Rapid Fire Q&A with Jason Fried and David Heinemeier Hansson

0:00Welcome to Rework, a podcast by 37signals about the better way to work and run your business. I'm Kimberly Rhodes, your host, joined by the co-founders Jason Fried and David Heinemeier Hansson. This week, we're going to do some rapid fire Q &A. Jason and David have not seen these questions, so we're going to get their raw, impromptu, on-the-spot answers. Are these from listeners? The only kind of answers you'll ever get. No, these are like random questions that I've made up. We've done a couple of Q &A sessions like this, and they've always been well-received. So we'll say a little bit of business, a little bit of personal, you know, just go with it.

0:35Let's go. Okay, for both of you, what's the most recent book you've read that you'd recommend? I know we've talked a lot about Blue Ocean Strategy. There's other books you guys have talked about. What about recent reads? Anything you've read recently that you would recommend to folks?

0:48David Heinemeier Hansson:I just started to read something that's way over my head called Something Deeply Hidden by Sean Carroll. It's about quantum mechanics. and I'm just fascinated by it, even though I don't understand it and it's over my head, but I actually enjoy reading stuff that's over my head. It's kind of fun. I like to listen to podcasts I don't fully understand. So it's a bit of a challenge. So I've not gotten through the book. I've heard it's very good. I've just started it. So that's my recommendation. David, what about you? I am almost done with The Fourth Turning, which is a historical analysis of the last five centuries worth of mostly American, but really Western history.

1:24how history has a tendency to repeat itself, that there's these cycles and generations that come. And they dive into just a lot of anecdotes, really, about how different generations of different times were remarkably similar to our time. Like, we love all of us to fret about what's going on right now because it seems like it's so different than what came just before it. And it totally is. But you rewind the clock about 100 years and you will see a remarkable number of patterns repeat. Really good book, The Fourth Turning. Is it true in some sense that this theory predicts all generations to come?

2:03Who knows? But it is a wonderful way to even just learn about the last five centuries worth of American history. Okay, I will link to both of those books in the show notes. Question, Dale, let's start with you. What's something you're using AI for that's surprisingly good? I almost reversed the question. What have I used AI for in the last few months where I have not been surprised about it being remarkably good? I mean, I'm actually struggling to come up with a great example of that because we are very much in the acceleration phase still. I remember back in December being blown away by Opus 4.5 and thinking like, okay, if this is all we get, it's amazing.

2:47We could just stay a year. Like, this is such a productivity boost. And then we got another bump in February. And now with these two new models with Fable and Saul, again, I keep going like, this got to stop, right? Like, we cannot keep accelerating this way, but so we do. So I've been using far more ambitious props. That's really the change recently, especially with this Fable model that Anthropic puts out. I've been using it a lot in the development of Machi. We're just getting finally, after a few months, ready to release a big new Quattro update, the fourth version. And I just keep upping the, I don't know if I want to say difficulty level, but the vagueness level.

3:31Like, I'll just ask it like, I'd like to have some end-to-end testing here. How would you lay it out? And that's really a distinct shift where I was so impressed by AI when I could just tell it to do something specifically and it would do it. I already mind blown. Now, I'm asking the AI, like, what would you do here? And the answers I'm getting back are just as mind blowing, actually more mind blowing, than just telling it to do something and then see it done very quickly. The fact that the latest models are able to analyze and design situations where I went like, this is really good. If I was sitting at 37 Signals talking to an employee and trying like, let me just get their take on it.

4:14And I was getting some of these responses. I would have been like, whatever this person is making or do for promotion. Because this is freaking incredible. Now, again, it's not in all cases. And then five minutes later, I'll be bewildered by something simple it seemingly can't do. But those bewilderments are getting fewer and further in between. It's actually quite rare. And whenever I hit one, you just switch to another agent. And seemingly that knows the answer. So I've not really been much disappointed lately. In fact, if anything, it just is accelerating this excitement for the new phase that we're in, that we can build these incredible things.

4:53I mean, I'm looking back upon just the last two weeks where I really got into a crunch mode with a match in, okay, let's get this fourth version done. And I look at the amount of stuff that's gotten done, and this is months, if not years worth of work. And it's so easy to get sucked into, like, the hysteria of that. and like, you know what? I don't even need to have any religious analysis here about what's going to happen in X number of months or years. I can just look at what's happening right now. I'm sitting in front of the keyboard a whole day. I evaluate what's done at the end of the day and I go, that's crazy.

