How to Live in Everyone Else's Future (with Shopify CEO Tobi Lütke)

Transcript: (disclaimer: may contain unintentionally confusing, inaccurate and/or amusing transcription errors)

Ben: Hello, Acquired listeners. Since we did our episode on Shopify six years ago, they have turned into a giant public company. They went public in 2015 for less than $1.5 billion. Very cute by today’s standards. They’re worth nearly $200 billion now and doing about $10 billion a year in revenue.

We’ve gotten to know Tobi and the team pretty well since then, and Tobi is an incredibly unique thinker. We were originally thinking, let’s do an ACQ2 episode on Shopify’s recent developments, everything that’s happened since, which ended up turning more into a philosophical conversation about the state of computing, how AI changes everything, and what’s possible to build in the future because of AI.

David: We also got into some of the more insane moments for Tobi as a CEO over the past few years, like when the company almost became a meme stock going from trading around a 20x revenue multiple before COVID up to 70x, then crashing back down after COVID in SERP. Today, though, they’re nearly back at an all time high, but this time because the business and the revenues are actually doing great, not because of a COVID in SERP–fueled bet on the future.

Ben: So please enjoy our conversation with the founder and CEO of Shopify, Tobi Lütke.

Where have you spent a disproportionate amount of your mental cycles in the last month?

Tobi: It’s a privilege of a lifetime to be part of another platform shift. It’s just every single time I think I have my head wrapped around it, it’s remarkable how quickly one normalizes to  completely futuristic new things appearing. We are all like, yeah, I guess I can just paste this into ChatGPT and ask it questions about it. I’m still hung up on we are software that no one wrote and I have to interview it to figure out its capability.

My time is really just interviewing new models that are coming out for their capabilities, just trying to figure out how to make them idealize non-judgmental teachers for people they keep getting better, and how to create structure around judging new models for their capabilities, figuring out where the edges are of what they can do right now, but also make it so that I can very quickly retest any new model coming against those edges to figure out if we’ve crested another—

Ben: Do you have your own personal test harnesses built to try to understand capabilities and models?

Tobi: That’s exactly it. This is what I’m describing, but you’re nailing it. I have no Tobi eval. It’s literally a folder of prompts with expected and judged results. I run it against every model.

At some point, yes I have these chats and standard questions, and then I make decisions based on this. I can write code that does this for me and run this against every model. It’s funny how your thinking evolves while engaging with these tools. This is not how I would’ve approached this before, but speaking with a lot of machine learning experts, this is how they have been doing this, these ideas of building evals they’re called.

That’s not a term to most people and it’s a niche product of the machine learning world because it’s essentially just batch jobs, which just drive the same thing over and over again, and then judge results.

Ben: I know them as unit tests.

Tobi: Exactly. Unit tests were around, were a revolution in the early 2000s, end of 90s. I remember then people proposed automated unit tests as a thing.

I talked a few before about my wonderful meister where I did my apprenticeship with at Siemens, and he always said, you have two years to develop software after you start, and afterwards someone post cemented to it. You’re never going to change it again. That was true at in the 90s. That’s how people thought about software.

Ben: Software engineering was like civil engineering. You build the bridge, the bridge is done.

Tobi: Exactly. There an end date to it. Then if you wanted to avoid it, you rewrote it. Like Windows was a rewrite. Every major version, they took some of it, but it’s mostly stuff from scratch.

Software evolves forever. I still can’t really explain why bit rot exists. Why does a great piece of software really, really bad a decade later? It hasn’t changed. We have changed. Our expectation have changed, maybe.

Ben: As soon as the new iOS is announced, the old iOS feels like garbage. It’s not that your phone got slower. It’s those UI paradigms expired.

Tobi: That’s right. I see this really vividly. This such a good example because I do all these product reviews. It’s such a bad thing to do, but I just physically can’t restrain myself from hitting that update to latest beta button when it appears anywhere. I cannot, There’s a siren’s call.

David: Do you run beta software on your phone?

Tobi: First version. I just can’t. I can't. I have a single phone and I find all the bugs that are waiting for me as interesting constraints that I can play with just like my apps.

David: Do you have Craig Federighi on speed dial and you’re sending him screenshots?

Tobi: No externalities. I own it. I own my own problems. I am going to the first developer beta, and if five of my apps that I need every day don’t work, I use it as a delightful opportunity to learn about new apps.

Ben: You’re a public company CEO and sometimes your phone is just randomly crashing.

Tobi: It just doesn’t work.

Ben: But that actually is avoidable. But not for you.

Tobi: Whatever the consequences are—

David: This is why Tobi is special.

Tobi: Anyway, as much as iOS is just updating to a new UI paradigm called Liquid Glass, which was they much maligned right off the bat by people, somewhat rightfully so in some instances, but to be fair, Apple has an uncanny ability to introduce things that afterwards seem that was clearly the right thing. It’s not even clear that they’re better. It’s just very clear that they make everything else worse. It’s induced bit rot on everyone else in a way.

I’ve been living with this changed UI for a while and I’m in this product reviews with my own teams. We are looking at designs which are clearly good designs in the context of the world that we are now living in a way. It’s incredible what it does because I’m like, hey, designers. I know you need run on betas. You need to design for the device as it will be experienced by people in the future.

I find that just the biggest job that I personally can do for the company and then my company can do for my customers, which is live in everyone else’s relative future. You don’t even need to predict the future that well if you just live in everyone else’s relative future, which really is just a couple of high conviction, high courage update to beta clicks.

David: It’s not that hard to live in the future.

Tobi: Exactly. You just go experience, develop taste for it, and figure out how to make your own decisions. This is, of course, extraordinary true, making sidekick inside of Shopify and having the AI assistant, which is now massive.

David: Tell us about that. Back to bit rot, I was going to ask before, a lot of code is now not written by humans, and it’s trivial to just be like, yeah, update this, make this better. What’s going on in Shopify?

Tobi: Exactly. What an incredible change. Code used to be, by far, the most expensive thing to produce. I think really, really good code for bearing things is still exactly like this, but around the margins, there’s a lot of different kinds of code that glues together. Inspired infrastructure can now just be treated as a first draft for something. I think that’s actually extremely healthy.

This is a property of almost every great piece of software. There’s a very clear split somewhere, there’s a layer. Like the Linux kernel needs to be developed in one way, and then the Linux operating system is just use the kernel’s infrastructure and pretends every computer’s the same because you don’t. It creates a pretend word that’s easier to reason about, so that then the software can be implemented.

Ben: Abstraction. It’s a beautiful abstraction.

Tobi: Yeah, abstraction, but inspired abstraction. Abstractions can also be bad. Abstraction is just another word for pretension. Everything that abstracts something just pretends it’s different from how it actually is.

Abstractions can be incredibly powerful. One could name which really stood the test of time, but a lot of abstractions are lossy abstractions. They actually pretend the system is a lot less powerful than what it actually is, which really, really matters if you evolve things on top and then you end up on this tower of abstractions. You sit on top, you want something, for which you would actually have to go down here and connect. And then you have a spider web of cables left and right, and it’s all like spaghetti in the end.

Ben: This was like the world of, to make it concrete for listeners, the write-once-run-anywhere mobile idea where oh, we’ll just cross-compile to all of the different phones back in 2012 or so. Phone gap or [...], the abstractions ended up being too lossy where you didn’t actually have access to the capabilities.

David: [...] HTML5 debacle of, okay, we’ll just write for HTML5 and it’ll run like a native app. No.

Tobi: And very, very good software companies believed that to be the way, like hybrid applications over rage in the beginning. Remember, the phones were really, really slow back then. The phones are insanely quick, and even today hybrid apps are not that great.

Facebook famously was all in on this idea. Xamarin is a good example. There are a lot of these picking the wrong abstractions early, really constraints what you can do on top. Very often when abstraction is wrong, it means you can’t actually reach great software anymore. You actually pretended the world is simpler, but you don’t actually allow most lot-bearing pieces into a pretended world that are at some point required tool making the software just really, really good.

Again, I’m a tool maker. Let’s keep beating on this Xamarin thing that no one’s heard of, but I think we’ve described it enough so that people might have an idea for what it might be. Just imagine you use very bad software and the team that started the project decide early to use Xamarin, and then they never had a chance of making this as good as all the other things you really appreciate on your phone due to this decision, no matter how good they are.

Ben: It’s like building your house on top of your kitchen table instead of on the foundation of your house.

Tobi: That’s a very good analogy. Tthere are early decisions that will constrain how good you can make something. In our parlance at Shopify, that means certain ideas, certain abstractions, certain choices bring the ceiling down, like if you have a scale from 1 to 10 of how good something can be. Now 10 is world-class masterpiece and 1 is horrible.

What you hope to do is that you use abstractions to make it easier to reach a good number. It brings the floor up. It’s at least this good. That’s what most abstractions do. From business value, a lot of businesses choose tools to make it at least this good, because that is a passing grade. It clears the hurdle.

David: Why would any customer use Shopify? I could go build my own e-commerce checkout payment system.

Ben: My store’s going to be at least this good even if I do zero customization.

