes
0:33:12 from the ivory towers
0:33:15 and you have to give them to the tinkerers.
0:33:18 And the people who remember how to memory align things
0:33:22 and what, you know, how to get the most out of a hardware.
0:33:23 Like that’s a different crowd than the people
0:33:27 who were writing Python neural topologies.
0:33:30 And it’s, we need everyone on this.
0:33:32 And then we’re gonna get the best outcome
0:33:34 because it is foundational for our future.
0:33:36 – Yeah. – Also, we need to figure out
0:33:38 how to make crypto closer aligned with AI
0:33:43 because it’s a natural way for how to pay for micropayments
0:33:45 and tokens and so on.
0:33:47 And this is really, really important now.
0:33:48 – Well, and solve so many.
0:33:50 I mean, it’s interesting.
0:33:52 Crypto solves so many AI problems
0:33:54 and AI solves so many crypto problems.
0:33:55 – Yeah. – They’re–
0:33:57 – These should be symbiotic
0:34:00 and just the sine waves of excitement of these two areas
0:34:02 just happen to be dissonant right now
0:34:04 and we need to make them harmonious partly
0:34:07 because of the bad big companies.
0:34:08 (laughs)
0:34:09 Sorry.
0:34:10 Well, that’s great.
0:34:14 So we are ready to take questions.
0:34:16 Any questions you may have?
0:34:18 Yes, sir, back there.
0:34:22 – If crypto plays out in exactly the same–
0:34:24 (mumbles)
0:34:30 – It’s like, I think very naturally the browser
0:34:32 will just like be able to move value around
0:34:34 and you will want to move value
0:34:36 in a way that can be addressed by the browser.
0:34:40 This is like, you can get pretty close to that
0:34:42 but you would have to squint
0:34:44 and you have to deal with like poor UX
0:34:46 and you have to deal with hard on ramps.
0:34:50 I think, I don’t think anything kind of magical happens
0:34:52 other than that people will have,
0:34:55 they need more utilitarian ability to like use money.
0:34:57 Again, I think if you run the counterfactual
0:34:59 and say open AI just for whatever reason
0:35:02 you couldn’t work for Stripe and just decided to,
0:35:05 I don’t know, like use some level two blockchain
0:35:10 and to pay for tokens as a initial MVP.
0:35:15 People would now move incredible amounts of money around
0:35:18 and build startups that might have these kind of ideas
0:35:19 that Sam talked about on stage,
0:35:24 about people sharing value of creating GPTs
0:35:27 and like all of the, you know, the blockchain,
0:35:30 like if you look at Open AI Dev Days,
0:35:31 especially the last 20 minutes
0:35:36 and you open at the same time VS Code to Ethereum contract,
0:35:39 you can basically write it out while he’s talking.
0:35:43 It’s like, so, we have this incredible infrastructure
0:35:45 but we can’t use it part because of fees,
0:35:46 part because of speeds,
0:35:49 because like you gotta solve the infrastructure problems
0:35:50 which I know we are doing.
0:35:52 Like this is people are building through this particular winter
0:35:53 in a way that is like,
0:35:57 seems utterly ideal from my perspective.
0:35:58 So that’s great.
0:36:03 There is, we need to bring killer use cases.
0:36:06 We need to find, like we need to have people
0:36:07 want to use this.
0:36:13 Unfortunately, there is like the focus on these kind of things,
0:36:16 especially once you need culture and social things,
0:36:18 is annoyingly not linear.
0:36:20 And that’s really trips everyone up.
0:36:21 It’s like, it’s nothing, nothing, nothing, nothing,
0:36:22 everything.
0:36:25 And because we have to get,
0:36:28 we have to like toil in obscurity
0:36:31 to build up the infrastructure
0:36:33 until it can support the kind of way
0:36:35 that people want to use it.
0:36:37 And then we need to catch lightning in a bottle.
0:36:39 Yes.
0:36:40 – Thanks, Sabi.
0:36:42 I had a question.
0:36:44 So Shopify works with a lot of merchants and brands
0:36:47 that might not be as tech savvy
0:36:49 or might not understand everything that’s going on
0:36:51 in like AI or crypto.
0:36:54 How does Shopify think about messaging to those merchants
0:36:56 and saying, hey, this is the future.
0:36:58 These are things that you should be using.
0:37:01 – Yeah, I think people actually hire us
0:37:04 for having them be ideal for whatever the internet is
0:37:07 and for whatever their buyers,
0:37:09 like they want to continue buying.
0:37:11 Like again, we start before mobile phones.
0:37:14 And so like just making sure that everyone’s store
0:37:16 is available to some mobile phones
0:37:18 back in those days, responsive, whatever,
0:37:22 just had it tremendously because their colleagues
0:37:24 who didn’t have that like just like started doing poor
0:37:29 and poor as money moved to the small form factors.
0:37:32 And they started out competing their peers in the industry.
0:37:35 So like as this is a super old and probably dumb example,
0:37:38 but like what I’m saying is like the demand has to come up.
0:37:42 Like the merchants offering like you can’t solve this problem
0:37:44 just simply by bringing the supply side up,
0:37:45 the demand side needs to go up.
0:37:50 The buyers should want to spend on the blockchain.
0:37:54 Now, that is the huge beauty.
0:37:58 You have 3% to work with here.
0:38:00 Like the credit card fees are high, right?
0:38:02 Like there’s a lot of interesting incentive systems
0:38:03 that you can do.
0:38:06 But what I found was the most amazing thing about sort of
0:38:08 when I came back to crypto,
0:38:10 paid more attention, smart contracts, if you’re,
0:38:14 is that, and I actually think even the practitioners
0:38:19 in the crypto space are under appreciating this
0:38:22 as a talent is there’s never been a field
0:38:25 of such applied widely distributed
0:38:28 and high quality thought applied game theory, right?
0:38:31 Like it’s the amount of the constructions
0:38:34 and the incentive systems that are being built
0:38:36 in this community are like tremendous.
0:38:41 Like the artists that have done really, really well
0:38:45 through NFTs like a such a good example.
0:38:48 This is the first, and now this is like,
0:38:52 man, this was the first business model of the arts
0:38:55 that we’ve gotten since patronage times, right?
0:38:57 Like there’s a huge breakthrough
0:38:59 and we’ll make a comeback.
0:39:02 This thought, this quality of thought has to be applied
0:39:06 I think to all sorts of other areas and in different ways.
0:39:09 And we can use the infrastructure that’s been built
0:39:12 to do something way better than cash back credit cards
0:39:15 like the first extra 3%.
0:39:17 Now it needs to also be meet the world where it is.
0:39:21 It requires some kind of escrow systems for charge back.
0:39:22 There is real distribution.
0:39:24 There needs to be like it,
0:39:26 not everything has to be trustless.
0:39:30 We should bring trust progress into the systems.
0:39:33 Like it’s called a wallet, which I think is a good name.
0:39:36 But in my wallet is, I don’t think I’ve cashed my wallet.
0:39:40 It’s all cards, it’s like licenses.
0:39:41 It’s like driver license and so on, right?
0:39:45 Like it’s attestations that I’m, who I say I am.
0:39:49 And bringing in trust brokers into the system
0:39:51 through the primitives we have now and saying,
0:39:54 yeah, that person is who we say they are.
0:39:56 Yes, that person can be shipped to.
0:40:01 Like if you write this thing on a package destination,
0:40:05 FedEx knows how to unclog this into an actual destination address.
0:40:08 We have to build this with the industry.
0:40:11 And then at some point,
0:40:15 we’re going to just have a much better environment for commerce.
0:40:18 And much like it’s not slightly better than using credit.
0:40:20 That it can actually be 10 times better.
0:40:22 People move around more since COVID, right?
0:40:25 Like how often have you gotten a package
0:40:27 to some other address that you’re no longer there?
