Fireside Chat with Tobi Lütke & Harley Finkelstein | DotDev 2026

True. It's true.

Pretty good, right? That was our first site really in Ottawa, Canada. Which is by the way, like you made that sound like this super lame thing. Like the trick was the moment you said that you never had to have that meeting because no one wanted to come to Ottawa. Exactly. And now look, we have 2,000 of them here.

All right, that's amazing. Now, here's one more. Raise your hand if you were at the legendary garden party of 2017. Look at this.

You want to do it? But like I remember so it's like so okay I'm not sure if you have a bit on this because you know.

But like so this was actually was one of my favorite things in the entire Shopify history actually. Because so we had Unite first day we did in San Francisco. And it was actually I think a phenomenal event at least you know just like I think it like all came together. Pretty sure we talked about Remember Me on stage, which then eventually got renamed to Chai. So, that's doing pretty well, I think. Yeah, keep that going. It's had a good run so far. Yeah, it'll make it. So garden party. So, second day we woke up elevator and elevator didn't work and you know, just like the entire city of San Francisco did not have power.

If I remember right at least because some kind of suicidal squirrel jumped in a transformer somewhere and like which was very funny because you know from us building tech companies in a you know far reaches of the world here like in Canada at this moment being in San Francisco is like you know this is this bit of pilgrimage for people in tech and there we are in a you know high castle of the technology industry with no socket in the entire district yielding any kind of electrons. So anyway, like so what we had to do is like pivot the team like I think we were opening we couldn't go into the building because the fire marshals told us that they were not sure that the sprinklers work because I mean I don't know why the sprinklers need electricity but because that seems like a bad idea to begin with but somehow we couldn't use the buildings and we pivoted like the entire team just like said hey like let's move it in a parking lot and we did impromptu things people found generators and battery packs and portable PAs and stuff like this. And so, we I think we had to kill the conference.

Why are we dwelling on this because I like just I think because it's like this is such a Shopify thing, right? Like it's because I'm pretty sure day two was better than it would have been with original planning and it would have been a great event any like even there. So like there's just something absolutely amazing by you know we have like thriving on change as a core value ever from day one and you know just there was no better example than that pivot in like a couple of hours making a better event you know out of that situation. So.

Yeah. I wrote all of those.

Yeah, my test apps for.

It's pretty amazing. Yeah. More importantly, does anyone remember Toby's hat phase? Yeah. Yeah. I'm glad you're bringing it back today again. I have the microphone. I can do whatever I want. Yeah. I mean, you've aged, but me on the other hand, look how great I still look the same. Yeah. Yeah. Okay. All right. That's it.

Okay like I mean I my I found this the one weird trick to predicting the future is actually just like live in everyone else's relative future because then you don't actually have to predict anything. You just look around and solve the problems appropriately to the stuff you already sort of encounter, right? So, this is like, you know, like from day one, like I don't know, there grew up in the '90s and we were like when the internet came to town, at least in Germany, this is sort of a way we talked about it. And you know, you grow up in a as a teenager with like understanding like you're sitting on like an important truth. There was no way like the future wouldn't be completely different given what you know was possible and the fact that none of the adults around you seem to care about it didn't seem to be terribly relevant there you know like sometimes it's like it's funny to look back at these things I recently saw a list which I really enjoyed of all the movies that could own would be over in one in it if cell phones would have already been invented yet you know like so you know sometimes like the past looks really funny from a modern technology perspective but as it happens it's harder to predict so you know William Gibson said future's already here it's just not evenly distributed just immersing yourself in all the latest technologies is important obviously the first of all this is fun second of all this an amazing excuse to my wife for buying everything so doing field research here and so then with AI I this is sort of supercharged, but you got to like sort of fine-tune this. Like what specifically?

I wake up right now and it's like, you know, it's like a cycle of you wake up, you see the craziest thing you have ever seen, you go back and do it again, right?