Tobi: Well, if you set out to your own from scratch and this is not your core business, what you get back might actually be completely underperforming what you wanted. You can’t even use it.

David: Might be, will be.

Tobi: Very, very, very likely. It becomes one of us ever like Dev March Projects that you just pull resources in. It takes incredibly good people to rescue a bad project, which is hard to even convince good people to do this.

Ben: Windows vista.

Tobi: Yeah. You end up in this world where bring the floor down is what people hope for. What a lot of people miss is very often the same systems bring the ceiling down. So what you are trying to do as a tool maker, an infrastructure builder, as a product maker, is you want to make a tool that brings the floor up significantly, but doesn’t constrain the ceiling. That is extremely hard.

It might not take decades to build it, but it does take decades to build the mental picture to really understand all the aspects of a problem domain in such a way that you have to look around five corners to know, am I going to restrict someone’s potential with making this choice that’s here ahead of me?

Modern operating systems are wonderful. They don’t usually lower the ceiling. You can build any amazing piece of software on top of Windows, Linux, and OS X now. We figured out how to do this because we have been building operating systems for a very long time, and there were clearly attempts that were wrong.

There are many such spaces, e-commerce software being one of them, where we want to build something that works like this. However, it also matters what is the distribution inside of the user’s office system of who people who hit 10 out of 10, versus if your flow is 7 out of 10, which is really good, is everyone clustered there and one or two people managed to take it to storied heights? That’s also bad.

Again, this is why I think AI is so exciting because its principal ability is really to help massively shift the scatterplot of where people end up in as compared to their own vision. I want people, whatever their vision is, unencumbered is a 10 out of 10 on the scale. I want them to hit it. If they don’t, there needs to be good reasons, and ideally their own. Like the choices they made, may make mistake, maybe their products are not good.

Ben: Not that the platform constrain them into not being able to make something great.

Tobi: Computers should never constrain people. This is not the computer’s role. The computer is to get people to build things where they could not possibly imagine having solved otherwise. This is for sorting, this is for ordering. Computers are there to make humans better, never to constrain what humans can do.

Ben: What does that look like from a product? I could imagine, but what products are you building that take advantage of that?

Tobi: Usually when it is a 10 out of 10, very often people are working with experts. Some people will really dedicate themselves to really understanding all the ins and outs of a really complex piece of software.

In many cases, that’s the only way we could get there. Although (I think) the role of software is to minimize the need for this thing, we want to make it approachable so that people can (again) accomplish your initial goal—get online, have a checkout, be able to sell the product that they’ve spent a lot of time with—and if they have additional time to invest, be able to start with (we call) hill climbing.

Now with AI, you can ask. In fact, we are going to increasingly ask you what your goals are. Or at least, have you ever tried to find ways to deduct that from what people are doing with the software or just straight up ask?

My theory is that people are very happy to tell us what their goals are. A lot of software just doesn’t do it because it can’t action it. It maybe the law, we could put a to-do list there and help you with a record keeping, but we can’t do anything about a to-do list before AI. Now again, with the Agentic Flow, we increasingly can.

I think the way to get to 10 out of 10 is just really sit on the same side of a table, have an intake. You very naturally tell us what you’re trying to accomplish, and then we can show you here’s what a prototype would look like of this idea of that you have, and also implement just these obscure best practices. I’ve spoken with one of our customers, they’re selling rugs and—

Ben: Rocks? Oh, rugs. Okay.

David: The woolly stuff on the floor.

Tobi: Yeah, beautiful carpets, really. I should use the word carpets because that looks better with my accent. They were just not making progress for Europe strategy and about to pull out.

Then they used some tools that we wrote out, not really knowing what it would be used for. Just change the pictures. Literally just update the images of all their products, away from Malibu beach houses to provision apartments.

This was not a new photo shoot. This was actually just do this for me. And they said trip it, because four years ago, three years ago, this would actually technically not have been possible to do. Now, it’s about the most boring thing you can imagine. This is how much the world has changed.

David: At a minimum, it would’ve been like, okay, let’s book a photo shoot in Paris.

Tobi: Allocate some capital organization. It needs someone’s attention, probably someone’s—

David: Get some product there. Let’s lay out there.

Tobi: Exactly. Now this is just done and this has proven out. Again, I’m sure they don’t run these AI-changed product pictures now in production. I’m sure they did the shoot afterwards. But they also discovered something over their business, which would’ve been entirely obscured to it. But this is an applicable lesson across the entire platform and something we can absolutely help very small businesses that don’t have the resources to go reshoot all their pictures.

David: And shoot, like this stuff is now so good that you probably don’t need to do the shoot. Even if you are the biggest business in the world, differences are undiscernible.

Tobi: When you talk with [...] person in the world, they already operate like this. The IKEA catalog has no photos in it. It’s all renders. They create really, really good versions of—

Ben: Car commercials are not videos of cars driving. It’s renders of the cars.

Tobi: Yeah, and you only know because the professional drivers on closed roads have disappeared, which was required before. They don’t need to show us anymore because…

David: I hadn’t noticed that.

Tobi: There’s no one actually driving.

David: That’s the giveaway.

Tobi: Yeah. I think the wave that really hits for me the most is again, I grew up on all the cipher books that everyone my generation really grew up on, or many did anyway. They’re very, very optimistic, but also far future like Asimov and so on.

David: What’s your favorite, by the way?

Tobi: I think Snow Crash is my favorite, and Neuromancer. I recently reread Neuromancer. It’s a really, really good book. In so many cases, especially highlighted this, there’s usually a catch-up chapter. Here’s how we get to the starting point of a book.

One of the events is always in 2110, the pass the Turing Tests. There are celebrations in the streets and then something turns out usually right after, to set the stage. What no sci-fi offer did to predict is that the Turing Tests just passes by.

David: We just blow by it and nobody noticed.

Ben: There was no New York Times front page article that we passed the Turing Test. It’s not even like the tech community was exploding about it. We all were just like, oh yeah, we’re in the generative AI right now. You just can’t tell.

Tobi: Clearly, GPT 3.5 is significantly worse than something you can run on your phone right now locally, like pass a Turing Test with flying colors. This is crazy. This is the craziest thing about the actual world if you think about it.

Ben: Maybe I’ll take back my previous comment. What was actually happening when we passed the Turing Test and no one cared was we were also focused on, oh, is this the path to AGI? I wonder if AGI will be the same thing, where we do actually achieve AGI but we’ll actually be talking about some other future milestone that we’re wondering about.

Tobi: I think humans are incredibly good at moving a goalpost. There’s some expectation-related version of hedonistic treadmill that we sometimes talk about, which just moves forward. We adjust to our circumstances incredibly quickly.

Almost everyone listening lives their own dream life from many years ago. People accomplish their goals and we never stopped to think about it. I think it’s really, really hard for us to judge our own environment and what’s going on around us.

Like on a societal level, this even seems to replicate. Again, the Turing Test was one of the best takes we’ve had. Now, people criticize it because it’s not good enough—I suppose it wasn’t—but it was still the best test that people generally refer to about figuring out if something is intelligent and we passed it. There were no parades, no front page news, no ticket tape parades. We should have.

David: It does seem though that there are a few CEOs, of which you are one, that notice this. You started this whole conversation by saying, you have the privilege now of living through another platform shift. I don’t think the world has quite realized yet that there’s a platform shift.

Tobi: But I wish people would, and I wish people would seep in the remarkableness of it. I don’t know if you’ve ever done this. I love the VR [...] Meta Quest and so on. I use it more than most because I use it for racing simulator preparations and so on. It’s an incredible device.

VR was another one of those things which we thought we would get at some point. Then it came and people use it sometimes for their fitness workouts or whatever. It just folded into the ambient world, at least for some.

Ben: Which we take it wildly for granted that technology is cheap and plentiful today.

Tobi: Exactly. $299 or whatever for a remarkable device. Augmented reality, also something people talk about. You can buy them next week. Like an event next week after we film this, and then afterwards you can just purchase. [...]

David: Meta glasses. Amazing.

Tobi: So that’s all around us. One thing that really struck me is when you play with Meta Quest VR, there’s one game which I really highly recommend people. It’s called Job Simulator. Have you ever encountered Job Simulator? You have to try it out.

Job Simulator is this satirical take of it pretends to be in a far future, and there are museums where people go to figure out what the job was. It starts you on a cubicle, you have to turn on the computer, and to turn it on, you actually have to plug a power cable into the wall. Otherwise it doesn’t work and you just think it’s broken.

It’s so wonderful because it plays on the future historians and robots getting details wrong. I actually think it’s profound and actually brilliant commentary on the world, but what idea this nails is that (I think) we will have VR simulations of these times. This is the golden age of humanity right now. It’s the end of the beginning.

I think in a thousand years, we will study these times the way we are studying Cicero’s times over the end of a Roman Republic and as a significant part of history. And because it’s where we figured ourselves out in a lot of ways. People look at the news. There are a lot of current affairs and events. I think they will fade out behind simplistic terms.

Obviously, there was strike in any time during great change, but it’s so unmitigated good. There are so many things we are bringing to world, these gifts of incredible technology and progress that are coming rapidly, that are all the types of things that are going to give individuals with vision so much more agency, which is the thing that’s always at the top of a funnel of anything that people truly care about.