0:40:30 Like this, it would be wonderful to make addresses pointers
0:40:32 that can be updated in the background and so on, so on, so on.
0:40:38 So like work with the actual infrastructure of the world
0:40:41 and modernize it on these new systems.
0:40:45 And so that all that want to accrue to people who want to take up one,
0:40:47 like who actually live in a real world
0:40:52 and not just like the hindsight synonymous, you know, abstraction.
0:40:55 I think that’s absolutely doable.
0:40:58 But man, do we have to get fees down?
0:41:00 And this is like it’s never going to happen
0:41:03 if we go back to a $50 transaction.
0:41:04 Yes.
0:41:07 We’ve got time for one more question.
0:41:10 Yes, right in the front row, since I saw you first.
0:41:13 So Shopify has a really successful partner program
0:41:16 where engineers are building themes and plugins and apps.
0:41:19 Could you talk a bit more about some of the challenges
0:41:20 starting that system up?
0:41:23 And then also, as you see more and more people speak the language
0:41:24 of NFTs and the EVM,
0:41:27 how that can evolve as we get more interoperability.
0:41:30 This is definitely, like, you’re sort of hinting at
0:41:34 how cool would it have been to have these pieces of infrastructure primitives
0:41:37 and we could have governed the entire partner system on a token economy.
0:41:38 Like, that’s totally true.
0:41:43 I think this is, I really hope someone is going to explore
0:41:46 how to build systems like the Shopify ecosystem
0:41:53 because there’s a lot of sharing and a lot of multi-volume transactions
0:41:56 which we are all doing with ledgers in the background.
0:41:59 Putting it up, like putting up any two-sided marketplace is tricky.
0:42:02 Here you have to overcome the chicken-necked problems.
0:42:05 It’s tricky, but if you manage to do it,
0:42:08 this is also really, really self-sustaining afterwards.
0:42:10 We have one of the greatest things.
0:42:14 We have about five different customers
0:42:19 with merchants who start on Shopify who now IPO’ed and are public companies.
0:42:23 We now have one who is building an app,
0:42:26 like on the Shopify app with their own private company,
0:42:32 Public Company Clavio is now in their Bay Bay successor, which is awesome.
0:42:36 Again, this is a wonderful thing of what you can do.
0:42:40 Companies are estimated to produce about six times more value
0:42:47 than they capture to the world, like enlightened self-interest novice.
0:42:53 The constrained vision in so-called terms is real.
0:42:57 And there are so many ways to build,
0:43:01 like so many businesses lend themselves to being turned into platforms later.
0:43:03 The problem is everyone tries to build a platform first,
0:43:05 and that’s really, really, really hard.
0:43:10 It’s much better to extract a platform out of a working piece of software
0:43:13 that people are already using, and so that helped us.
0:43:17 People were already trading themes as zip files,
0:43:19 and we were like, “Okay, well, we can help with that,”
0:43:23 and people are already building things for themselves
0:43:26 on the APIs for integrating their own ERP systems,
0:43:31 and then building this to the point where you could list this ERP connector
0:43:37 and build it like our SDKs made it so that it can be in so many stores.
0:43:41 But, de-forward, we hinted people and helped them into the right direction,
0:43:45 and all these kind of things ended up being incredibly valuable for the community.
0:43:50 I mean, it’s definitely one of the coolest aspects to build,
0:43:53 but again, at the end of the day, what all of this is,
0:43:59 is you build really good software that solves real problems very, very quickly,
0:44:03 and then you build a bit of a game-theoretical system on top,
0:44:12 in which people being selfish about wanting to build something accrue a huge amount of value
0:44:19 to the community, to Shopify, to the merchants, and so on.
0:44:26 I mean, I feel like the most sophisticated thought along those lines on Shopify
0:44:30 is a fairly amateur sort of baby’s first incentive system
0:44:37 compared to some of the things that were built on Ethereum and the blockchains
0:44:42 in terms of sophistication, except everything in the world of crypto
0:44:45 was so pointed at financialization and finances,
0:44:49 but I think people might have missed what could be done
0:44:52 if you think more about products rather than financial products,
0:44:55 and I think that’s hopefully going to be the next wave of crypto,
0:45:00 it’s going to be more products and solving real problems for people.
0:45:01 I’m that amazing insight.
0:45:04 I would love to please join me in thanking Koby.
0:45:05 Thank you.
0:45:08 [Music]
0:45:10 [Music]
0:45:12 [Music]
0:45:14 [Music]
0:45:16 [Music]
0:45:18 [Music]
0:45:20 [Music]
0:45:22 [Music]
0:45:24 [Music]
0:45:26 you
2
0:00:04 When Shopify was founded in 2006,
0:00:07 it was just a year prior to the launch of iPhone.
0:00:10 Needless to say, smartphones were a far cry
0:00:11 from the ubiquitous accessory
0:00:13 that now billions carry around.
0:00:17 And fast forward to today, nearly 20 years later,
0:00:19 where Shopify merchants can accept cryptocurrency,
0:00:21 create immersive shopping experiences
0:00:22 with augmented reality,
0:00:26 and even use AI to generate product descriptions.
0:00:29 So how does a nearly $100 billion company
0:00:30 stay ahead of these trends?
0:00:33 And what will shopping look like in the generations to come?
0:00:36 Plus, why would A16Z never have invested in Shopify
0:00:38 in its early day?
0:00:40 And why has the CEO of this company
0:00:42 chosen to continue writing code?
0:00:44 These are just some of the questions covered
0:00:45 in today’s episode,
0:00:48 the conversation the Shopify co-founder and CEO
0:00:52 to bias Luki and A16Z co-founder, Ben Horowitz.
0:00:54 This episode was also originally published
0:00:57 on our sister podcast, Web3 with A16Z.
0:01:00 So if you are excited about the next generation
0:01:03 of the internet, find Web3 with A16Z
0:01:04 wherever you get your podcasts.
0:01:06 All right, let’s get started.
0:01:14 – Welcome to Web3 with A16Z,
0:01:17 a show about building the next generation of the internet
0:01:20 from the team at A16Z Crypto.
0:01:22 Today’s episode features Tobias Lutke,
0:01:26 CEO and co-founder of the e-commerce platform Shopify.
0:01:29 Speaking with A16Z co-founder, Ben Horowitz,
0:01:32 at our second annual founder summit in November,
0:01:35 they discuss what it takes to build a breakout startup
0:01:39 in a crowded category, the changing face of retail,
0:01:41 how to affect change in the workplace,
0:01:44 and how to handle individual emotions and corporate culture,
0:01:47 including dealing with calls for activism,
0:01:50 as well as the value of embracing negativity.
0:01:52 They also touch on the moral imperative
0:01:54 behind creating quality software,
0:01:58 the symbiosis between AI and crypto, and more.
0:02:01 As a reminder, none of the following should be taken
0:02:04 as business, legal, tax, or investment advice.
0:02:07 Please see A16Z.com/disclosures
0:02:08 for more important information,
0:02:11 including a link to a list of our investments.
0:02:19 – All right, so for those of you who don’t know Toby,
0:02:20 which you probably should know Toby,
0:02:22 you can look him up as well,
0:02:27 but he started a snowboarding company
0:02:29 in 2004 called Snowdevil,
0:02:33 and he parlayed that into Shopify,
0:02:37 which now serves millions of merchants
0:02:40 in 175 countries around the world.
0:02:42 How many countries are there, like in total?
0:02:44 – I think that’s most of them, there’s 200?
0:02:45 – Yeah, 200, yeah, but–
0:02:48 – There’s changes around the margins.
0:02:51 – You’ll get them all at some point.
0:02:54 And then he was also interesting for this crowd.
0:02:58 He was one of the first adopters of Coinbase Commerce.
0:03:02 So Toby’s been kind of big on crypto for a while now.
0:03:04 But let’s get started.