5:28David Heinemeier Hansson:Like that is just a truly staggering bulk of work. And again, this doesn't apply to everyone. Where I really find actually it applies so well is when you just have this single individual meet working on a single task where I know whether we're done or not. I know whether it's good or not. I can steer it because I know what it's making. That's when you get the most pure acceleration as soon as you start working with other people, as soon as you start working on like, is this actually what we want or not want or who are we making it for? You're not getting the same sense of acceleration because humans are still in the loop.

6:01So anyway, I'm still very excited. Jason, is there anything new you've been using AI for? anything that's surprisingly good?

6:09David Heinemeier Hansson:Nothing new. It's just getting better and better. I mean, the thing that I just find most impressive about it is the fact that you can ask follow-up questions, which is a very basic thing. This is like GBT1, I think. You could basically do this, like, tell me more, or I don't get it, or, but what about this? And people talk about, at its basic level, people talk about, like, AI is sort of a better Google. But the thing is, you could never do this with Google or search engines at all. You could ask a question. you wouldn't even ask that question, but you could, you get results, you go somewhere else and you're kind of stuck with what you got.

6:40David Heinemeier Hansson:You can't ask a follow-up. You can't go deeper. Actually, you could search again, but it's all lost. And this is just a tiny little fundamental thing, but this is what's so good about talking to human beings and talking to interesting people. Oh, tell me more about that. What do you mean by that? That alone, and it's just getting better and better and better and better at that and holding the context in its head and just the way you can push it and prod it to learn more. I just find it to be incredibly fascinating. And that alone for me is just the biggest breakthrough is the follow-up question, the second order question, which is actually where the most interesting things happen anyway.

7:11David Heinemeier Hansson:Like when you're interviewing somebody, you're talking to somebody, you catch a thread and you want to go on that thread. That's what you would do in an interview or when you're talking to an interesting person, that's where you'd go with it. And you couldn't do that with computers before at all, basically. Now you can. And I find that just to be fascinating. Okay, one AI use that I had recently that kind of blew me away. And keep in mind, I'm not as deep into the AI programming that you guys are doing every day. But I was looking for a venue for an event that we're doing and had put in some search criteria for it to like look in a city for this thing I was looking for.

7:42And it came back with what I remember that you liked about this other city you did. So it was like giving me context of like, I know that you already did this. I know that's what you liked about it, even though I hadn't prompted it in that way. Fascinating. That it was like going back into my search history. Like you already did this once before. Here's where we landed and why I know you liked it. So I gave you these other options following that same line. This is the main thing that Open Claw and those persistent agents are getting people super excited about. Persistent memory. which is actually straight to your point.

8:19That is what's so interesting about humans. You get to carry on the conversation. It's not the one lot back and forth. And with some of these long-running agents in particular, the Open Claws and Hermes and the like, now you're carrying that conversation. Tell me more over weeks, months, maybe even years. Okay, let's go to customer service. I want to hear about a positive customer service experience you've had recently and what made it so memorable. I always find when I have a great experience, I'm like, oh my gosh, this feels like it's a standout. So I'm curious if you guys have had any standout customer service experiences lately.

8:54David Heinemeier Hansson:The one I always come back to is just like basic business stuff, like our landscaper. It's just like you can just talk to a person and you have a quick conversation and something gets done. It's like, it's so beautiful and it's so hard for that to exist in the world in most situations that I find just simply going up to someone who's doing something and asking them a question and getting a response and like getting a quote or like, that's actually the biggest thing is like getting a quote. Like we have some hedges that need to be trimmed twice a year. And I just go up to the guy and be like, how much?

9:24David Heinemeier Hansson:And he's like, I don't know, 1300 or wherever the number is. And I'm like, okay. And then we just like, we're done. Like that. Why can't more things be like that? Like, I don't want to wait for a quote. I don't want to get a quote via mail. I don't want you to follow with me later. Just tell me what you think. If it's reasonable, we'll do it. That's not customer service necessarily, but it actually also is customer service. It's not like there wasn't a problem there. But just these fundamental base interactions between two human beings, someone has a simple thing, someone can provide a service, and it's just done with no paperwork, no complications, no back and forth.