Everything that people really love, all the products, all experiences, other people. It’s things that people fought for is passion, the results and a residue. It’s a “museum of other people’s passion projects,” as John Collison once said.

Ben: Such a good quote.

Tobi: Such a good word. I think that is probably the best tweet on the entire X platform. It is so good.

Ben: We’ll link to it for the listeners who haven’t seen it. I’m sitting here thinking, if I could describe your life, you started a company in the previous technology era, the web era, and then took it into the mobile era. You have a family, you have hobbies, you’ve talked about racing, and the entire world is going through what I think you think is the most transformative change in the last hundreds of years.

What should you do with your current life in reaction to that, knowing that you live at one of the most crucible moments in history when we look back at it?

Tobi: Me specifically? Or is my advice—

Ben: Maybe you, specifically, and then generalize it. What should people do if they believe this?

David: It sounds like you’re interviewing models every day.

Tobi: But once the floating points in them, just to be clear.

David: You are interviewing foundational models.

Tobi: Exactly. It’s an important distinction.

David: Good distinction.

Tobi: Here is what I’m doing. This comes fairly natural to me and I think it’s a little bit a job for me to transmit some of the awe I feel about the times. I think that what we think of history currently is wrong. Again, people in a thousand years from now who are also going to play 2020 simulator…

David: I’m never going to get this image out. I love it.

Tobi: …they will think about these decades and this time differently because what they will do in due time is reintegrate technology into our own history.

Our own self history is one of sometimes decline. We are talking about the Golden Gate Bridge being built in 2–3 years. We are rightfully saying we can’t do this anymore. We are looking at great works and infrastructure projects and our great generation, the wars and these things, look at this and say we can’t experience a world right now where people build Bethlehem Steel or people build Henry Ford’s factories, where they’re just physical manifestations of incredible power of human beings coming together and creating things is manifest for us to see in front of our eyes.

Therefore people judge that we no longer do this. However, we are actually absolutely doing this. It’s just the great works are kernels of operating systems, the browser, lots of the software platforms, the phones, the incredible advances in silicon lithography. Again, the models that we create and train, and the data centers that support them.

Ben: But some of it is actually physical building, laying the energy infrastructure and the power generation, the power transmission, the buildings themselves, the heat management, the cooling.

David: Less than three years ago, ChatGPT didn’t exist.

Tobi: But it is physically not as imposing. Like a data center, I love the blinking lights down the aisle, but man, if you have been to what a refinery looks like or a natural gas liquidation plant, it just hits you on a level that requires zero analogy to understand how incredible of an achievement each of these things are.

The bridges we drive over are like man, this is humans being their best selves, creating, shaping things, very often not stopping at utility, but also going for just beauty by itself. The ornamentation is lot-bearing because we are creating things. There’s no ornamentation on the outside of it, of a data center that eventually will train super intelligence if that were to.

Ben: In fact, we hide it. We don’t put logos on it. We try not to have it on maps.

Tobi: Exactly. It’s going to be as obscured as we can make it. I think this is one of those things in a world that sends people astray. I love to transmit this because I have cultivated every technologist has to, an appreciation for the abstract in the same way that people naturally come preloaded with an appreciation for the concrete.

I can look at a piece of software and appreciate it in the same way how I can look at the pantheon. Unfortunately, that is not something that comes easy and is also probably not valued with the time that it takes. You really have to rewire your brain in a very significant way.

David: Any human can look at the pantheon and be like…, but not any human can look at a piece of software and say wow.

Tobi: Exactly. We even have the thousands of years of understanding about why the poantheon is like. We can start with math to tell you why the ratios are correct. We are now close in the abstract realm. We can’t describe why software code truly. Like its taste. Some people know when we see it. It’s something you cultivate. It’s just abstract on top of abstract.

If you build a tech company, you have to do what I’m talking about to yourself. I see this in my customers. The reason why my customers love building physical stores, retail stores is because they can stand in them, they can see them, they can judge every piece by touching it. They can watch their customers come in and look at things.

They learn what the first thing is they look at and how they handle it. Do they pick it up? Is it clothing? Does it feel right? Do they put it down? Is that something for them? And they will rearrange the store every day. They’ll hill climb inside of it because they can judge everything. It’s concrete.

With an online store, it’s not. We try really hard to tell you how many people there are and tell you stats about it. It’s always abstract. It’s numbers. It’s not real. It doesn’t hit us even if you cultivate an appreciation for looking at the numbers and looking at stats and become data-driven.

If you would do an analysis of how a day went in your mind while being in an fMRI machine, it would be very small. It would be neuro frontal cortex only that would light up. If you would think about your day in the shop talking with people, it’s the entire brain lighting up. We are predisposed to appreciate the concrete. Of course we are.

Ben: We have this exact thing with Acquired. We release an episode, numbers instantly hit hundreds of thousands, and then over the course of months hit millions. I look at the dashboard and I feel nothing. Then David and I do…

David: But then we do an event.

Ben: …one in-person thing at Radio City Hall.

David: With a tiny, tiny fraction of those people.

Ben: And my whole…

David: Blows up the whole brain.

Ben: Emotions I’ve never experienced before. Memory is created, nostalgia is created, relationship is created. You’re on cloud nine for weeks, and that was nothing compared to what was in the dashboard.

Tobi: It’s so true. I don’t know how to solve it. I think it might be not solvable because it might honestly depend on our hardware at some point. We can convince our incredibly malleable frontal cortex of anything and it can learn to appreciate the abstract, but the lizard brain is not convinced.

Ben: Is the age of VR the answer? Is the next stage after charts, graphs, and tables to create full, abstract experiences of—

David: The millions and millions of people there [...]

Ben: If you were standing in your store right now, this is what it would look like as people were shopping.

David: This is how busy your Shopify store would be.

Tobi: I actually think it will get better then not nearly the same. I just don’t think it’s reachable. I do not think you can replace it. That’s a real constraint and we just have to put on it.

David: But that’s a good point though, that you were saying a minute ago about part of your job and part of the industry’s job is to convey and describe that to the world.

Ben: What should other people do if they agree with you that this is this moment of great historical significance?

Tobi: The actual thing I do is I found myself a thing I really, really care about and try to bridge from a future to people I care about. I build Shopify, I try to understand what the future looks like with any trick I in the book, and I want to help with what’s going on.

Finding a task that you care about inside of the change is I think the best position anyone can take. Doesn’t have to be a software company. So much of music is going to change because we can generate quite extraordinary music from nuance directly. That’s seen as a threat, similar to how words can be generated with AI in an even more obvious way.

It’s an integration. Help everyone figure out what is the clearly human part. Here’s a funny thing. There are a couple of worlds where AI has already massively. Obviously math. The calculator has been there for a long time. The best and fastest calculations are clearly not done by humans ever since a calculator came around.

Chess is such a good example of this. The best game of chess in every year of the last 20 years has been played by machine against machine, and no one gives a shit. No one watches those matches. No one looks at the… no one studies them, but everyone knows who Magnus Carlsen is. We actually care not about chess. We care about humans playing chess.

This is the one thing people really do get very wrong here. We need to figure out how to treat what the new capabilities as instruments that we can play, then explore that space and teach others. I think that’s going to be the most fulfilling and fun position because it allows you to seep yourself in the things that were previously impossible while they are still hard, figure them out, which is always one of the most fun things, and then teach others. I can highly recommend it. It is incredibly fulfilling.

David: It’s funny. We were talking to another CEO yesterday who is a huge American football fan, and last year wrote an AI agent to do his fantasy draft for him.

Ben: Scour the whole internet, take into account all the likelihoods to be injured and all that.

David: Incorporate stuff that people generally like. First he was like, what do people not incorporate in their fantasy drafts that they should.

Ben: Well, first it was what do people incorporate, but they actually don’t know all the information? Great. Go and just all that information. Now, what are the factors that I think people aren’t thinking about?

David: One of those was propensity to be injured. That [...] should be really important. Most people don’t account for that. Great. Write an AI agent does that.

Ben: Or if they do, they’ve incomplete information and you could go get more.

David: Yup, so wrote it, crushed it, won the league, all that. This year, everybody in his league is using an AI agent. Then it was like well, well shoot. Shouldn’t NFL head coaches do that too?

Ben: One thing on this, now everybody’s doing this thing. I’ve been thinking about when you really see the full capabilities of the latest thing in AI and you get a demo of something or you get to use something for the first time, that you’re like, this isn’t how I was using it before and this is not how other people use AI. You think, whoa, I have an edge. I actually can do something that people didn’t think was possible.

It’s become clear to me that even five years from now, people who are successful, it’s not going to be people that are making things in an unassisted way. Everything is going to be assisted. The best people in the world at anything, necessarily will be the people who best know how to use these tools.

Do you agree that this is the new bar to be a human out in the world, is understand how to best be a user of AI tools?

Tobi: Yes, but I think for a bigger reason than just the AI tool. I think in terms of stability, which even our industry had, like the things versus not changing that much.

Ben: [...] 2015 to 2021?