0:03:08 So you started as a snowboarding company,
0:03:12 and I always find that funny because of all the CEOs
0:03:16 that I know you’re probably the least snowboard kickback,
0:03:19 like let’s hit the powder dude.
0:03:21 So how did that happen?
0:03:23 Like why did you start out that way?
0:03:28 – Yeah, you know, when you talk to governments,
0:03:32 they always, Shopify is doing a lot about entrepreneurship,
0:03:35 which I was a lot of entrepreneurship,
0:03:38 and it’s such a brilliant institution.
0:03:39 I know everyone here is a card-carrying member
0:03:42 of the entrepreneurship is good club,
0:03:45 but politicians usually, they’re sort of vague on it,
0:03:50 but they say like they want to pass government programs
0:03:52 and policies to cause more entrepreneurship,
0:03:55 which I always think is the wrong way around.
0:04:00 But there’s actually, so the funny story is that,
0:04:03 so I was in Canada at this point, I came from Germany,
0:04:06 living in Canada 2004, three, four,
0:04:10 and I found out I cannot actually get a job.
0:04:13 I was trying to be hired as a computer programmer locally,
0:04:15 but I found out I did not have a work permit.
0:04:16 I didn’t know what a work permit was.
0:04:20 So turns out that you are in Canada,
0:04:22 allowed to start a company even without a work permit.
0:04:23 So there’s one government program
0:04:27 that actually has led to what company is being created.
0:04:29 – Well, that’s like the best government program ever.
0:04:32 Just like the payoff on Shopify paid for itself
0:04:33 a million times.
0:04:35 – And it turns out Canada is very cold.
0:04:37 So I was doing snowboarding, so I knew something about it,
0:04:40 and I started an online store for snowboarding.
0:04:42 It’s kind of a pretty simple story like this.
0:04:44 – That’s amazing.
0:04:47 So basically the key to your whole career
0:04:49 was not being able to work.
0:04:49 – Yes.
0:04:51 (laughing)
0:04:53 In fact, you know, what was in my mind too is like,
0:04:56 I was like the solution with programming at this time.
0:04:57 It was like sort of a Java word,
0:04:59 kind of was kind of very oppressive.
0:05:04 And I wanted to reclaim it as a hobby.
0:05:06 So I was like, cool, if I sell snowboards online,
0:05:08 I can make money while snowboarding,
0:05:09 and I get to do more fun stuff.
0:05:14 And it’s been extremely unsuccessful as a strategy.
0:05:17 I basically never had time to snowboard again
0:05:17 after I started.
0:05:20 (laughing)
0:05:22 – The end of your snowboarding career
0:05:24 was the start of your snowboarding business.
0:05:25 That’s pretty remarkable.
0:05:27 – Maybe I would project more back at this point
0:05:29 if things would have gone in the counterfactual way.
0:05:31 – Yeah, now the other thing that really strikes me
0:05:34 about Shopify is that, you know,
0:05:36 and I hate to say this as a venture capitalist
0:05:38 who some people think is a smart person,
0:05:42 but we probably would have never invested in Shopify
0:05:43 when you started it
0:05:47 because it was probably the most over-competed
0:05:48 category in the world.
0:05:52 Like everybody had some kind of e-commerce platform.
0:05:56 I mean, there was just so, so many of them,
0:06:00 Magento, this, that, the other.
0:06:05 So did you go, oh, no, no, no, like,
0:06:07 I’m gonna build this thing.
0:06:09 It’s gonna dominate everything.
0:06:10 Then I’m gonna have a network effect
0:06:12 and I’m gonna build, you know,
0:06:14 like one of the biggest things that anybody’s ever seen.
0:06:19 Like, how did you like get all the way to where you are?
0:06:20 Like, what was that path?
0:06:24 – Yeah, no, there’s no plan like this.
0:06:29 Just like, I think, for every entrepreneur,
0:06:31 it’s a useful thing to think about
0:06:33 what is your energy source, right?
0:06:35 Like what is actually driving you?
0:06:37 And sometimes it’s greed.
0:06:40 And we, you know, like the sort of negative emotions
0:06:42 channeled into building actually
0:06:44 the most powerful energy sources.
0:06:47 I had this, you were right.
0:06:48 Like, I mean, Netscape, IPO,
0:06:52 but like, you have to explain every platform
0:06:54 in terms of an old platform and, you know,
0:06:57 online hosted CS catalog with buy buttons
0:07:00 was like kind of a way you would mark
0:07:03 to describe even what the internet would be used for, right?
0:07:04 Like, so e-commerce was like right there
0:07:07 from the beginning, I thought there would be lots of software
0:07:09 that I could use and I didn’t find anything.
0:07:12 I was actually really upset with the quality of the software.
0:07:15 And so like, I channel that into building something
0:07:17 and that was my energy source.
0:07:19 And I just love the building.
0:07:20 – That’s actually, by the way,
0:07:23 one of the best entrepreneurial ideas there is,
0:07:25 which is this stuff sucks so bad.
0:07:27 Yeah, I’m so mad about having to use it
0:07:29 that I’m going to build my own.
0:07:32 – Totally, I honestly, it fuels me to do it this day.
0:07:34 I’m just, I still think it’s ridiculous
0:07:37 what people are subjected to in terms of quality of software.
0:07:39 It’s like, oh, it’s weird.
0:07:41 I probably did some like neurological,
0:07:43 like kind of rearrangement there,
0:07:44 but like to me, this is almost a moral issue
0:07:48 of like when people like confused with software
0:07:50 and they think about themselves as like being inadequate.
0:07:53 It’s like, how dare software make humans feel that way?
0:07:54 Right, like it’s just not okay.
0:07:57 So you want to make it a more approachable,
0:07:59 but like, I think the thing that people might have missed
0:08:02 in e-commerce or like, it’s not that I had a great insight.
0:08:05 I just sort of organically got there is bad.
0:08:06 Yes, there was a lot of e-commerce software,
0:08:09 but it was all built for already rich retailers
0:08:11 who needed to move online.
0:08:13 Then that was a phrase people used.
0:08:17 And there was no one started software
0:08:20 for starting more entrepreneurial process online.
0:08:23 Like actually the digital native, what we call it now also,
0:08:26 like if you like that term,
0:08:29 because there’s a totally different set of requirements
0:08:30 for the people who are just starting out.
0:08:32 And so that’s what we did.
0:08:33 So this is the thing, right?
0:08:38 Like it’s like, when you talk,
0:08:40 that’s like, I don’t know, 60 words per minute or something,
0:08:42 and you type fast, it’s like twice that.
0:08:45 But like, if you can do customer consultation
0:08:49 in your own brain, the brain runs at like terabytes per second
0:08:50 by bandwidth, right?
0:08:53 Like so, if you know the use case,
0:08:56 if you’re building for yourself or your past experience
0:08:58 or some kind of, if you have a really good model
0:09:01 of your customers or use cases in your mind,
0:09:03 you can do customer development at high bandwidth
0:09:06 with low latency and you can just build something
0:09:08 that you know needs to exist.
0:09:10 – Yeah, no, that’s incredible.
0:09:13 So one of the things, you’re also kind of been
0:09:19 a kind of a special kind of, or a different kind of CEO
0:09:24 than some of the ones that people read about
0:09:27 such as Adela or Bob Iger and that you have a reputation
0:09:31 for being extremely hands-on, including writing code.
0:09:34 You still have the fury that started the company.
0:09:38 How do you think about that in terms of,
0:09:40 you’re not kind of paying attention
0:09:42 to all parts of the company equally.
0:09:46 You’re clearly doing certain things and not others.
0:09:49 How do you think about that as an effective
0:09:51 kind of management style?
0:09:56 – Yeah, I mean, I imagine the people in my company
0:09:58 would disagree with your assessment
0:10:00 about some effective management style.
0:10:02 So well, it’s worked pretty well.