Like I just like look I'm a I'm having so much fun with you know sort of open claw type of build. I think it's like I think if you haven't done this like it really is very very worthwhile to just like go through a trouble of setting up an open claw like a Hermes agent any of these kind of things. There's services that can do it for five bucks. It's a mess like a little bit but like it's a glorious mess. It's a mess that looks like the you know internet mess of the '90s. So like I'm getting all sort of warms and fuzzies there. Like this is the tech industry I fell in love with the but like so I'm a maximalist on this. I honestly think some of these experiences are some of the most profound and honestly compelling product experience I've had in my entire life. So I also went to the extreme of like basically giving access to literally everything like so I do it safely. I'm a perfect company CEO like I was a technologist. I write my own filter gateways like this. So, but it's still first of all ever since last you know you all know this September last year of December really the models have gotten so good that they just constantly blow our minds, right? Like especially Claude is like Claude is a little hacker. If you tell Claude you want something, it'll do it and sometimes in the most dodgy way possible. It's like absolutely hilarious. So you know like maybe give you one example there like is then you know like I was sort of initially slow. I gave it access to more things. At some point I needed my like I just had all sort of problems walking around in the house with like dropping my like Wi-Fi and blah blah blah. So I gave it an API key to just like access to my UniFi system, my Wi-Fi system and like said like go fix it and it fixed it which is incredible. But like the thing that was really really funny was like I you know I was talking about like medical thing. I came back from a sport event. I was like really really hurting. I'm like just like is it bad if these parts hurt? And it like was like okay now it's just like it'll probably heal. So okay. So, and then so next morning I get a message from my open claw and he was doing a like workout that my professional trainer there we did a workout I got a message saying given that you know I had lower back pain after the workout that I'm clearly in right now I should really go in the sauna so it turned the sauna on.

So yeah, so I ended up having to go into logs to figure out how this happened. So like through the UniFi connection, it like figured out that there was a sauna. I mentioned sauna before at some point and so it actually found it on the network found that it's like has a local access figured out how to like talk to it and turn it on and I'm like dude honestly I was actually really touched. It was like one of the nicest things, you know, like so, you know, so you just go to this incredible experience with these incredible events and so like I mean I have so many claw stories and but like yeah, it's just like again immerse yourself in the technology and sit with like the incredible things, you know, the biggest change in the world here with AI is really that like I think incredible things happen in the world when one of two things happens. Something that was currently like previously either really expensive or really hard becomes like you know 10 times cheaper or 10 times easier to do or if something that was previously impossible becomes possible. Those are the biggest shifts. And so I think we have seen all these three things happen at the same time here where like basically everything that was previously impossible is now just simply hard. A lot of what used to be very very hard is now actually easy and a lot of what was expensive is now quite cheap with maybe proviso that tokens are quite something. So hopefully that's temporary. We'll see. But you know that's like the most rich environment for innovation that is possibly imaginable, right? Like and so I think that's it but again it feels like I've had the privilege of seeing many many shifts in a industry. I've been there for again like just even the sort of initial contact internet. I've been there for you know the emergence of web 2.0 which is sort of not documented this way, but really is like I think then the technologists of us who were really really just incredibly interested in the web and its potential essentially raised our hands and said hey from an infrastructure level browsers broadband like and kind of sort of depth of our understanding of a medium. I think we are here and we know how to make great systems. And also you know maybe business model like with software service like that so all rolled in I think web 2.0 is underdocumented but plays an important role there because it gave essentially this world of technology the business model that it didn't have during which caused automatically you know the dot com crash right the dot com crash was like all the promise was there literally every company it could launch was right Like every one of the business models made sense even no matter you can make fun of I know Pets.com all day but like pets.com was a totally sensible e-commerce store right like.

Yeah. Exactly. So you just needed like infrastructure that like if your first order of business is you know pay a million dollars to Larry Ellison for getting an Oracle database so you can start programming then he's probably you're probably not going to connect with your market at the price levels that they want. So you know just but they were all right and so eventually this started working and coming together and you know things that were previously impossible became possible like using credit cards securely became possible and so you know just this shift I think is worth tracking and then immersing yourself and just like playing be playful with it do shenanigans technology shenanigans are like the toys of the crap people do on a weekend are the multi-billion dollar companies of all, you know, and so it's like.

So your our meetings are super chill. But that is amazing. But I want.