9:59David Heinemeier Hansson:I just, I don't know. I find those to be fleeting experiences and incredibly pleasurable experiences. And you just wonder why everything can't be done that way. So that's what comes to mind when you ask that question. David? Funny on the same note, careful what you wish for. I remember just the number of times I've sat at a restaurant thinking like, why can't I just like pay on my phone? Like, I don't want to sit around. I don't want to wait. I want to. And then you get it. There's this very nice place in Malibu where I go for lunch and it's on a nap. And I was there yesterday to have lunch with someone.

10:31and we're both there on our apps and trying to scroll through these menus. And I'm like, this is such a deeply asocial experience. Like we're spending all our days on these fucking screens and go to lunch with another human for the purpose of like, whatever, having conversation. And then we have to start that by both sitting and scrolling on an app. It was stupid. And it was also stupid because it was slow. It was slower than even like not the most attentive form of service. and then I got the opposite experience at the end of the day where I went to another place for dinner and it was just, this guy had just opened this new place a few months ago and was the owner there.

11:11And I sat down, quickly glanced at a piece of paper. Remember those? Where like the menu on it with little prices and I just said like, I want that one and he was like, great. And then that was that. I was like, oh yeah, you don't know what you've lost in terms of service even when it's not the best. In this case, That was a very good exchange, very quick and easy. And you get to deal with the owner. That always feels better. Then the thing I thought I wanted, like, can I just pay on an app, was actually total shit. I was actually going to add, the reason I thought of this question is because I recently had a great customer service experience.

11:46I was booking something for an event and I sent an email in, immediately got an email back asking for a quick call. Fine, whatever. A little annoyed, like I just want to do it online. but in talking to the person, I was talking to the owner and he wanted to spend, he's like, let me just spend one minute to tell you about how we got started. And he wanted to tell me like his company's origin story. Him and his wife started this, it's a coffee cart, started this coffee cart business in whatever year. It was like the highlight of my day hearing this guy's one minute story about how we started. And then I wanted to support his small business.

12:21And literally his email was like, thank you for supporting our small business. That also made me think of you guys and how it's so driven on like, we're small. It's a small company. Like we started this. It's not this giant conglomerate.

12:33David Heinemeier Hansson:Those are nice experiences. Yeah. It was great. Okay. A question about something you've either started or stopped and you can answer it either way or both ways. Something that you've started or stopped that has made your work or life better. Anything you've stopped doing that you're like, oh, my day is so much better because I don't do this. My work life is so much better because I started something new. I'll combine the answer here that it's both. A few months ago, I had the flu for like a week. Never a pleasant experience. But what came out of that was I suddenly didn't like Red Bull anymore.

13:09And it was just like the taste was actually repulsive. And I was like, you know what? I know that drinking a can of Red Bull every day, probably not the healthiest habit in the world. but I kind of like it in part because it's not the healthiest habit in the world right like I live an otherwise fairly healthy life and it's good to have some indulgences and a can of red full a day is about as uh low on the risk tolerance um list as that can be I just didn't like it right and I tried actually several days in a row like maybe it was just one day let me just keep trying and now here I am like fucking craving caffeine and I don't like coffee and I I don't feel like I can do matcha every day either.

13:51So I'm like, I stopped something that was supposed to be good for me. And I want to start again. Can I get my old addiction back? Can I somehow, like, does it taste okay now? And unfortunately, I'm not quite sure that we're there yet.

14:03David Heinemeier Hansson:I don't have anything in the short term that I feel like I've changed much. Summers strangely don't lend themselves to that for me. Like, it's usually kind of an end of the year thing. It gets dark early. And I just feel like those are moments where maybe something else will come up that I'll start to do differently. But yeah, I feel like I don't have a good answer for that one. Okay, that's okay. We're going to go to my last one. What's your favorite part of a typical work week? I mean, other than recording rework, of course. I mean, I just like when we're working on something that's like close to shipping.