0:10:07 It’s working, but it certainly doesn’t proxy
0:10:11 to what people sort of, like it’s not a CEO job
0:10:14 out of central casting that I’m trying to perform, right?
0:10:17 I tried that on and it didn’t work for me at all.
0:10:20 I actually find, like I value being in all the details
0:10:23 and I ask it from all the people who are around me
0:10:25 and report to me, like almost everyone,
0:10:29 especially after COVID, I turned over everyone,
0:10:33 but one of my executives during the first 10 months of COVID.
0:10:35 And now almost everyone reports to me,
0:10:38 it’s actually a AX founder, maybe someone be acquired
0:10:41 or someone who started a company before coming to Shopify.
0:10:45 So because there’s this, I just want,
0:10:48 I find it actually really inefficient,
0:10:50 not understanding the details of what we’re doing, right?
0:10:51 Because then–
0:10:52 – Well, then you’re guessing.
0:10:53 – Then you’re guessing, right?
0:10:57 And you’re like, well, when something goes wrong,
0:10:59 now you not only need to fix it,
0:11:02 you actually have the first cram for,
0:11:04 like you have to learn like three months
0:11:05 before you can actually do a fix.
0:11:06 If you understand what’s going on
0:11:09 or what ought to be existing, it’s much faster.
0:11:13 And you want to train people on the company mission, right?
0:11:16 Like because there’s a thing that you want to exist,
0:11:18 like even if you don’t know exactly what it is,
0:11:21 like you know which direction you want people to go, right?
0:11:24 And getting everyone aligned to go in this direction
0:11:27 is only possible if you can paint picture
0:11:30 of like how does this area actually sum together
0:11:33 all the way to helping the mission overall.
0:11:38 And so, and this is sort of a sort of unique thing
0:11:39 about our times right now.
0:11:41 It’s like the infrastructure keeps changing,
0:11:44 like here’s crypto, here’s AI, here’s,
0:11:46 I mean, I started again almost,
0:11:48 like next year is 20 years ago, right?
0:11:51 So there was no mobile phone, like that offer quality
0:11:53 that we have now.
0:11:54 – Right, that’s right.
0:11:55 You started before the mobile phone.
0:11:58 – Right, so it’s like, that was one of the things
0:11:59 that just sort of happened.
0:12:03 And, you know, SaaS software was new at this time.
0:12:05 So, you know, the platform’s capabilities change
0:12:08 and you want to adopt them to like your mission
0:12:10 to what is now possible.
0:12:13 And so you have to be able to understand, well,
0:12:15 but in a kind of like ideally in your head,
0:12:21 in the world where, I take a very long-term view
0:12:24 and my biggest fear that I have
0:12:27 is that Shopify winds up being a fantastic solution
0:12:30 to a problem that people no longer have, right?
0:12:32 Or no longer solve this way.
0:12:33 So I’m trying to like need to keep it current
0:12:35 and being able to run the counterfactual about
0:12:39 this is not possible, this is what we’re trying to accomplish.
0:12:41 But if this would have been possible from the beginning,
0:12:43 we would have built this entirely different thing.
0:12:45 So now our job’s to get from here to there
0:12:48 because, again, we want to keep it current and idea.
0:12:49 – Right, you know, that’s such a great insight
0:12:53 because they think that, you know, when I work with CEOs
0:12:55 and you think about the job,
0:12:58 so much of the job is making high quality decisions
0:13:00 and setting the direction of the company.
0:13:03 And in some respects, neither feels like work.
0:13:07 You know, because you’re working on something else,
0:13:10 like how does AI work, what’s possible,
0:13:13 in order to make decisions and set the direction.
0:13:16 But what you’re doing seems actually removed
0:13:20 from anything that the company needs to do right now.
0:13:22 And I think so many people kind of neglect
0:13:24 that kind of thing for that reason.
0:13:27 The other thing that I always find is,
0:13:29 at any given point in time,
0:13:32 some things in the company are very important
0:13:33 and very high leverage.
0:13:35 And some things are just not important
0:13:37 for you to put your time into.
0:13:39 And people often get lost in that.
0:13:43 They think, oh, I’ve got to talk to this person
0:13:46 and then that person and then the other person and so forth.
0:13:48 And it’s like, no, no, you don’t.
0:13:49 You’ve got to make good decisions.
0:13:50 You have to set the direction.
0:13:52 And if you don’t do that, nobody’s doing that.
0:13:54 And then the company’s going to fail.
0:13:57 So that’s such an amazing way that you’ve gone about that.
0:13:59 – If you’re in a small team,
0:14:01 like for a new thing that your company needs to do
0:14:04 and you’re in a small team of people who are like,
0:14:08 know or like, figure out the shape of it early,
0:14:09 if it’s actually as important,
0:14:11 it will itself have a gravity
0:14:14 to change the way the company thinks about quality of software
0:14:16 shipping, the way the architect thinks.
0:14:22 And it’s also after this comes up as a new thing,
0:14:26 you then have the ability to like a type of legitimacy.
0:14:27 As a founder,
0:14:30 you have a huge amount of legitimacy in a company already.
0:14:33 But like if you also are there with the new project,
0:14:35 that’s important, that needs to be a resource and say,
0:14:38 this is what we’ll do.
0:14:40 You don’t have to have all these conversations.
0:14:43 People love clarity of like, this is what we do.
0:14:45 Like it has a gravity all to itself.
0:14:48 That’s unlike anything you can accomplish
0:14:50 by one-on-one consultation,
0:14:53 trying to talk everyone through the entire decision matrix.
0:14:56 And so what is Toby doing over there?
0:14:57 – Oh, that must be important.
0:14:59 – Yeah, that’s exactly right.
0:15:01 We’re doing like a job site click now.
0:15:04 And like, then I don’t have to figure out like,
0:15:06 yes, there’s like 15 teams that are already doing AI stuff
0:15:09 and they ought to be in some kind of constellation.
0:15:12 But now this AI assistant thing is going to run through
0:15:16 and the company, everyone wants to be a part of it.
0:15:18 And this is like, all these problems kind of disappear.
0:15:19 It’s like, it’s maybe not the most effective-
0:15:20 – Like a magic trick.
0:15:21 – Yeah, yeah, yeah.
0:15:23 Yeah, actually, I think by the way,
0:15:25 like everyone in this,
0:15:28 like everyone here has clearly read your books.
0:15:32 And if you haven’t done do that, like I think-
0:15:33 – Thank you.
0:15:35 – The best books on leadership, honestly.
0:15:38 Thank you so much for sharing all this.
0:15:40 I think in some of what you described,
0:15:43 but there’s only very few ways to do a change management.
0:15:45 One way of doing it is to something super, super,
0:15:49 super surprising and then explain everyone why you did it.
0:15:51 It’s the fastest way to get a large company.
0:15:53 In my shop, there’s about 10,000 people,
0:15:55 much, much, much faster way to do a change management
0:15:56 than hand-to-hand combat.
0:15:58 – It’s like shock therapy.
0:16:00 It’s like, what the fuck was that?
0:16:02 Oh, well, this was that, that was okay, got it.
0:16:06 Very, very, very high retention on that.
0:16:07 So let’s talk about, you know,
0:16:10 one of the things that has come up, you know,
0:16:13 a lot over the years that you had probably, you know,
0:16:16 maybe the best statement on, I thought,
0:16:20 is politics, you know, politics, activism,
0:16:22 how that’s kind of come into companies
0:16:25 has taken over a lot of the activity of companies.
0:16:29 And how do you, I guess, how do you think about that?
0:16:33 And then why’d you say what you said?
0:16:35 And then how’s that going?
0:16:39 And then how are you able to sustain it particularly
0:16:42 as things get even hotter now with, you know,
0:16:44 what’s going on in the Middle East and so forth?
0:16:47 You know, it’s gotten even more intense than it was,
0:16:49 you know, amazingly in 2020.