Tobi: Exactly. That’s that time (I think) that entire 2000, like 2010’s decade in tech was obviously people will rattle off a couple of things that changed. We had the mobile phone, we had the platform, and we just scaled the ideas that all come from the previous decade. It’s pretty obscure, but then when things change faster, it becomes very obvious.

There are two groups of people. People who fundamentally have fallen in love with problems and people who fundamentally have fallen in love with solutions. I think that sometimes, of course for people who really fall in love with solutions, obviously, then grief, if our solutions become unideal.

That makes perfect sense and I’m sure it sucks. It happened to me too. I love computer programming, low-level stuff for just no reason. I will never implement most algorithms again because when I type the function name, there’s great text already and I can hit tab and then it’s implemented. And sometimes better than probably my naive first implementation would’ve been anyway.

There’s some artistry loss and some skills that I no longer need. But again, the people who love to solve problems are the ones who are just so enabled right now, and I think everyone starts there. No one falls in love with solutions at the beginning of their careers. You can’t.

I think people need to remember how they got to where they are right now, and realize that this is just a bit of a calcification of who they were. They need to take a step back and realize probably life was more fun when I was walking around looking for how I will do my craft, how I will do my thing.

This even gets so far, even in a corporate world. Like job descriptions, job titles. We are almost pushing people into this, into foreign love solutions. You’re a sales operations expert. What is that? That’s a title for a particular way how companies end up solving a particular operational problem they had at some point. What you actually are is you’re an amazing human being, got hired into a company, and the company wants you to add as much of what you’ve got to the mission.

Clearly, you can zoom out. Everyone does this when going out with friends and venting. You’re talking about how the place could be better. Well, forget about the title. Figure out how you can contribute? Everyone who’s really immersing themselves in these tools.

I sent a memo to my company and we set the expectation that we require that people reflexively reach for AI now. We require it because it’s unfair not to, because the people who do are otherwise going to be the people who sequester all the best careers to themselves. I think it’s fair to tell everyone that that’s what it take now.

David: To be clear for everyone who didn’t see or forgot the memo from just a couple of months ago—that feels like a few years ago—you’re now required in Shopify to use AI as your first pass at everything, right?

Ben: Which was controversial when you sent it, and now seems obvious.

Tobi: Exactly. I didn’t think I was sending a terrible or controversial message to begin with, but then someone taught me that it’s being leaked in bad faith, and then I just posted it. That worked quite well, at least there. I was surprised how many people cared about that because it says a couple of things.

If at the beginning of a project, you use AI to prototype it, and then if AI’s really bad at prototyping it, make fun of AI. It’s not about what you make with it. It’s about that you did it because it might surprise you. Also, very earlier in the conversation we talked about evals. If AI is bad at the thing that you asked it to do, now you have an eval for the next AI that’s coming out. So you’re building your quiver of tests.

Ben: What are some of your evals? You mentioned you have this folder of instructions. What are some of your tests?

Tobi: A lot of this are things from coding for instance, where misunderstands a certain prompt, which it had enough context. I like the term context engineering because I think the fundamental skill of using AI well is to be able to state a problem with enough context, in such a way that without any additional pieces of information, the task is plausibly solvable.

It’s a really tricky thing. Now with a lot of agents, the agents end up being pretty good at figuring out what’s missing and go get it themselves. But I think the crazy thing about AI is it can do this. I think the skill people should build is trying to not require it to do it.

Advice is a useful skill to build. First of all, it makes much cheaper to use AI. It makes much higher likelihood of success. But—this is really funny—I think it makes you better in so many other ways that has nothing to do with AI.

I write much better emails since I became a good context engineer. I now realize how much is unsaid in instructions, and why so many things so much of what people describe in companies as politics is actually bad context engineering.

It’s really funny. I so often end up doing the, what is the root note of a problem here that we all have, and you realize after digging 1–2 layers down, there’s just a fundamental disagreement on an assumption about how to do it. Like what is good? What does good look like for this area.

Ben: David and I have just finally started taking some things that we previously thought, oh, that’s part of the secret sauce of Acquired, and who knows why my neurons fire that way. It just comes out and it’s part of the magic.

We started actually writing down instructions so that AI can do things that we previously thought were part of our magic. It’s less magical than you think. We’ve just never had to actually codify how we do the job before.

I think that’s probably a lot of people. It’s a failure of introspection of what actually makes the work that you make good. Once you actually write down your process and all that and have an evaluation framework for what makes it good, it actually is quite executable by AI.

Tobi: A hundred percent. I don’t know how far you’re down this journey now, but there’s an entire world that opens itself up to you once you start doing this. I have this tool we have at Shopify. We use Anthropic’s term, we call these things constitutions which is maybe a little bit too grand. But figuring out a document that is actually all the things that are not platitudes, where all the things where someone else, some other company would plausibly take the other side off. But both of the real choices. If no one would take the other side, it’s platitude.

Teamwork is clearly something every company wants. Usually, the company values end up being just platitudes. It’s very funny. You end up creating a list of these things that you uniquely would pick a side on.

David: The trade-offs that you’re choosing.

Tobi: Exactly. Write on your uniqueness. And then you get in an absolutely fascinating word, because first of all you can use it. You can hold up to your book and just say, okay, well hit me here. These are things you need to know in your, again, context-engineered window.

But it gets way more interesting when you use your favorite episode’s transcripts. You add this and say, what’s missing from a document? You get suggestions, and you’re like—

Ben: Could we have made Costco better in some way?

David: And it’s scary.

Tobi: What you will realize is this is another head-climbing exercise because this constitution will get much, much longer, and you actually have to collaborate with AI on it. These things are incredibly valuable.

This gets too much into the weeds, but inside of Shopify there are so many things that can be done like this. We have product principles, which I wrote at some point and have since been hill-climbing in a loop like this. Any product in our internal systems when it goes into review time’s, we can run against the product principles and also the specific principles against office area.

It’s not like some control mechanism. It’s honestly just, here is where interesting discussion should happen. It guides the conversation because they often, what we find, is the product principles are wrong or imprecise or stated too broadly with not enough provisors. And because you want to write for a large audience in generalities. But when you do work with AI, you can actually say, okay, well actually let’s write all the statements next to the conviction.

Ben: It’s interesting you need deterministic instructions for a non-deterministic machine.

Tobi: They are (of course) non-deterministically integrated in the end. But this is also the beauty of it because sometimes it tests you. Sometimes it says, hey, this is contradictory and it isn’t. Then you have to have a conversation with it about it.

You realize, you know what? That mistake in the way you read that you made probably happens non-zero times of humans too. Finding a more precise version of stating the thing that you’re trying to go for, that example is also really valuable. This is why I’m saying you have to interview these things.

One thing that I’m really noticing with the latest series of models, specifically [...] and pro, is that I feel like I’m pretty good at asking questions and I can take these conversations with AI into very, very, very interesting places. I have started now asking the AI what my next question should be, and I get better ones very often. That’s pretty impressive.

Again, it wants to predict. And they have no memory very often with all your private conversations. When I’m at my absolute best, I ask better questions when I’m at my average, and it can execute against the idealized version of myself. It’s really, really fascinating and really valuable.

I have run a keyboard logger, what is the active window and take a screenshot of a window of my machine every 10 minutes, and archive this going on 15 years now. It’s something I’ve always been running, and it just [...].

David: So you have an archive of your daily digital activity of the last 15 years?

Tobi: Yes.

David: Holy crap. Wow. What a gift. Nobody has that.

Tobi: I would never imagined this ends up being as valuable as it is.

Ben: It’s how Tobi became Tobi, how Shopify became Shopify.

Tobi: It’s basically that. All of my weekend projects are actually just running through all of this and putting it together because (of course) it’s dispersed, keyboards, lots of noise in it,

Ben: Different formats over time.

Tobi: Yeah. Every once in a while I use Vim, at which point there’s a lot of what’s in the keyboard buffer is complete nonsense. But the nice thing is with enough AI tools, you can turn this into pretty clean timelines.

Ben: Without you having to write a script that’s like, okay, every time I’m hitting a queue, disregard.

Tobi: And of course a format change a million times, and there are missing bits and pieces. Then I have my calendar which I’m now adding to it over the times, and a lot of emails, just headers of who I converse with and so on.

David: Do you have that in a repository? Then you can just, whenever you say, whatever AI model you’re using, like, hey, I need you to make something like Tobi, just look at this, then do it like this?

Tobi: This is exactly it. The actual format was I just always turned it into text files which is the best format in the world because it lasts forever. Now, I’ve been doing more and more work on analyzing it, cleaning it, and adding more things because the value of this is getting incredibly high.

I’ve also been doing my notes. I kept them so there are their times and so on. I have a lot of digital artifacts about myself, which again—

David: What prompted you to start doing that?

Tobi: My initial thing was I just wanted to know what am I spending my time on in general. There were some tools that came out.

David: [...] stuff.

Tobi: Exactly. I forgot the names of them, but there were some tools. The things directly about me personally, I want to keep on my machine. That has been a bias which probably comes from when I started using computers. And these are interesting challenges. Anyway, I have all this. I don’t even how specific got there.