14:32David Heinemeier Hansson:I just find that to be really fun to be like we're a day away or two days away and we're really honing in. We're really kind of tightening it up, really polishing it. And then we get to announce it and we get to show it off. Like whenever that's happening, it's exciting. So for me, it's always like near the end. I mean, I like the initial stages of things, too, because that's when you get to really come up with the idea and think about it broadly and kind of hone in on what it's going to be and detail it to some degree and like hand it off perhaps to somebody and like get them excited about it.

14:59David Heinemeier Hansson:But then there's that lull in the middle where like it's just kind of happening and nothing's going to ship today or tomorrow or the next day. But then you're like, like, for example, this week, I think we're shipping a good handful of really cool stuff we've been working on for a week or two. And I just like the energy of like, gosh, are we almost there? I mean, I just get impatient. And so I push a little bit and I start to look at all the little small things because I know once a million people see it, it's a different story. And so you just kind of look at all that stuff. And that's just really particularly exciting for me.

15:25David Heinemeier Hansson:I like those moments. I like to have those moments. And luckily, these days, you can have those moments a lot more frequently than you could in the past because we can get stuff out the door faster than we ever could before because of AI and some other changes and stuff. So I'm very excited about those kind of weeks. I don't like the weeks when nothing's going out the door. David, what about you? It's very similar. And I'm feeling it right now because, as I mentioned, I got this crunch going to get the Omachi 4 shipped and spend a couple of months sort of in that middle phase. We're making a big change.

15:56I didn't want to rush that out because lots of users and whatever. I got to live with it for a while. It's fine. But it's not exciting. The exciting part is that I can see the finish line. and it just gives you such a surge of motivation to sprint through everything that's left. And of course, it always turns out there's way more left than what you think it is. You're just like, oh, I can be done on whatever, Monday or Friday or whatever day it is. And then you sprint towards it and you get there and you're like, fuck, there's still 50 things left. But what I so enjoy about that final phase is none of it is planning.

16:33Every bit of it, every forward motion is like, fix this thing, ship it, great. It's better. What's next? And you get handed essentially a roadmap to completion by just finding the next thing that's broken, the next thing that's not right or could be improved. And then you fix that and move forward. So the rate of dopamine hit from designing a solution and fixing that problem becomes tighter and tighter and faster and faster, which is just a sort of great spin wheel for mustering the enthusiasm and really the time it takes to just finish some of these things. The family was out for a couple of days here going somewhere else.

17:11And I was like, I'll get it done. And I just ended up spending an unusual amount of time, like this Sunday, I was up until late working on this thing. And then again, yesterday, up late, working on this thing. I just go like, you know what? I don't want to be in this mode every day. But when I'm in that mode, it's actually the best mode. It is that flow state where you're constantly being driven forward, pulled forward. I don't have to muster anything special. And again, as I often find it, when these projects are personal, individual, in the sense that I can just keep moving, I don't have to stop, wait for something else or someone else for something to happen.

17:48A lot of it can accelerate by this AI stuff. I'd often find in the past, I'd sprint towards a finish line and then I'd have these five things I don't really want to do. And the reason I don't really want to do them, they're kind of annoying, or I feel like, oh, I got to dive into something. And now you don't have those as much, because now you have this whole team with you, a little clever computer people in the chips coming out and wanting to help you get to the finish line. So you can just keep moving, right? Like those roadblocks in the past that would just go like, well, I'll just slow down and then I'll go again, are gone.

18:21Again, that's also why you can't do it all the time. You can't sprint as much, just 365 days a year. But like those weeks or days up until something is done oh my god yeah i think like if you like problem solving

18:34David Heinemeier Hansson:which i think we do it's nice at the end because there's dozens more problems to solve and they can all be solved quickly yes versus like early in the project there's a lot of big problems that are kind of slower and and you don't see the end of them quickly so right at the end there's like you know nine tweaks in in the next hour we're gonna make to this little thing and it's like so fun to just nail it and get it just right and to have that you know the dopamine hit essentially of solving a problem. Next one, solve a problem, bring it up, next one, and just move through things very fast. So that's always a fun thing at the end.

19:03Okay. Well, those are my random questions for Jason and David. If you have a random question for Jason or David, send it in. You can send us an email to rework at 37signals.com or send us a video recording. You can do that at 37signals.com slash podcast question. This has been an episode of Rework. Rework is production of 37signals. You can find show notes and transcripts at our website at 37signals.com slash podcast. and full video episodes are on YouTube.