0:16:51 – That’s a complex topic.
0:16:56 I can, here’s, I mean, there’s a lot of conflicting ideas
0:16:58 about what people want companies to be.
0:17:02 And, you know, I think, especially in 2010
0:17:06 and maybe now again, a lot of people would like companies
0:17:08 to play the role of some kind of phantom limb
0:17:13 of what I think the whole left over by the stories
0:17:16 that people are surrounded with, like maybe with governments
0:17:17 or like, like, and so on.
0:17:20 I suddenly, people want, like I have gotten tons
0:17:23 and tons of petitions for Shopify to take stances on issues.
0:17:27 And I think everyone’s feeling the same thing.
0:17:30 This was weird to me, but like initially,
0:17:32 but like, I found myself generally agreeing
0:17:34 with what people identify as the problems
0:17:36 when this happens.
0:17:38 But I found myself, they really, if ever,
0:17:41 really agreeing with the proposed solutions
0:17:41 to those problems.
0:17:44 And I think it’s kind of important to say like,
0:17:46 hey, unless we are all aligned on this,
0:17:48 we’re probably not doing this.
0:17:51 Because, you know, at the end of the day,
0:17:53 like companies ought to be thought of
0:17:54 as a bit of a simpler thing.
0:17:57 It’s, you know, real love entrepreneurship.
0:17:59 Entrepreneurship is, you know,
0:18:01 liberal values kind of institution of, you know,
0:18:05 you’re like reaching for independence,
0:18:07 reaching for like, to better yourself,
0:18:09 like building something like enlightened,
0:18:10 the ideas of enlightened self-interest.
0:18:13 There is a bit of a political coding in this,
0:18:14 people at least think so.
0:18:16 And someone else might take a totally different opinion
0:18:17 about all these kinds of things.
0:18:18 So that’s cool.
0:18:20 The cool thing is companies themselves explorations
0:18:22 into a set of ideas.
0:18:24 Sometimes encoded in a mission sometimes by the founders.
0:18:27 And if we make every company the same,
0:18:29 then we lose that.
0:18:31 That’s actually like a weird form of diversity itself,
0:18:32 if you think about it.
0:18:36 So anyway, I went ahead and just like said,
0:18:40 like look, we cause more entrepreneurship to exist.
0:18:43 And sometimes there are social causes
0:18:45 that are aligned with this.
0:18:48 And then maybe we become active because of our mission.
0:18:50 But outside of that, you know,
0:18:53 everyone makes money and you can do whatever you want
0:18:54 in your after hours.
0:18:58 And that’s just a simpler way of thinking about the business.
0:19:02 Of course, that’s not universally loved,
0:19:05 but like after being there, they clear about it
0:19:05 and explaining it.
0:19:09 They explain, you know, we give everyone when we hire people
0:19:11 a letter of like, here are the reasons,
0:19:13 like before you sign.
0:19:15 Here’s the reason why you might not want to work
0:19:18 for Shopify because like here’s a set of ideas
0:19:19 that you disagree with.
0:19:22 And then that clarity is wonderful.
0:19:24 We keep everything so nebulous.
0:19:28 I think the world of marketing and PR
0:19:31 has just done a number on us all to be so non-committal
0:19:32 to everything.
0:19:33 – Yeah.
0:19:35 – And the problem is if you’re non-committal,
0:19:37 you leave a lot of vacuum.
0:19:38 – Right.
0:19:40 – And that vacuum is going to be filled by someone.
0:19:43 And it’s usually the people who want to fill it
0:19:46 with bad faith takes, who fill it.
0:19:50 So if you just clear about it, it works sometimes,
0:19:54 like this is something I might get into hot water with,
0:19:57 but like sometimes it’s actually good to heighten
0:19:58 the misalignment for the kind of people
0:20:00 that you don’t think should be working in a company.
0:20:04 Like, hey, we work hard and it’s kind of an-
0:20:06 – Yeah, we have poor work-life balance.
0:20:07 – Yeah, yeah.
0:20:08 – Sorry.
0:20:09 – That’s a good-
0:20:11 – You are free to work elsewhere.
0:20:15 – If it’s true, put it up, like light it up on a sign
0:20:17 because otherwise you’re going to end up
0:20:20 with the weirdest divisiveness in the company
0:20:22 because of basically false advertising.
0:20:25 So I think it’s better to just be clear.
0:20:29 So I mean, what is it like?
0:20:31 There’s a lot goes into this.
0:20:33 Like again, culture is unstable.
0:20:36 We know from the internet that every community
0:20:40 that anyone likes Reddit, like Hacker News,
0:20:42 if anyone actually likes this page,
0:20:47 it’s a product of unbelievably dedicated moderators
0:20:51 who are keeping this discourse well.
0:20:54 I think that’s an idea that I think companies
0:20:55 have started adopting too.
0:20:57 Like once you’re a couple of thousand people,
0:20:59 your Slack channel is the internet basically
0:21:02 and you need to apply similar rules.
0:21:04 You need to be able to take people to a sign saying,
0:21:05 hey, here’s the thing you’re causing.
0:21:07 When you’re sending a message to a channel
0:21:09 with 3,000 people in it,
0:21:11 that is basically the same as you spending a day
0:21:15 typing individual text message to any person in the channel.
0:21:16 And would you do that?
0:21:18 If no, probably don’t put in this kind of stuff.
0:21:22 Yeah, that’s really such a great insight that you had that.
0:21:24 Hey, I actually kind of agree with the intentions
0:21:26 that people have on most of this,
0:21:30 just not with the ideas on how to achieve those intentions
0:21:35 because that’s where politics gets really dangerous
0:21:38 in a company and I have a funny story about this.
0:21:41 So my great uncle Harold, my grandmother didn’t speak
0:21:43 for 20 years, brother and sister.
0:21:45 And I asked my father a few years ago,
0:21:46 I said, like, what happened there?
0:21:47 Why didn’t they speak?
0:21:50 He’s like, oh, well, ’cause your great uncle Harold
0:21:51 was a Trotskyist.
0:21:53 And I was like, well, what was grandma?
0:21:55 She was a Stalinist.
0:21:58 They’re two slightly different kinds of communists
0:21:59 and they didn’t speak for 20 years.
0:22:01 And that’s kind of what happens inside a company.
0:22:04 It’s not like you’re a good person, you’re a bad person.
0:22:06 This is a moral issue.
0:22:10 It’s like, well, yeah, we all want way fewer people
0:22:12 to get killed in this current conflict,
0:22:14 but how you achieve that,
0:22:16 that gets very complicated very fast.
0:22:19 And so if you let people attack each other,
0:22:22 shut down conversations, intimidate each other
0:22:26 inside your company around that, it’s just all bad.
0:22:27 – Yeah, but fighting gets the fiercest
0:22:29 when mistakes are the lowest.
0:22:31 You know, like when people almost agree,
0:22:34 when they become mortal enemies.
0:22:35 – Right.
0:22:36 – It’s a very funny effect.
0:22:38 I mean, you can.
0:22:39 – Right, in a way, the closer they are,
0:22:42 the more, yeah, yeah, yeah, exactly.
0:22:44 It’s not the way you would think.
0:22:45 I think at the end of the day,
0:22:48 basically what you have to avoid is divisiveness.
0:22:52 I see we have a no asshole rule in a company,
0:22:56 like most companies, I think, many companies choose.
0:22:58 I think it’s actually a valid opinion
0:23:00 to choose the opposite and say,
0:23:03 we are perfectly fine with brilliant jerks.
0:23:05 Like that’s actually a stable equilibrium
0:23:06 you can choose as well,
0:23:09 but like you have to be all in on that.
0:23:12 And you know, I think it’s actually useful
0:23:16 to reframe like active divisiveness
0:23:19 as like hovering a part of the same company
0:23:21 as a act of Nassau.