Ben: What’s the most useful thing that you found to input that into?

Tobi: I have a period of time where I went from full-time programming to fundraising, to seeing company things more clearly, and just charting it as a time spent in different applications is interesting. From seeing what I typed into what applications like text docs and stuff like this, I also can track some of my beliefs changing, which is a really interesting experience. The brain is fundamentally far to be a backwards-facing narrative consistency optimization device rather than actually a chronological data bank.

Ben: I read a great book recently that described a key element of the brain as the press secretary when anybody asks you why you did something. The book’s called The Elephant in the Brain. You have this press secretary that comes up with the rational answer of why you probably did it, but that part of your brain was actually not present at the time of decision.

Tobi: I think that’s very, very fundamental. The brain optimizes for narrative consistency. Having some of things I typed on many issues that face a company, it is just fascinating. I haven’t come to real conclusion, but I have the ability to do it. I might be a bit slow to really do the study well. You just have to eat a bit of humble pie.

David: At a minimum, what an amazing archive. You think companies have archives and it would be board minutes. You have a moment by moment source of truth of how this happened.

Ben: I’ll ask you the question and then I’ll give you some context to try to lead the witness a little bit. What are some of those things that you think you changed your mind on over time, that you used to feel strongly on and you feel differently now? Maybe to give the context, everyone remembers 2021. Shopify is the media darling, crazy stock runup. I think revenue multiple went from 20x to 70x, something crazy like that.

David: Human brains and narratives, COVID happens, people at home, Shopify.

Ben: That does this, then does that. I think you have an internal rule at Shopify. You don’t talk about the stock price, but we talk about it here on air, spitting distance of all-time highs again.

I just want to throw that out as some context if you choose to adopt it for the answer to the question of what are some things you used to believe that you no longer believe?

Tobi: This is interesting because now I can do, like we talked about the archive. I can now go and check this. But here’s what my brain potentially narratively lays out for me, is that I have always seen the stock price extremely different. I work on the fair market value of a company and which is an unknown thing, but the stock is a betting market on the fair market value. It’ll oscillate over the thing.

I felt the company obviously fits better in the times because now e-commerce is more important. I think that actually increases the fair market value of a company. It was like we trained for this moment for us. But it went really, really high and really, really low. This is the oscillations over this. It’s whiplashy inside of companies.

But I don’t know. I feel through this, I ended up having my convictions pretty unchanged. The thing that will happen where I get caught up is there was a bubble of belief that we just had to massively increase the size of a company, the head count and so on to deal with this, and for that it would be okay to change the hiring standards.

But I changed my mind on a lot of things. In fact, I think I would change my mind on every opinion I have if someone shows up with good data on it, and I think that’s actually really, really important.

Ben: Can you remember something from the early founder–type days before you became a manager, an executive, and a leader that you’ve done a 180 on?

Tobi: Straight up, I didn’t believe in leadership. I just didn’t think leadership was important. I felt that if you get really good individuals and can point them in the same direction, or even without that, all the right things would happen. I fought this partly because people really disliked leadership. Clearly, the right people would show up.

I believe those things are utterly incorrect now. I think people love leadership. They love complaining about it, too, which they should. It deserves all criticism it gets, but there’s something absolutely wonderful being part of a group that’s being led.

I’m not obviously a leader, especially outside of Shopify. I love just being part of a team, and nothing more fun than having someone who has a vision and building something. Every hackathon I go, I’m just like, hey, I’m taking a vacation from leaving. You tell me what to do. That’s so good. So that’s a distance traveled. That’s an almost 100% swing on this particular issue.

The first version of Shopify, everyone made the same salary. We’re like, hey, let’s not have a difficult conversation by just paying everyone exactly the same. We’ll figure this out.

David: How long did that go for?

Tobi: Well, not long but like my first 10 people or so..

David: I'm going to say Ben and I do the same. It just like, hey, everybody in choir makes the same.

Tobi: I think that’s it. I actually really like this. I know how egalitarian socialist that’s coded, but like, I love starting there. I think that would be cool if that would be perfect way, a good way of doing it. Actually rejecting these things one step at a time is actually perfect, because now you don’t have a derivative idea of something, you actually have some, you’ve ran the test for you.

Ben: It strikes me that when you started the company, you knew nothing about building a company or leadership or having anything of this scale. You had to, from first principles, go from doing it wrong to doing it right along the way. Where are some places where you found shortcuts where you didn’t have to learn it yourself and you could look to some other leader or bring in someone great and that person actually just had like a whole bundle of shortcut.

Tobi: One of my most important realizations at some point.  I found myself at some point only doing things I dislike doing. I’d like to think that even the things I dislike doing, I can learn to, I think everyone can get to 80% of state of art pretty quickly on any topic. I was just miserable at doing this the reason why, like, everything I like doing, I had no problem delegating to other people because I know it’s fun. Everything I didn’t like doing, I did myself because I didn’t feel like I should ask people to do it. Luckily, I did it. I overdid this to a degree that I had to that it broke. It’s really good to overswing.

I find that if something just works, you don’t have to counteract, but if something really breaks you when you just have to examine the entire picture. I just ended up realizing everything I don’t like is someone else’s dream job, that's incredibly liberating. Since then, I’ve been creating lots of beautiful dream jobs of stuff that I don’t want to know. People ask me a lot about, hey, Tobi, why do you actually still do programming? And why do you find it valuable? I have a mind, I can’t find things that aren’t valuable fun either. It ends up being the same to me. I did, I do trust that if my mind wants me to go in a certain direction, that there’s a premonition, that it will be valuable in some odd way even as analogy also as like, perspective or because I meet people that are just.

Ben: Probably the successful founder archetype, by the way. If it really comes down to one thing. I think it’s having that intuition where the overlap of things that are interesting and exciting to you and value creative, just like happened to be perfectly overlapping.

Tobi: For instance, the leadership thing. A lot of this came from learning an instrument. I really like playing guitar, I’m very bad at it, but the funny thing then is, I got fascinated with music theory and different genres, and I got so much more out of music. Eventually we got to spend some time with jazz musicians who are very high near the top of a pecking order in that world, especially guitar related. Only after I understood how a jazz band works, did I realize what leadership actually is. Because this is where the business world is just incorrect.

I feel like it's  so chain of command but a jazz band, you have a clear leader and sometimes the leadership changes and shifts around, but what the leader does is not dictatorally say what is playing and how we are spending our time, and what to eat. They just invite other people that have merit for what we want to do. In fact, it may be not even just invite any people and ask everyone to add what they’ve got to a very basic set of conventions where it’s going to be in a key, it's going to be a temple.

It’s emergent, the piece of music is being created at the moment the next note is entirely dependent on the merit, skill and taste of the people who are showing up. I think this is the best metaphor for a great team is like, an elite. There got to be constraints. It needs to be in the key, it needs to be at a temple. But sometimes even the members of a group propose changes. We go to a different set of course, we go to a different key. We go from major to minor. People play dissonance to create tension, which is against the rules to then cause a merging back of a music. Bringing it back home feels incredibly powerful to everyone listening.

David: Yeah.

Ben: That’s mastery, knowing the roles so that you know when you can break them and why.

David: This sounds to me like the definition of a technology company.

Tobi: Exactly.

David: Henry Ford couldn’t set up the assembly line like this. Every worker needed to do the same thing, and every car needed to come out the same way. It had to be any color you want as long as it’s black. Software doesn’t work that way.

Tobi: Software doesn’t work this way. People do try to build software this way and really, really give, try to turn it into something that actually resembles what a car assembly line looks like after it’s created. That’s really, really dialed in. Everyone does one widget and you get 2% out of individuals that you hired. Because they are incredible, they have a lot to give. They have a lot of priors. They give them the room to contribute but then hold them to it. But you would be amazed how often this happens. I’ve learned so much from video games about business.

David: You’ve talked about this a lot. Wait when you stopped asking for personal research here, when did you start playing video games with your kids? And how did you do it?

Tobi: I think we didn’t do screens until they’re two years old, which is picked. I don’t think that’s necessarily even good advice. I think maybe it’s, it’s fine.

David: A two year old can’t play video. But my daughters is four and is like, she’s just on the cusp, started the way

Tobi: The way we played video games was we used an iPad. We were lying on the couch together. I start Minecraft and they tell me what to build. They point at things, I build them. This is us playing video games together. Some of my most fond memories because even two year olds would just basically just know words, letting them choose what we built. I played Minecraft a lot before that. In various instances, I never built such interesting things as the kids told me to do. We built tree houses and  things on top of tree house. I wouldn’t have imagined this because there’s so much creativity there.  That's already a small team.

I allowed them to be the jazz director of our play session.  I think video games are very malignant and nothing is like, some of it deserves criticism, but it always question what does it replace. In those moments, I probably needed some downtime and this was some way for us to do this. I have a tricky job and sometimes I need to shut off my brain too. I’d rather do it. Doing something together with a two-year-old being creative than being upstairs playing Minecraft or so like not exactly what I would be playing, but,  I think this is the most important question.We did this very early, and then it became much more intensive during COVID because I was really worried about how cocooned everyone was.