So yeah like yeah yeah my clankers have access to my heart rate and you know like put it into a little DuckDB and warehouse because why not? I at some point destroyed it. I would like it to do it and since then it's been dutifully doing it. It's wonderful. I mean the amount of shenanigans that are running right now in the background while we are sitting here on the stage for you is absolutely mind-blowing.

Everyone wants to hear from the boss that their meeting raises the heart rate more than anybody else.

Yeah, I mean, look, I love this community. I think it's I mean it's so crazy from my perspective at you know at some point you know we like again like since we invoking so much of the past and you seem to want to lead us down a road here of nostalgia and whatnot like at some point in like 2006 7 I was working heavily on you know APIs with a few of my friends there and you know like we were like well like there's a lot of internal teams at these companies you know this was a big part of our thesis was like software that tries to be everything to everyone is software that no one ought to be using or that is only being sold by doing you know playing golf with a boss because it's completely disconnected from the consequences of actually having to use it. This is the kind of thing we were like really really upset with and we thought was gumming up the software industry and which we appreciated. So we wanted to build super high quality software that scales with everyone and these kind of things right. So, however, of course, if you want to be everything, like if you want every feature or you want to be able to help people with their missions, like you kind of have to build a lot of things. And we were like, let's build infrastructure instead of features. And because infrastructure is composable, composibility to software is essentially what compound returns is to finance. So if you build great infrastructure and invite a community of people who are just like really really close to the problems then that would work better as internal teams and the with the people we were talking to but like if you can allow the internal teams to you know go one step further and instead of making a piece of software a one of one for one particular use case but like follow sort of a technological shape that Shopify took you know shop ids and like make it so that multiple people can use it and we create the conventions then I think we would follow our mission which is really to create more entrepreneurship in the world even more closely because of course these internal teams might actually sit on a real insight they want something to exist for themselves why not make it really easy for them to share it then put this like sort of terrible version first version of a directory up and then allow them to tell each other about it. Right. So this is like at some point this happened there and then you know being in a room with such like you're all such unbelievable impressive builders and entrepreneurs yourself is like is incredibly gratifying. Right. So and so yeah we are standing like and I love the way you put this. we are standing shoulder-to-shoulder. It's at the end of the two best things you can possibly build in a world that are durable are systems and systems of incentive specifically, right? Like so incentive like you can build okay one systems of incentive. Let's go with that. And so we are all standing shoulder-to-shoulder because all of us are doing the same thing. We are trying to help these incredible impressive entrepreneurs who are doing this incredibly brave and courageous thing of putting themselves out there by building a store, a business, a system usually in the face of as you all know like significant amount of opposition. There's a lot of people who tell you you shouldn't. And they are doing their thing and they want like the kinds of people who are motivated enough to reach for independence are people who also want to do it their way and so be finding points of distinction is so important. We are trying to do as much as we can with our template systems but really at some point your business needs to be different because like putting a different skin on top of the same business is not a different business. You have distinction for your products which we want you to spend all your time on. So we want to show up with as much supportive infrastructure as possible. We want to show up there with you get the software that like it should just work and for anything that's like a real reach where you're like ambitious about it or so you can use we bring the infrastructure to you. You're like taking the other side of this particular LAU and converting it into something that's really perfect for what people are trying and creating this point of distinction. And so I think this just works so beautifully. I think what's really then going on in the current times on top of this is that you know the like this is such a opportunity-rich space and you know we were all so limited by you know the speed of really building software that now we can like this has been truly augmented we all know it we all feel it we don't need to spend time continuing thing we can talk about shenanigans that the AI can do, we don't need to talk about it sort of relative merit anymore. And you know with that the like opportunity plus tooling leads to like an explosion of creativity, right? And so again we can give so much more distinction. We can give so much more variety. We can make the web so much more fun. Again, I'm sure you all saw Ben and Vanessa then team's work on you know kind of going back to the basics like saying hey you know it's really really really really great that browser let's actually like celebrate it fully instead of wrapping it in JSON and so you know so there's like from our perspective we like let's you know go to the stuff that matters and you know like build opportunity space and invite everyone over because you know like our corner of the internet is a pretty good one.