0:23:22 And I think that just,
0:23:25 that actually kind of transcends a lot of these conversations
0:23:29 because it gets away from the issue.
0:23:30 It’s never about the issue.
0:23:33 It’s actually about the behavior by people around the issue.
0:23:34 That is a problem in a company, right?
0:23:39 So just say, if you manage to make it a norm,
0:23:41 to say, this is just simply not what happens
0:23:46 on company funded internet servers such as Jack,
0:23:47 everything gets simpler.
0:23:51 – Take that outside, outside the virtual.
0:23:53 – Maybe one last point.
0:23:55 One book that really helped me figuring it out
0:23:59 was Thomas Sowell’s “Conflict of Visions,”
0:24:00 which is an underappreciated book.
0:24:02 I think it is very, very good.
0:24:04 Even if you read the first three chapters,
0:24:07 suddenly you realize why everyone’s actually kind of right,
0:24:12 except comes from a different sort of philosophical prior.
0:24:14 – Yeah, yeah, no question.
0:24:17 Actually, so let’s, you know,
0:24:22 you were into crypto very, very early on and why,
0:24:27 and then, you know, how do you think, you know,
0:24:29 given kind of how things have unfolded,
0:24:31 how the technology has developed,
0:24:34 how the kind of regulatory uncertainty has played,
0:24:36 like how are you thinking about it now,
0:24:39 kind of in the context of Shopify in the world?
0:24:41 – Yeah, a couple of angles to that.
0:24:45 First of all, I make a habit,
0:24:49 so growing up in a very small town in Germany,
0:24:53 not being, like I didn’t know anyone
0:24:54 who was into computers where I was,
0:24:59 like until late in my teens and I moved away, basically.
0:25:03 I think I built some skills around like outside observation,
0:25:06 like just sort of monitoring,
0:25:08 like I tried to figure out everything that’s going on
0:25:10 and, you know, like Silicon Valley,
0:25:12 then I figured out what that was,
0:25:16 but like I downloaded like the Linux kernel source code
0:25:19 and mailing this back in most of their use net,
0:25:21 like so I could study it all.
0:25:24 So I’m like, it’s weird because I’m clearly an insider now,
0:25:27 but like I’ve sort of lived and built skills
0:25:29 around outside observation.
0:25:31 So I make it a habit to like find the areas
0:25:36 where there’s the most passion and most higher stakes
0:25:39 and the most talent and people are just having a brainstorm
0:25:40 about what the future’s like.
0:25:44 I find that this is, to me, like watching television.
0:25:45 It’s so great.
0:25:50 Like so crypto is such a wellspring of brilliant insight
0:25:55 into like just like, you know, amazing technologies
0:25:57 and so on because of that I studied.
0:26:02 I find like consumer internet, like mobile apps in China
0:26:06 around 2010, 2015 so times like this, these pockets.
0:26:08 So that was sort of my angle.
0:26:11 I did make a Ruby implementation in 2012
0:26:14 of a blockchain paper of Satoshi’s paper.
0:26:15 So like it’s been on my mind
0:26:18 and just because I also love decentralization of,
0:26:20 I mean, just like giving people power,
0:26:23 but entrepreneurship, Shopify is just like give people.
0:26:25 – It’s actually the true power
0:26:27 of the people is decentralization.
0:26:27 – Exactly.
0:26:30 – And the false one is communism by the way.
0:26:33 – So there’s a hundreds of ways have that drawn me
0:26:36 into this which are my personal ones.
0:26:39 What I was hoping for is like something more practical
0:26:42 because my favorite thing is if people that I,
0:26:45 things that I find super interesting end up like coming
0:26:47 sort of in the vicinity of commerce
0:26:49 at which point I get to play with,
0:26:53 like I can do it for a birthday rather than in the evening.
0:26:58 So, so, you know, I love from technology perspective,
0:27:02 I love it from a sort of philosophical perspective.
0:27:06 You know, obviously at some point people like want
0:27:11 to use it for clearing, like buying products and so on.
0:27:13 We’ve supported this in any way we call it
0:27:15 for as long as you could have.
0:27:19 The demand has been low for using it for physical products
0:27:22 because you know, there’s (indistinct)
0:27:24 perspective, there’s some advantage credit cards,
0:27:27 but like it is crazy to me that at this point
0:27:29 that the internet just doesn’t have native currency.
0:27:30 We will get there.
0:27:33 I think people in this room hopefully will play
0:27:36 a leading role in this.
0:27:41 I, you know, again, from a personal philosophical
0:27:46 perspective, you know, I was on the internet in the 90s
0:27:51 and of course used Netscape and it came along mosaic
0:27:54 actually before that has just this incredible experience
0:27:57 of like in the future, like everyone will be able
0:28:02 to contribute and this is going to distribute power
0:28:05 in a way that is like unanticipated and not understood.
0:28:09 Shopify itself is a little bit like stuck
0:28:11 in this sort of 90s view of the internet.
0:28:14 Like it’s, I know it’s hosted software.
0:28:17 So it doesn’t conform directly to everyone has
0:28:19 their own server, but like you have your own website.
0:28:20 It’s like something you can own.
0:28:22 Like you can own a domain.
0:28:26 You can own, you know, the design of your system.
0:28:28 You can retrain the pens that is important.
0:28:32 And then I think this is like people have like my one
0:28:36 criticism to the previous sort of wave of crypto has been
0:28:40 that this idea of decentralization is too technical.
0:28:43 Like it’s like, I know people in the community here,
0:28:46 but like no one else cares about that definition of like
0:28:49 what we need to figure out like crypto needs friends
0:28:53 and the best angle to find friends is to find the fellow
0:28:58 travelers who are trying to move power to individuals again.
0:29:02 So that people can just like are not relying on like a
0:29:05 central marketplace for instance.
0:29:08 So somewhere that just has its own gatekeeper rules
0:29:10 or where the fees move to exactly what your margin is
0:29:14 and eradicate your ability to actually have a sustainable
0:29:15 business.
0:29:17 And so that’s really, really attractive to that.
0:29:20 – Yeah, so you, it almost needs a campaign,
0:29:22 like kick the ass of the tech overlords.
0:29:25 Like we’re going to give the power,
0:29:27 we’re going to give the technology back to the people
0:29:28 we’re going to take it from.
0:29:29 – Well, the, yeah.
0:29:32 – The Internet and Meta and Amazon and et cetera.
0:29:34 – Yeah, yeah.
0:29:37 I think what started with everyone start your own web
0:29:40 server, everyone write your own HTML.
0:29:44 It’s actually a life and well if you just change the framing
0:29:46 away from the exact technical implementation.
0:29:47 – Right.
0:29:50 – And so that’s a better, I think.
0:29:51 – Yeah.
0:29:51 – A larger group.
0:29:54 You can build a, you can build a movement around that.
0:29:56 You can’t build a movement around it.
0:30:01 It must be using the yakme.js, right?
0:30:05 Like it’s like, that’s not how you start a movement.
0:30:06 – Excellent point.
0:30:11 One last question then we’ll go to audience questions.
0:30:16 So you’ve been doing a lot of kind of experimentation
0:30:19 now with AI and then interestingly, you know,
0:30:22 kind of asked with crypto, there’s this huge push
0:30:23 to regulate AI.
0:30:28 You know, what do you think that means for kind of the
0:30:33 industry and the world and it’s a good idea, a bad idea.
0:30:37 – I think regulating AI at this point is a bad idea.
0:30:42 I think regulation around crypto, I might not be quite so.
0:30:46 Like, I think a regulatory clarity is what we should want.
0:30:49 I think you can get that by having no regulation
0:30:51 because it’s just a regulated field, but it should be
0:30:55 extremely permissive and for experimentation.
0:30:58 I think we found out some of the things that the financial
0:31:03 system had are things that are deeply desirable in crypto
0:31:04 as well, one by another.