I was really worried about the lockdowns and what I would do to kids, especially in that more, my boys were in ages where failing softly was just really important for them, you just have to make mistakes. I’ve got to where I’m by just having failed more than others, most of the time. I like failing at things, I like being bad at things which is something I want to cultivate with them too, of it being very agent in the world, making lots of choices for you to stretch yourself.

David: It’s so fun. I was so pre-bias this way anyway, but my daughter’s, just at this age, she’s about to turn four. Last week, we have a steam deck. I’m previewing my carve out for the next episode. This ongoing saga of David’s carve outs about.

Ben: Will he or won’t he get a steam deck?

David: It’s gone from me playing this thing to her being like, I want to move the character. Then like, watching her figure out how to use a joystick, and the first night is like watching a monkey try to use a joystick, and now she’s moving the character around. It’s like, whoa.

Tobi: It’s so cool. It’s so important to not jump in and do it for them, obviously you can do everything better, this is such an important thing to let them actually. They asked me for help and I give them pointers of how to do it better. It takes a lot. It’s so hard. You want to short circuit it.

David: Last night she was, oh, I want to talk to these people. I was like, okay, hit the A button. Then watch her figure out what the A button is, unfortunately she was just spamming the A button. I’m like, okay, no, no, no. You see, or figure it out.

Ben: Imprecise instructions. Hit the A button.

Tobi: Need better context interview.

David: Yeah. Yeah, exactly.

Tobi: Just like a recent thing is I’m dyslexic. Two of my kids are also dyslexic, it’s frustrating because it just gets you, it makes some things quite harder. Then the school, it’s wonderful. Her middle kid finally got frustrated with just not being able to type. Well, I’m like, okay, well what are you going to do about it? And let’s build a habit somehow to do it. Then he was doing a typing trainer, and it’s just, my God,  he was very good at it and did it for 10 minutes a day, but man, it’s that boring. Then I went out and clearly someone must have solved this. We ended up finding on Steam a game called, I can’t remember of a name, but it’s like, it’s like a rogue light typing game where, you’re in the middle and words come at you and enemies and you need to type them to shoot them then you can choose upgrades. Not all my kids are playing this type trainer.

David: Even though the other ones don’t need to learn it yet.

Tobi: The younger one’s also getting good at typing now because they’re all like doing 10, 10 finger typing and getting to like this. I think he went from 16 words a minute when he started, he’s now at like 70, he types like average speed of most people, that was  months.  I wish I would’ve learned to type better. Let me tell you, as a video game, you’re going to fall into this thing and you come out on the other side being an excellent typer. We have not integrated video games correctly into a society either. We are treating them too frivolously and as a family who solves problems with choosing video games as a thing that we’ve done periodically, this has been just one of those, I talk about it parenting hack, it’s amazing, it works super well.

Ben: I have to bring it back to Shopify, is that okay? I have a company specific or company journey specific question for you. If you’re an outside observer to the company, it seems like this meme stock-ification thing that happened in 2021 is a very consequential event to the company then you have this 80% draw down. You’re in this  public market valley of despair situation. What is it like living through that? And if you could build a time machine, what would you go tell yourself?

Tobi: It is very hard for me to, I’m trying to key an interesting answer to almost any question. Because that makes it a puzzle of a good conversation. How to take this in the most interesting direction. I find this one is really stumping me very often because it just wasn’t that interesting. Maybe the most interesting thing about what you’re describing is that it wasn’t interesting in a significant way. If I would have a time machine, I guess the only thing I would tell myself is that I should quickly get the credit rating and the facilities down to do a buy bag. Once it goes down, that would be good. I would be opportunistic about it.

David: I love that spoken like a true entrepreneur.

Ben: Own more of my company.

Tobi: Yeah. It would be awesome. I love his company,  best investment I make for sure. I don’t think it was interesting. I think it was inside of a company, we were fair market value of a company. It actually didn’t go down this time that the stock went down. It did not change.

Ben: It was actually intrinsic.

Tobi: Everyone in the market are wonderful because people buy and sell and someone buys it with a high conviction and someone sells it with low conviction. If you have a lot more people with low conviction then the price is going down because the clearing price shift and the low conviction is not due to what the company do very often. Low conviction might be just because the vibe changed because people’s conviction that there will be bias tomorrow changed and.

Ben: Interest rates changed.

Tobi: Interest rates changed. Russia invaded Ukraine and every stock moved in the entire market. But none of us companies has shifted in fair market value outside of maybe military contractors. It isn’t that interesting and it doesn’t affect the company a lot, but it hangs over company. It comes up a lot. I’m usually somewhat upset with is sell side analysts who gave buy ratings at these ratings.They are the only professionals whose job is to inform the public with good advice and like telling people to continue buying. If companies go to 70 something times revenue, it feels like an asset test for your quality of your predictions.

Somehow ends up being entirely unproblematic for people doing that where somehow the fact that they did that and made some people purchase at these valuations, which then lost a lot of money afterwards on the sell side was probably Goldman Sachs because they do fundamental analysis and realize that this is probably overvalued and sell it to people who then lose their shirt. That feels bad. I love the fact that I’ve made a lot of money for a lot of people who had conviction, the company. Who purchased Shopify early, I made sure the most different thing I’ve probably done to most companies of my cohort is where they went public very early. Like I went right. It’s like over 10 years ago.

Ben: 2015

Tobi: Depending on how you count, Shopify, 2006 launched the product Shopify, which really is beginning of that we were public I guess nine years later from there.

David: Today short. Today it’s like well, maybe in decade three we’ll go public.

Tobi: We went public with every dollar that we ever raised still in the bank and add a billion and a half valuation.

Ben: We went public at a $1.5 billion valuation.

Tobi: Yeah.

Ben: It’s funny, we did a Shopify episode and that’s a crazy low number by today’s.

Tobi: Yeah. I love it because it means like, hey I love the facility of public companies. I think it’s so cool.

David: You’re public means people participated in your growth company.

Tobi: People can buy it on conviction and say, I like what this company’s doing. I like its positioning. I’m going to put this into my, hopefully tax free savings account for a long time. I am very motivated by building something that becomes of more and more value to the people who have high conviction.  I love it and I’m the person trying to convince everyone else to take their companies public because I think it’s just the right thing to do. Again, it’s also valuable for a company in many ways. It’s not difficult. Sometimes it’s annoying, that’s true. But everything is.

David: To your point about like for every job you dislike, there’s somebody else who’s dream job.

Tobi: Exactly. Yeah. Frankly I thought we were data driven before we went public, and we really got data driven through process of taking the company public. I learned a lot. Again, 20% of the process was, I felt pretty sorry for myself for having to do, but 80% ended up being utterly valuable. Just like every challenge you said.

Ben: What are some other things in this 2021 period as historians, David and I can’t help but zoom in on these moments of great tumult for companies and see how they changed. I’m curious, one thing you’ve shared with us before is that, even though everyone’s issuing all these buyer ratings, you actually felt like we’re not a very well run company, and you actually had to change a bunch of stuff, including people. Can you walk us through some of the things in that 2021 forward period where you look back and go, okay, we’re a much more resilient company today because we took X, Y, Z actions?

Tobi: I change the way I work inside of a company because I just felt I need to. I have special privileges in terms of history, such as I’ve seen every version of it. Therefore I have an enormous amount of context. My own context engineering is easy because it’s all preloaded with every important decision with, I’ve been doing shop for 21 years. That’s helpful. That helps a lot in product, and it helps a lot in strategy, and it helps a lot in these things. During this time, Harley who I’ve been working with for 15, maybe 16 years now, is he’s

Ben: The president of Shopify.

Tobi: He’s president of Shopify. He's an absolute, wonderful human being and incredible storyteller. Just creates lots of capacity for me to be more internal focused. Other than Harley, every other executive is more recent now, and it’s a combination of things. It’s not because anyone’s bad, but like with everyone of a previous team, it became a really, really good optimization function to learn a particular way to make me look another way. To speak about your area in a way that sounds really rigorous and good. You’re probably not deciding over my current biggest problem, which is the one I want to work on, therefore I’m going to look elsewhere.

The thing that COVID did is like, there was just no possibility of me being diverted in attention because I just went through everything and through every project, every spoke with every team directly when, when something was really stuck in engineering, I talked with engineers and so on. This is how I used to run the company before I got into taking a little bit more of an abstract view of a company. I did more of this and realized this is what I love doing. It’s tons of work, but I would probably do it if you wouldn’t pay me for it. No shop have books like this. Every four weeks I review every project we have internal ledger like this GSD thing.

Every project goes in, AI analyzes it against all our product principles. We are caulking with gas now. Building a company that’s just fundamentally built around a certain conviction about how a cohesive way of how great products are created and getting to explore if  you’re right about it, and finding, figuring this out, is this actually a good way to work together? It's cool. It's something that is unique to, I think, founder run companies. They're enabled in such a way that you can really build them differently.

David: Yeah. That was the thing  when you started the answer. The thing about this period in history was it actually wasn't that big a deal, and it just smacked me in the face like, only a founder led company could say that. Because if you weren’t a founder led company, you would’ve gotten fired.

Tobi: Yeah.