0:31:06 I’m sure there’s like, everyone can take a different
0:31:09 position on this, but like that’s there.
0:31:14 In an AI case, man, like I just like, you gotta,
0:31:18 like you can regulate institutions or like ground rules
0:31:21 for markets, but regulating technology is like,
0:31:24 that seems like a crazy position to take.
0:31:27 Like suddenly floating points are like illegal.
0:31:29 That’s after you do too many of them.
0:31:34 – It is so weird, you know, I had a conversation
0:31:37 with a policymaker and I was like, so if this was nuclear,
0:31:40 you want regulate nukes, you’d regulate physics.
0:31:43 That’s your idea like physics is not illegal.
0:31:48 – Yeah, throw away those textbooks immediately, please.
0:31:53 – So again, I think maybe even some of the things
0:31:56 that people have identified as potential problems
0:31:59 in the future, I guess we could agree with,
0:32:04 but like again, the proposed solutions to this are crazy
0:32:06 because we get into these insane places where like,
0:32:09 cool, maybe analog computing is actually better
0:32:10 for training neural nets.
0:32:13 So that is not floating point operations anymore.
0:32:15 I was like, are you doing that now
0:32:18 because of regulations instead of because of pragmatism?
0:32:20 Or, you know, it just, I don’t know.
0:32:22 It just seems too early.
0:32:27 And it, I think that it’s a very, very bad idea right now
0:32:30 for any governments to set ceilings and rules
0:32:34 that are especially onerous for startups,
0:32:35 because I think all you’re gonna get out of that
0:32:39 is just like, you’re gonna get the place
0:32:42 where it’s gonna be all behind four-pay APIs
0:32:44 that are gonna be all controlled by the few.
0:32:49 I mean, that’s strictly better than it doesn’t exist, granted.
0:32:52 But like, clearly we need open source.
0:32:54 Look at everything that happened just now
0:32:58 after AI became available first
0:33:02 and then thanks to Metta and releasing amazing Lama models.
0:33:04 We know so much more about what Nuance do.
0:33:08 We know so much more about how to accelerate this.
0:33:11 You need to rescue these things sometimes
0:33:12 from the ivory towers
0:33:15 and you have to give them to the tinkerers.
0:33:18 And the people who remember how to memory align things
0:33:22 and what, you know, how to get the most out of a hardware.
0:33:23 Like that’s a different crowd than the people
0:33:27 who were writing Python neural topologies.
0:33:30 And it’s, we need everyone on this.
0:33:32 And then we’re gonna get the best outcome
0:33:34 because it is foundational for our future.
0:33:36 – Yeah. – Also, we need to figure out
0:33:38 how to make crypto closer aligned with AI
0:33:43 because it’s a natural way for how to pay for micropayments
0:33:45 and tokens and so on.
0:33:47 And this is really, really important now.
0:33:48 – Well, and solve so many.
0:33:50 I mean, it’s interesting.
0:33:52 Crypto solves so many AI problems
0:33:54 and AI solves so many crypto problems.
0:33:55 – Yeah. – They’re–
0:33:57 – These should be symbiotic
0:34:00 and just the sine waves of excitement of these two areas
0:34:02 just happen to be dissonant right now
0:34:04 and we need to make them harmonious partly
0:34:07 because of the bad big companies.
0:34:08 (laughs)
0:34:09 Sorry.
0:34:10 Well, that’s great.
0:34:14 So we are ready to take questions.
0:34:16 Any questions you may have?
0:34:18 Yes, sir, back there.
0:34:22 – If crypto plays out in exactly the same–
0:34:24 (mumbles)
0:34:30 – It’s like, I think very naturally the browser
0:34:32 will just like be able to move value around
0:34:34 and you will want to move value
0:34:36 in a way that can be addressed by the browser.
0:34:40 This is like, you can get pretty close to that
0:34:42 but you would have to squint
0:34:44 and you have to deal with like poor UX
0:34:46 and you have to deal with hard on ramps.
0:34:50 I think, I don’t think anything kind of magical happens
0:34:52 other than that people will have,
0:34:55 they need more utilitarian ability to like use money.
0:34:57 Again, I think if you run the counterfactual
0:34:59 and say open AI just for whatever reason
0:35:02 you couldn’t work for Stripe and just decided to,
0:35:05 I don’t know, like use some level two blockchain
0:35:10 and to pay for tokens as a initial MVP.
0:35:15 People would now move incredible amounts of money around
0:35:18 and build startups that might have these kind of ideas
0:35:19 that Sam talked about on stage,
0:35:24 about people sharing value of creating GPTs
0:35:27 and like all of the, you know, the blockchain,
0:35:30 like if you look at Open AI Dev Days,
0:35:31 especially the last 20 minutes
0:35:36 and you open at the same time VS Code to Ethereum contract,
0:35:39 you can basically write it out while he’s talking.
0:35:43 It’s like, so, we have this incredible infrastructure
0:35:45 but we can’t use it part because of fees,
0:35:46 part because of speeds,
0:35:49 because like you gotta solve the infrastructure problems
0:35:50 which I know we are doing.
0:35:52 Like this is people are building through this particular winter
0:35:53 in a way that is like,
0:35:57 seems utterly ideal from my perspective.
0:35:58 So that’s great.
0:36:03 There is, we need to bring killer use cases.
0:36:06 We need to find, like we need to have people
0:36:07 want to use this.
0:36:13 Unfortunately, there is like the focus on these kind of things,
0:36:16 especially once you need culture and social things,
0:36:18 is annoyingly not linear.
0:36:20 And that’s really trips everyone up.
0:36:21 It’s like, it’s nothing, nothing, nothing, nothing,
0:36:22 everything.
0:36:25 And because we have to get,
0:36:28 we have to like toil in obscurity
0:36:31 to build up the infrastructure
0:36:33 until it can support the kind of way
0:36:35 that people want to use it.
0:36:37 And then we need to catch lightning in a bottle.
0:36:39 Yes.
0:36:40 – Thanks, Sabi.
0:36:42 I had a question.
0:36:44 So Shopify works with a lot of merchants and brands
0:36:47 that might not be as tech savvy
0:36:49 or might not understand everything that’s going on
0:36:51 in like AI or crypto.
0:36:54 How does Shopify think about messaging to those merchants
0:36:56 and saying, hey, this is the future.
0:36:58 These are things that you should be using.
0:37:01 – Yeah, I think people actually hire us
0:37:04 for having them be ideal for whatever the internet is
0:37:07 and for whatever their buyers,
0:37:09 like they want to continue buying.
0:37:11 Like again, we start before mobile phones.
0:37:14 And so like just making sure that everyone’s store
0:37:16 is available to some mobile phones
0:37:18 back in those days, responsive, whatever,
0:37:22 just had it tremendously because their colleagues
0:37:24 who didn’t have that like just like started doing poor
0:37:29 and poor as money moved to the small form factors.
0:37:32 And they started out competing their peers in the industry.
0:37:35 So like as this is a super old and probably dumb example,
0:37:38 but like what I’m saying is like the demand has to come up.
0:37:42 Like the merchants offering like you can’t solve this problem
0:37:44 just simply by bringing the supply side up,
0:37:45 the demand side needs to go up.
0:37:50 The buyers should want to spend on the blockchain.
0:37:54 Now, that is the huge beauty.
0:37:58 You have 3% to work with here.
0:38:00 Like the credit card fees are high, right?
0:38:02 Like there’s a lot of interesting incentive systems
0:38:03 that you can do.
0:38:06 But what I found was the most amazing thing about sort of
0:38:08 when I came back to crypto,
0:38:10 paid more attention, smart contracts, if you’re,
0:38:14 is that, and I actually think even the practitioners
0:38:19 in the crypto space are under appreciating this
0:38:22 as a talent is there’s never been a field
0:38:25 of such applied widely distributed
0:38:28 and high quality thought applied game theory, right?