David: Exactly.

Tobi: Exactly.

David: A hundred percent.

Tobi: That’s the craziest thing. Again, I really make a massive distinction, founder run companies, ideally we should find a different term in companies. They’re just so different from more managerially run companies, more traditional companies. They are just completely crap in so many ways. If you judge them basically just from perspective of companies, companies have figured out and are operational.

Ben: All the most valuable companies in the world are founder run or look founder run in some way.

Tobi: Yeah because there's a playbook for major run companies and they’re bad at adopting to the new things because they’re fundamentally, again, no one carries the full legitimacy inside of a company. Legitimacy is a currency in the companies. No one carries a full legitimacy in a company. Even the CEO, the CEO just has a lot of input on the plan. The plan carries the legitimacy and so in a managerially run company, the plan for the next year or years, if you’re in a good one it’s the thing that everyone defers to. The CEO is an avatar and advocate for the plan, so changing anything requires an enormous activation energy where, if a legitimate legitimacy is invested in an individual, then you can pivot on a dime.

The people who don’t like this, don’t work in these companies. You have broad agreement on the way to work. That leads to bad outcomes often. But like if the individual who’s playing this role is very conscientious and is very open to feedback and like change mind their mind, like strong opinions weekly held, if better data shows up and so on, it can operate at a quite a good batting average, which makes they have highlights and the accomplishments will end up making up easily for the.

David: It’s funny, I’m remembering maybe it was a Don Valentine thing or maybe a Mike Maritz thing. We're sitting here folks who are watching. We’re recording in Sequoia’s offices because my home is under construction, so here we are. But I think Sequoia had a rule or one of their playbooks of evaluating founders has to be right a lot. That’s the judgment you need to make as a shareholder in a founder-led company. Do I believe this founder is going to be right a lot? That is the case that like, great.

Tobi: Yeah. Exactly. That’s what you need to aspire to.

David: Actually, I guess that was borrowed from Bezos. That’s a Bezos.

Ben: I think be right a lot is an Amazon leadership.

David: That’s right.

Tobi: It’s a really good one. It’s so to the point.

Ben: What do you think, let’s draw a clearer distinction between managerial led companies and founder led companies, because I would argue Satya is not Microsoft’s founder, but it’s starting to feel a lot like a founder led company or I’m trying to come up with another good example.

Tobi: But this is a thing, Satya is my go-to example of someone who transcend

Ben: Jamie Diamond.

Tobi: Right. Here’s the way I would frame it just to have some term to hang an analysis on which can incorporate people in extraordinary individual like this. I think both of them have given refounding events to their companies. I think they have made such large changes with such clear breaks from the past that they are actually either the authors of their company’s cultures now,

David: Certainly, certainly in JP Morgan’s case, we went through the whole thing with it, like bank one this is JP.

Tobi: I know lots of people from Microsoft

David: Absolutely

Tobi: Microsoft like Satya did too.

Ben: And these things aren’t as significant as what Warren Buffett did to Berkshire Hathaway. It’s completely unrecognizable. Microsoft still resembles the Microsoft of old.

Tobi: It’s [inaudible:01:25:21] opposite on many convictions. It’s like using WebKit for its browser, not Internet Explorer. It's very pro open source. It’s more open source than Apple is for sure.

Ben: It's shocking that they’ve become a great business without owning the important platform of the era. That was the entire thing that their success was predicated on in their heyday. They don’t have that importance now that relevance, and yet it’s still a great business.

Tobi: Yeah. You could argue, it’s actually more of a platform now. I don’t know, this is not the episode about Microsoft. I appreciate Microsoft deeply, and it’s full of incredible individuals and it's hard to find fault there. It’s just  such an incredible company and such a storied business. Again, I asked Bay to build a company that lasts for a long time like Microsoft has been incredibly successful at this and managed to reinvent itself. As a second most valuable company in the world, I think for, for free or something.

David: It's interesting to hear you include CEOs like that in cadre. There’s a group, we do need a better term.

Tobi:  Frank Lipman goes and companies and turns them into sufficiently moves them to be given them refounding events. I think this is actually an employable strategy. It requires enormous in [inaudible:01:26:44] fortitude and will probably not work often.

Ben: And the existing shareholders handing over the control for someone to. In a one-way door manner, take the company over.

Tobi: My sense is we will see a lot more of it. One of my executives, Cas, is going to open door and obviously that story will have to be written about Hovis Coast. I think that is a business that everyone agrees requires a refounding event to have a shot. You see more recognition of that coming within him coming in and I think compensation system even set up to facilitate this thing.

Ben: It’s $0 salary and it’s all based on the, if the stock does well. Something like that.

Tobi: That’s a founder type conversation like a compensation that is similar to study.

David: Jamie Diamond, I’m going to put half my net worth into this company that I’ve just joined as CEO.

Tobi: That’s right. People underestimate how expensive hedging is. But all this to say, we’ve come here to talk about founder led companies and managerial companies. Currently, people don't understand how stratospheric different they are. The thing that will contribute to understanding this in the future is because we actually, I think will determine that there is a middle ground here that is you can get it from both sides, that is super highly enabled individuals and with an acceptance for things going wrong, as long as the balance asks towards great success is what you want right now, the only people who are in a position to ever get their other founders of companies, because the shareholders didn’t traditionally let them. I think jobs got back there, but wouldn’t have if you would’ve come in as a new hire. Like at Apple. I think they’ll find new tools and then I think people are going to realize.

David: People appreciate this now in a way that they didn’t before. How the playbook of VCs used to be step number one fire founder .

Ben: It’s like we’ve seen enough iterations of the simulation now where you can say, oh, actually it is worth it to vest all of our trust into this one person.

Tobi: There are going to be plenty of individuals who are not with that trust. It will fail and it’s risky and people don’t like taking risks, especially investors. I’ve opened door that isn’t working, but people who made this deal with cash are going to not look very smart. It’s high alpha upside downside, which many people avoid.  We will see this being [inaudible:01:29:43], anyway it will be more common.

David: More and more companies are operating in the technology world. Which is going through this ridiculous, crazy platform shift as we talked about in the first half of this conversation. You have to operate this way as a company.

Tobi: This is super important because we are not talking about a generic. There’s no question of where in the spectrum you want to be. If your business is a widget supplier to General Motors or something just operational effectiveness, sweat every detail. Frederick Taylor stopwatch everywhere being founder run Bayer doesn’t do anything. It just adds chaos and adds hard to predict quality control, I imagine. There might be someone here, this might be an industry which is being reinvented right now and maybe,  I think the default position should be you want these to be actual companies. This is why I also feel like it’s actually better to call the founder run even public companies ventures or missions or something. It's probably nonsensical to make, try to make that change and make that stick because it will never stick.

David: Is this part of your process of discovering that leadership is important and for you and for Shopify?

Tobi:  Yeah, absolutely because I felt very, very dumb the moment I made this realization.  I feel like I’ve had everything basically through most of my life to make this realization very, very early. It took me way too long but it’s that leadership and consensus are two sides of the same spectrum. They don’t intersect at all. Every time consensus makes a decision, it’s the absence of leadership. I also think not enough people know this, and I don’t think. I still feel like it is a bit of a secret because even in our language, consensus is a positive term. Maybe it’s fine, but consensus is always the absence of leadership. Everyone abdicated their leadership responsibilities to cause a consensus to emerge and prioritized the presence of a consensus to finding a best solution. Some people are so clever that they can manufacture consensus around what they would have chosen anyways and operate this way. That’s the only time these things combine, usually it’s you decide one thing or another.

David: Do your alarm bells go off if you start sensing consensus around something in Shopify?

Tobi: Even worse than that. I actually use it when I don’t want decisions to be made.

Ben: If you want to punt something, you’re like, kumbaya, we’re all going to get in the room.

Tobi: I think most companies play too much with pricing all the time. It just causes all these problems with data and it makes it hard to predict and then suddenly you test discounts and then you cheapen the product because now people are just going to wait for the next discount and stuff like this. Obviously, obvious things are valuable strategies, but I think people overdo it and make full-time  positions, changing pricing and so on. Every five years I say, okay, let’s look at pricing.

From first principles, get to a pricing. If that happens to be the same we have today, we don’t do anything. In other cases, we make changes and then it goes back into fulfillment. We are not looking at pricing again for five years just because otherwise it’s too distracting to a company. I used to just forbid it from doing pricing, now I have a much better way I create a pricing council. The council can, if they come to a consensus that something is a better idea, they can then pitch me on it. It’s like super easy to say no at this point and they don’t get to a decision anyway. They don’t have to deal with it anymore.

David: You have two layers of protection. Perfect. You’re not going to get to a decision and then you still have the beat detail.

Tobi: Yeah. Once you see through the matrix,  there are moves become available to you, which are highly efficient.

Ben: This is like whenever a company tells us that there’s a VTE forming to look into something or some like committee or like what, so it’s nobody’s job.

Tobi: No, I think that V teams can super work. It just, you have to have a leader. That needs to be extremely clear who’s responsible.

Ben: This is the old Apple thing, the DRI, the directly responsible individual also. It has to have someone's name next to it. Is it?

Tobi: Apple might be too, but Amazon is definitely big on the DRI thing. Yes, exactly. I’m a fan of making this quite fluid.  I’m going to be a team member in a half day project. I’m not a leader, I think very often, if not in commonly the person who’s a leader of an area is different from the highest paid person, you’re not actually making the choice of leadership sufficiently is one test I apply. By default, MO in every company is the highest paid person making decisions. This is actually why I take a $1 salary.

Ben: But that you benefit by far the most if the company does well.

Tobi: Of course because I can and it’s fun.

Ben: What’s the term the hippo, the highest paid person in the organization is always the person that’s right in a room?

Tobi: I want to signal that I will come with opinions. I want everyone to give me better ones because that’s fun. So consensus, decision making and leadership. I think a spectrum. If you see instance of leadership, you also saw a company that has the ability to avoid needing to make every one of its decisions via consensus. The problem with consensus decision is on that spectrum from one to 10, it’s usually a six or a seven.

Ben: Mm.

Tobi: By the very process of making a consensus, you optimize for lack of downsides. People are almost always better than five because they don’t get the middle. You get the best version that everyone can agree on.

David: It’s like back to the floor in the ceiling thing of. Consensus will get you a floor, but it will cram that ceiling floor.

Tobi: Exactly. That’s exactly what it does. While in a presence of leadership, you can hit any number one to 10. If you’re a company that wants to hit 8, 9, 10 or die trying , then you have to.

Ben: Rule by fiat. It’s the only answer. More charitably, like when we’re looking at creating the best products and the biggest companies and the most successful organizations, what you are looking to do is create an extreme outlier. Extreme outliers can only be created pick your term, whether it’s by an auteur designer or rule by fiat or leadership rather than consensus. Outliers come from a single vision executed by a team of people, not an agreed upon vision, executed by a team of people.

Tobi: That is a hundred percent true. This is very, very consistently demonstrated in every area where it’s quantifiable how good things are. I don’t think a great book has been written as a big collaboration between multiple people. Books are hitting are as good as they are because they’re always single author. There’s a supportive agent in the editor, clearly. But that's a supporting role.

Ben: This is so interesting because I completely agree and I'm obsessed with auteur driven things, and yet we operate a partnership where neither of us are the boss.

David: Yeah. But it’s a totally different thing. We’re a two person partnership. We are one person.

Ben: I suppose a two person committee. We’re like a married couple at this point. It’s like a fused brain.

Tobi:  I doubted that you wrote by a committee. I think every single time I see a great partnership, it's when I see people who are actually learned to respect and trust the other person, like each other over in which areas they tend to have a better, higher betting average.

David: That’s actually true. We don’t operate by consensus at all. We operate by, if one of us wants to do something, we do it.

Ben: We should search our iMessage thread for defer to you.

David: That’s just like anytime one of us wants to. We’re both founders, anytime one of us wants to do something, we do it. That's what it is.

Tobi: Exactly, I work very closely with lots of people. We’re give a right to overrule me. My co-founder of Shopify, Daniel Weand was a brilliant designer, chief design officer eventually. He had [inaudible: 01:38:54] over every feature you shipped and like this, this was, I made the decision on what was what we were building. I wrote most of a code, even, especially because at some point it was the two of us only. He said no to features that I wrote and wanted to ship because we couldn’t figure out how to make the UX code for it. That was important because Shopify’s killer feature 2006 was we had brilliant UX. We were the linear of our time with Figma. It’s always been incredibly important. Giving the people who have a highest context and the greatest, the offer of ship over aspects of the company is also not consensus. That is actually straight up leadership again. The leader on that aspect becomes that person. It’s important that you need to be able to realize.

David: You shouldn’t listen to this and be like, oh, I need to be like Tobi and I need to be the leader on every single thing in the company.That’s not the lesson.

Tobi: I absolutely am not, what I don’t believe in is pushing decision making out towards its grassroots or down organization idea. I also don’t believe in top-down decision making. I believe in pulling decision making inwards. I create the highest ideal liquidity environments in which we can make very good decisions quickly, but where it’s very clear who’s making, making decisions and who’s consulting on it.These review cycles I talk about are that like, we are going to a thousand projects. It is a crazy thing to do. Text starts at like 8:00 AM and usually runs to 10 at night, three days out of every month and the follow ups from this are probably for a weekend. I do follow up reviews and stuff. I happen to like working, which is good.

Ben: You sound locked in.

Tobi: It’s good. This is what I want to do. I’m doing a thing that I didn’t set out to do. I’m trying to build at massive scale, like a societally important piece of software. If we subtract Shopify out, if it disappears overnight, there are a lot of countries which would have no GDP growth,

Ben: Second order of GDP because of Shopify is massive of all your customers.

Tobi: Yes. Plain GMV like gross Pro profit dollars, just sales across the platform is like, I think a third or more of a Canadian entire GDP. Like if you,

Ben: What's the number-ish?

Tobi: It’s like 400 billion things. It’s just absolutely crazy.  I didn’t hardly mix it, so I don’t talk about it.

Ben: I think that would make you either the second or third largest retailer in the world if all of that GMV was accruing to one retailer instead of the distributed

Tobi: Yeah. Which obviously it isn’t. This is a highly virtualized way to think about it. In terms of the checkout, the shop fair ships, in the top three busiest checkouts on planet Earth. At a growth rate that might end up being the biggest. That’s a tremendous responsibility, which I don’t operate too much like on regret minimization. A slightly different framework there basis famously talked about future of regret minimization.

David: What is the nuance?

Tobi: My biggest regret would just be, hey, I didn’t take it seriously and responsibility seriously enough. I think it sucks when people use bad software. I think it sucks when anyone’s underperforming their own potential due to tools. I can do so much here.  I’m dedicating myself to this thing. Very early we talked about fund run companies versus managerial companies and how in management a CEO is being fired and results are bad, which I think is fundamentally, that’s, that’s multiple levels of wrong. The way to judge a CEO is out of what opportunity did you what company? And maybe even for what reasons. You look at something like eBay, which by rights should have all of commerce on the Internet.  They have their first and had everything figured out on PayPal.

David: Incredible business model asset light.

Tobi: Shopify does not own drive. They had the entire thing in the PayPal acquisition, think about the staff. It’s like Peter Thiel coming in, Elon Musk max left Chen and like a million others besides got David Sack.

David: The guys that started YouTube

Tobi: Yeah, exactly. Think about the talent density that they’ve got. From that moment what, what’s eBay now?

David: Not that.

Tobi: You can even play, this is the craziest thing. You can even play that game with Microsoft. Obviously the codes have now said that Apple can charge 40% on all economic activity on top of platform with some emergent providers and maybe cover out. That means Microsoft could have done this with a pc. They got an anti-trust over shipping of web browser, which is so insane if you think about it. What they should have done is have an app store and charge 30% of literally all money on [inaudible:01:44:27] right now. What trillion dollar company would that be? Obviously I’m not advocating for this particular reality. I’m very glad this didn’t happen. I think maybe the entire planet on Earth is off better because of this.

David: Well, we have Ben,  you have your theory that every great company has latent economic value within it that it chooses not to access story potential energy story potential energy.

Ben: Yeah. A hundred percent

Tobi: Out of what opportunity did you have, what is actually the right question? Of course, much harder to judge, but at least for the purposes of every CEO evaluating themselves on the job, that is a question they should ask themselves, that's certainly the question I ask myself.

David: Okay. Give us your version of the re regret minimization framework.

Tobi: I do agree with it. In general, I think it’s just lossy and I think it’s much enhanced.Before I knew that was the way Jeff talks about it, I always somewhat, maybe more psych philosophical. I just feel at the end of life you meet with people, the person you could have become. I think work and life try to minimize the difference between that person and who you actually ended up being. That’s certainly what I do. I think one ingredient in this

Ben: Say that one more time at, is it at the end of life?

Tobi: Yeah. At the end of life, you get the, essentially, the game over screen is you get this, you get to meet a person. You could have become, if you would’ve

David: The best version you could have been.

Tobi: The best version you could have possibly been. I think your goal on planet Earth is to minimize the difference between that and that potential. I don’t think this is a truth of any kind, but it’s a very, very valuable story to believe and a thing that I get a lot of value out of. I think one of the most important ingredients is not just think about future self regret. Don’t think about what you’re at a deathbed you would’ve wanted to do. You need to consult your inner 16-year-old too. It’s so important. We live in such amazing times and we have accomplished so much. Everything we currently have is something that we dreamed of at some point. It’s important.

David: Yeah. What a perfect place to end it in a Yeah. Couldn’t agree more.

Tobi: It feels like such a cheap way to make better decisions, but I wish more people would do it.

Ben: Well, Tobi, thanks for doing this with us.

Tobi: This was good fun. Love that. That’s great conversation. Thank you very much. Thanks for your platform. I, again, keep saying this, but I love your podcast, so great work. Thanks for letting us be a fly in the wall on your own discovery journeys, on we story IT companies and people.

Ben: Thank you. Thanks, Tobi.

Tobi: Awesome.

Ben: All right, listeners, we’ll see you next time.

David: We’ll see you next time.