0:38:31 Like it’s the amount of the constructions
0:38:34 and the incentive systems that are being built
0:38:36 in this community are like tremendous.
0:38:41 Like the artists that have done really, really well
0:38:45 through NFTs like a such a good example.
0:38:48 This is the first, and now this is like,
0:38:52 man, this was the first business model of the arts
0:38:55 that we’ve gotten since patronage times, right?
0:38:57 Like there’s a huge breakthrough
0:38:59 and we’ll make a comeback.
0:39:02 This thought, this quality of thought has to be applied
0:39:06 I think to all sorts of other areas and in different ways.
0:39:09 And we can use the infrastructure that’s been built
0:39:12 to do something way better than cash back credit cards
0:39:15 like the first extra 3%.
0:39:17 Now it needs to also be meet the world where it is.
0:39:21 It requires some kind of escrow systems for charge back.
0:39:22 There is real distribution.
0:39:24 There needs to be like it,
0:39:26 not everything has to be trustless.
0:39:30 We should bring trust progress into the systems.
0:39:33 Like it’s called a wallet, which I think is a good name.
0:39:36 But in my wallet is, I don’t think I’ve cashed my wallet.
0:39:40 It’s all cards, it’s like licenses.
0:39:41 It’s like driver license and so on, right?
0:39:45 Like it’s attestations that I’m, who I say I am.
0:39:49 And bringing in trust brokers into the system
0:39:51 through the primitives we have now and saying,
0:39:54 yeah, that person is who we say they are.
0:39:56 Yes, that person can be shipped to.
0:40:01 Like if you write this thing on a package destination,
0:40:05 FedEx knows how to unclog this into an actual destination address.
0:40:08 We have to build this with the industry.
0:40:11 And then at some point,
0:40:15 we’re going to just have a much better environment for commerce.
0:40:18 And much like it’s not slightly better than using credit.
0:40:20 That it can actually be 10 times better.
0:40:22 People move around more since COVID, right?
0:40:25 Like how often have you gotten a package
0:40:27 to some other address that you’re no longer there?
0:40:30 Like this, it would be wonderful to make addresses pointers
0:40:32 that can be updated in the background and so on, so on, so on.
0:40:38 So like work with the actual infrastructure of the world
0:40:41 and modernize it on these new systems.
0:40:45 And so that all that want to accrue to people who want to take up one,
0:40:47 like who actually live in a real world
0:40:52 and not just like the hindsight synonymous, you know, abstraction.
0:40:55 I think that’s absolutely doable.
0:40:58 But man, do we have to get fees down?
0:41:00 And this is like it’s never going to happen
0:41:03 if we go back to a $50 transaction.
0:41:04 Yes.
0:41:07 We’ve got time for one more question.
0:41:10 Yes, right in the front row, since I saw you first.
0:41:13 So Shopify has a really successful partner program
0:41:16 where engineers are building themes and plugins and apps.
0:41:19 Could you talk a bit more about some of the challenges
0:41:20 starting that system up?
0:41:23 And then also, as you see more and more people speak the language
0:41:24 of NFTs and the EVM,
0:41:27 how that can evolve as we get more interoperability.
0:41:30 This is definitely, like, you’re sort of hinting at
0:41:34 how cool would it have been to have these pieces of infrastructure primitives
0:41:37 and we could have governed the entire partner system on a token economy.
0:41:38 Like, that’s totally true.
0:41:43 I think this is, I really hope someone is going to explore
0:41:46 how to build systems like the Shopify ecosystem
0:41:53 because there’s a lot of sharing and a lot of multi-volume transactions
0:41:56 which we are all doing with ledgers in the background.
0:41:59 Putting it up, like putting up any two-sided marketplace is tricky.
0:42:02 Here you have to overcome the chicken-necked problems.
0:42:05 It’s tricky, but if you manage to do it,
0:42:08 this is also really, really self-sustaining afterwards.
0:42:10 We have one of the greatest things.
0:42:14 We have about five different customers
0:42:19 with merchants who start on Shopify who now IPO’ed and are public companies.
0:42:23 We now have one who is building an app,
0:42:26 like on the Shopify app with their own private company,
0:42:32 Public Company Clavio is now in their Bay Bay successor, which is awesome.
0:42:36 Again, this is a wonderful thing of what you can do.
0:42:40 Companies are estimated to produce about six times more value
0:42:47 than they capture to the world, like enlightened self-interest novice.
0:42:53 The constrained vision in so-called terms is real.
0:42:57 And there are so many ways to build,
0:43:01 like so many businesses lend themselves to being turned into platforms later.
0:43:03 The problem is everyone tries to build a platform first,
0:43:05 and that’s really, really, really hard.
0:43:10 It’s much better to extract a platform out of a working piece of software
0:43:13 that people are already using, and so that helped us.
0:43:17 People were already trading themes as zip files,
0:43:19 and we were like, “Okay, well, we can help with that,”
0:43:23 and people are already building things for themselves
0:43:26 on the APIs for integrating their own ERP systems,
0:43:31 and then building this to the point where you could list this ERP connector
0:43:37 and build it like our SDKs made it so that it can be in so many stores.
0:43:41 But, de-forward, we hinted people and helped them into the right direction,
0:43:45 and all these kind of things ended up being incredibly valuable for the community.
0:43:50 I mean, it’s definitely one of the coolest aspects to build,
0:43:53 but again, at the end of the day, what all of this is,
0:43:59 is you build really good software that solves real problems very, very quickly,
0:44:03 and then you build a bit of a game-theoretical system on top,
0:44:12 in which people being selfish about wanting to build something accrue a huge amount of value
0:44:19 to the community, to Shopify, to the merchants, and so on.
0:44:26 I mean, I feel like the most sophisticated thought along those lines on Shopify
0:44:30 is a fairly amateur sort of baby’s first incentive system
0:44:37 compared to some of the things that were built on Ethereum and the blockchains
0:44:42 in terms of sophistication, except everything in the world of crypto
0:44:45 was so pointed at financialization and finances,
0:44:49 but I think people might have missed what could be done
0:44:52 if you think more about products rather than financial products,
0:44:55 and I think that’s hopefully going to be the next wave of crypto,
0:45:00 it’s going to be more products and solving real problems for people.
0:45:01 I’m that amazing insight.
0:45:04 I would love to please join me in thanking Koby.
0:45:05 Thank you.
0:45:08 [Music]
0:45:10 [Music]
0:45:12 [Music]
0:45:14 [Music]
0:45:16 [Music]
0:45:18 [Music]
0:45:20 [Music]
0:45:22 [Music]
0:45:24 [Music]
0:45:26 you
2
In this episode of Web3 with a16z, Shopify CEO and cofounder Tobias Lütke joins a16z cofounder Ben Horowitz for a conversation recorded live at the a16z crypto Founders Summit.
Together, they explore what it takes to build a breakout startup in a competitive market, the changing landscape of retail, and how to drive workplace change while navigating corporate culture and calls for activism.
Tobias shares the story behind Shopify’s growth from a snowboarding store to a global ecommerce platform serving millions of merchants, discusses the moral imperative of creating great software, and offers insights on leadership, innovation, and embracing negativity as a tool for progress.
The episode also touches on the intersection of AI and crypto, the power of decentralization, and the next wave of technologies reshaping business and commerce.
Resources:
Find Ben on X: https://x.com/bhorowitz
Find Toni on X: https://x.com/tobi
Stay Updated:
Let us know what you think: https://ratethispodcast.com/a16z
Find a16z on Twitter: https://twitter.com/a16z
Find a16z on LinkedIn: https://www.linkedin.com/company/a16z
Subscribe on your favorite podcast app: https://a16z.simplecast.com/
Follow our host: https://twitter.com/stephsmithio
Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures.