Liam Martin & Nick Francis | Two Founders, Two Futures

Part one: Liam Martin

Timothy Allen: With you specifically, we have to talk about how much has changed in the last year since I spoke to you. If anyone wants to listen to the interview we did in Morocco about a year ago, by all means, but it’s me saying things like, what’s vibe coding? And what’s an LLM? And literally a year later, we’re driving around in driverless cars, talking to computers. Seventy per cent of my podcast is now AI. The whole thing has just gone ballistic. I don’t know how you feel about it, but I have this feeling that in the last six months the floor has kind of dropped from underneath me, and I don’t really know where we’re at. I don’t know when we’re going to hit that bottom floor. All right, let me be more specific. I remember you telling me with such joy a year ago that the AI couldn’t do something for you. And then all of a sudden it’s saying to you, I’ll tell you what, I’ll do a bit of research and I’ll come back to you. You were experimenting with the coding side of it quite intensely.

Liam Martin: Yes.

Timothy Allen: Even me now. I wrote a website last week, a new website. I’m really terrible at this and already I can do it. And part of that process was creating a plugin for WordPress, which it did like that. I didn’t even know you could do that. I thought I’d have to get a developer to do that side of it. No, it created all the elements of it. I would say I’m the control, because I’m not very AI savvy, but already I can do stuff.

Liam Martin: What I was talking about in my introductory talk here at Running Remote was how we’re moving from distributed people to distributed intelligence. So it’s not just people working in an office, it’s people working remotely, people working hybrid, people working with AI, AI agents. And that’s such a huge change from where we were even in Morocco last year. It’s moved from how can you pull in the context of what you need to accomplish, not just ask an AI what to do, but pull in all the data sets, the context, the information. And then once you have all that information, it’s incredibly easy. I had in my presentation today, up until about a week and a half ago, Claude Design didn’t exist.

Timothy Allen: We’re probably dating this podcast quite perfectly. I’ll put it out quickly, don’t worry.

Liam Martin: Claude Design did not exist.

Timothy Allen: I don’t even know what Claude Design is.

Liam Martin: Claude Design is basically, what if you could build an entire user interface for an application just by talking to it? So you can build Uber, as an example, instantaneously. Or you can build a presentation. What I did is I had all of my slides inside Google Slides, and I said, let’s toss this into Claude Design and just see what happens. And I also built another tool called Peppermint, which pulls in all of the context from my work. It’s pulling in my Slack, my emails, my video transcripts, all this kind of information. And then it said, well, number one, I think I can put together a really good visual presentation for you. But more importantly, before that, your actual premise is wrong. Two months ago you talked, in a video call on a podcast, about how we’re moving from distributed people to distributed intelligence, and that’s what should be the thesis of your talk, not what I had just spent the last eight hours working on. And I was like, you know what? That’s a better idea. In seventy-eight seconds it built a presentation that the AV guy told me was the best presentation he had ever seen in his twelve years of doing slide presentations for people. The animations were perfect. The transitions were beautiful. At one point, I think I talked to you about this in Morocco, I was really, really good at writing ad copy. I still am. I put my ten thousand hours into writing ad copy. I can write an email, I can put information down on our website, I can really convince people. And I had to mourn the death of me ever doing that again. I recognised a week and a half ago, I will never produce another slide deck for the rest of my life.

Timothy Allen: Jesus. Something you were talking about before, you had an AI basically monitoring everything you did, in order for it to understand who you were and what you needed. Have you gone further down that road now?

Liam Martin: Yeah, it’s called Peppermint, peppermint.com. That’s one of the applications.

Timothy Allen: That’s your company.

Liam Martin: Yeah. What it does is it allows you to understand everything that you’re doing, and it creates a digital twin. So yesterday, as an example, someone asked me, hey, did you submit your taxes to the state of Massachusetts? And my digital twin answered saying, yeah, he did it two weeks ago, but he actually had to use his credit card as opposed to his banking tool, and there are three core problems that he has to deal with. Can you tell our accountant about these issues? And that was my personal assistant that did that.

Timothy Allen: This is a pure digital personal assistant, pure AI.

Liam Martin: A pure digital twin that is a photocopy of me at work and understands everything that I know. And you can access it wherever you want. So if you’re in a Google Doc, as an example, you could say, at Liam, what do you think about this Google Doc? What do you think about point three? And it will literally give you my opinion.

Timothy Allen: Hold on a minute before you go on. Do you get your own opinion on stuff?

Liam Martin: All the time. Or even things like, I was working on this project two months ago and I can’t remember where the Google Doc was, can you find it? Yes, it was two months ago, here it is.

Timothy Allen: It’s a photographic memory for you.

Liam Martin: And then it also gives you the context of what’s going on. I got this from last year. We had Raj from Harvard, who did a really fantastic study with Zapier, which is one of the companies speaking here today. The CEO of Zapier ended up getting a digital twin, pulled all of his Slack conversations into it, and then if employees had questions they could ask Wade’s digital twin as opposed to himself. So we just expanded on the technology fundamentally. That was our big inspiration. So far it’s been really successful. We have hundreds of users currently using it. And it really allows you to get rid of that synchronisation tax that I talked about in Morocco. Sixty per cent of the work that is done inside an organisation is simply moving internal documentation around. I’m going to prepare this presentation for my manager, or get this ready for the board, or I’ve got to do this big slide deck for all the employees. All of that disappears with something like Peppermint, when you have a memory layer you can share with people.

Timothy Allen: Do you feel uncomfortable at all about it watching you the whole time? Do you put limits on it? How do you stop it diving too deep into your life? Maybe you don’t mind. That kind of stuff does make me feel a bit uncomfortable.

Liam Martin: It does definitely make people uncomfortable. Because I’ve built it, I understand how safe it is. We’ve built a PII layer, so it would never record a password, an API token, your social security number. But you can also tweak it so it would never record adult content or financial information if you wanted to remove that. And then you can give complete access to yourself, which is incredibly valuable. But if you want to give trusted partners access to that twin, that’s where the real magic happens, because you have that collaboration effect. I don’t have to be bothered by twenty-eight questions of which twenty-six my twin could have answered on my behalf.

Timothy Allen: Bearing in mind my knowledge of AI is a bit limited.

Liam Martin: It sounds like you’ve gotten a lot better at it.

Timothy Allen: I have, but I haven’t really. I think with me and AI, and this is where you and I differ, I’m very much into self-sovereign AI. I’m much more cautious than you. I’m not opening up as fast, I would imagine. So I’m thinking, if I want to use it in that way, I will need my own hardware, I will need it sandboxed, I will need all this kind of stuff.

Liam Martin: So another option we do have is the ability to bring your own API token. But it is a cloud system, so it doesn’t really work if it’s only available on your own rig, because then the collaboration component doesn’t work. But if you want to use it for yourself, that absolutely can be done completely privately and you can run it on your own computer directly. We’re getting to a point where AI is doubling in intelligence every three months. When we were in Morocco it was probably every year. So we’re going to this exponential point. I’m working in this, and I have to tell you, six months ago I bought a chalet up north in Canada, much like what you guys are building. I’ve got my own water, I’ve got my own electrical. I never thought I would be a prepper guy. And just the speed by which these things are moving. Have you heard about the p-doom rate?

Timothy Allen: P-doom? No, I want to hear about it.

Liam Martin: A lot of engineers talk about p-doom. That’s the percentage that AI will lead to our doom. When you hear a lot of these AI evangelists talk publicly about AI, maybe there’s a ten per cent chance that it could be dangerous, or a five per cent chance. If you speak privately with those same people, they’re around seventy, eighty per cent. I don’t think we’re going to have Terminator, but I do think within the next twenty-four months we’re going to have a bad actor that uses AI to hurt a lot of people. And that’s when I don’t want to be in Dodge. I don’t want to be in a major metropolitan city when something like that happens.

Timothy Allen: It’s funny, that’s a bit of a meme, isn’t it? I’ve heard this similar story recounted a number of times, that very well-known founders are leaving their jobs and their LinkedIn profile says sayonara, see you later, and they’re just disappearing.

Liam Martin: The head of security at Anthropic, when 4.6 came out, he moved to Austria and he’s writing poetry. So it’s real. It’s absolutely real. They’re nuclear weapons that are in your pocket.

Timothy Allen: I get fatigued quite easily with it, and the fatigue comes from constantly feeling like, oh my God, I’ve got to do something else now. I just get on top of it. I’ve got it editing my videos now. I’ve got it reminding me of things and telling me this might be a good thing to talk about. Or with you, for example, it would be, this is what you spoke about before, why don’t you. But then I hear stuff like that and it’s like, do I need a digital twin now? Have I got to do that?

Liam Martin: I think that all of these things are probably going to be really advantageous towards your overall workflow. And probably, what you’re doing right now, you can create more content, you can have more conversations like this.

Timothy Allen: Yes.

Liam Martin: Which is really the dream. Where it becomes dangerous is when that AI decides to completely replace you. So maybe you can use a HeyGen model to have a version of you, and then our two Peppermint twins could load into a HeyGen model to say, hey, let’s create a podcast between the two of us. What questions would we talk about? How would I answer those questions? We could probably do that by this time next year.

Timothy Allen: But hold on a minute, let me stop you there. Me personally, I wouldn’t want to listen to that. I want the human experience. From what I can understand about the trajectory that AI is on, you can see that what becomes the scarce property of life is the human experience, whatever that means. So that’s the thing that’s getting scarcer. The AI is creating abundance. And as you say, the AI could create a thousand podcasts with you and me talking for an hour each time. But it’s an AI you and me talking, and I would have to tell people that. It would probably be a gimmick for me, a one-time thing. Hey, listen to me and Liam talk, our AI twins. But then after that I would think, did a novel thought happen? Was a new thing spoken about? Or why am I even listening to that? Why not listen to a document?

Liam Martin: So I agree with you. However, if you look at Instagram right now, a lot of Instagram girls are not real any more, they’re AI-based. Images are already fundamentally at this point. You’re seeing a lot of that stuff inside TikTok. If you looked at your feed recently on TikTok, I would say for me personally about twenty-five per cent of the content on there is AI-driven, or at least AI-assisted. So where are we at next year? Are we at sixty per cent of short-form content that’s AI-driven? Maybe then we start to address long-form content. And I agree with you, it’s the human condition, it’s the human experience, it’s what I prefer, but the data would show otherwise. So I think you and I would definitely want that experience, and I want to stay with it. That’s one of the reasons why I built Peppermint. I want a human-centric reality for AI, meaning I want the AI to work around the human and help the human as opposed to bypass the human. This is something we all need to pay attention to, because there are a lot of people right now building tools that are going to replace humans. And once enough humans are replaced, you’re going to see blood in the streets for sure.

Timothy Allen: I’ll give you an anecdote. Obviously an AI writes my show notes now. Next Friday’s episode is a guy called Crémieux, a really interesting bloke. It wrote the show notes and I had to rewrite them myself. It got so obviously AI-generated. I actually said, no, stop, and I scrapped the AI’s idea. I couldn’t even send it back to zero. It was remembering, it was learning, it was doing all these things, but it was so obviously AI. So I am rejecting it. I am rejecting even the titles of the episodes. I had it recently picking viral titles for my episodes. And I got an email from a pretty good friend the other day that was written by an AI, and I know it, because I know why sometimes the AI is right. Big dashes throughout a document, all that stuff. So it’s really bugging me. And I agree, there’s a lot of slop on TikTok that people are just mindlessly watching, but that doesn’t mean they’re not going to reject it at some point as well.

Liam Martin: I hope so. The issue is that this is the worst that it’s ever going to be. So it’s going to get a lot better in the next year, in the next quarter. If we slow down, I would agree with your thesis. I just don’t know where the floor is. That’s where we’re at right now. I don’t know where we land, where we say, okay, we’re at a point where we’ve built the maximum amount of compute that we possibly can, and we’re at a point now where podcasts will be done exclusively by humans, AIs really can’t even do it. Is that going to be a point? Or what does the AI do and what does the human do? I actually think that we need, and you’re probably going to cringe at this one, government oversight for a lot of these tools. They’re like nuclear weapons in your pocket. And the things that I can do with AI today allow me to affect hundreds of thousands of people that I couldn’t have even dreamed of five years ago. So we’re in a really interesting spot. I think 2026 is the year where you’re going to be talking about what happened before 2026. I think AGI has actually already happened. I think 4.6 is AGI for Claude, and Codex 5.5 is definitely AGI. And I don’t want that to happen. That’s the issue. I don’t even think it’s a technological issue at this point, it’s a militaristic issue. It’s different nation states recognising that if we don’t build AGI first, we’re going to be subjugated by the country that does. And that’s a really negative spiral to have in society. I just don’t know how to get out of it.

Timothy Allen: Arguably that happened with nuclear weapons and no one blew themselves up yet.

Liam Martin: That’s absolutely right.

Timothy Allen: My fallback thought is always that people will crave the human element. Just because I’m watching myself, and I suppose I could call myself normal. I got a Waymo in Austin for the first time in my life, and I was like, oh my God, taking videos, showing my kids. Now I’ve already had three, I don’t care. And I get that same experience with AI. I remember the first time I tried to get an AI to create a podcast out of something. You remember, Google did it. And it even spoke it, and I was going, this is amazing. Never touched it since. Why would I? The moment I heard someone do that, I thought, yeah, I’ve seen that, been there, done that. I kind of get the feeling, because in the world of film and photography we went through this with things like Adobe, things started looking a bit samey. And then people thought, how do we make it feel a bit different, a bit real? So either the AI says, right, I’m now going to try to create things that look imperfect, which it probably will do. Or we just say, do you know what, the easiest way to navigate this is just to know when something’s AI and when something’s a person, and make your choice. And to be honest, when it’s things like what we’re discussing now, I see absolutely no point in listening to two AIs talking to each other. Why would I?

Liam Martin: I agree with you. And I hope that I’m not at the point where I get tricked in a year or two years from now, where I’m listening to a Joe Rogan podcast for three hours and then recognise, oh, that was entirely created by AI.

Timothy Allen: Do you know what would be fun, is if this was AI and I announced it at the end of the episode. No, it’s not, guys, this is two people really talking to each other. Tell me about the future of work. I’m interested in what you’ve just said about not knowing where the floor is, how far this goes. Because really you’re also talking about how far the number of people that get unemployed goes as well.

Liam Martin: Yes.

Timothy Allen: So talk to me about that.

Liam Martin: What I talked about in my opening talk was, I think we’re moving from distributed people to distributed intelligence. And we have to figure out where that floor is. Right now we’re in this state of flux where we really don’t know. Are our accountants going to be replaced by AI? Are lawyers going to be replaced by AI? Are doctors? No one really knows at this point. Actually this is a nice moment in time, because I think we’ll probably have the answer to those questions by this time next year, for sure, as to whether that’s going to happen. And I hope, I pray, that we can retain some level of autonomy there, because I don’t even know what the sociological implications of thirty, forty, fifty per cent unemployment might be. No one’s been there before. And if everything we’re saying comes true with AI, we could be at a point this time next year with thirty, forty, fifty per cent unemployment.

Timothy Allen: Really? Twelve months? People seem to make those kinds of predictions a lot. What were we talking about a year ago?

Liam Martin: I’ll give you a really good corollary. OpenAI has raised a ridiculous amount of money. If every single person on planet Earth, all eight billion and change, bought a subscription to OpenAI, it would still not equal the valuation that OpenAI is currently at. So that leaves you with one very clear conclusion, which is that their endgame is not for people to get subscriptions to OpenAI. Their endgame is to replace hundreds of millions of people’s jobs. So if I can pay two thousand dollars to have an accountant that does absolutely everything for me, and does a much better job than my current team of accountants that I spent sixty thousand dollars on last month for my taxes, of course I’m going to take that deal. And a whole bunch of other people will take that deal as well. But then what do the lawyers, the accountants, the doctors, the executives do with their time?

Timothy Allen: Well, I don’t know.

Liam Martin: And that scares the living daylights out of me.

Timothy Allen: I don’t know either. We’re going to all produce poetry?

Liam Martin: No way. There has to be some type of, you have to strive for something. At least for me, I have to have something that is difficult that I’m working my way towards. It’s just how I’m built. And if I can’t do that, I’m very unhappy.

Timothy Allen: You seem to have the same opinion as a lot of AI insider people, which is, I don’t know what’s going to happen, I can’t even advise you. Because it’s coming for, well, I say it’s not coming for all of us. The media has been through an event like this. When we went from analogue to digital it absolutely annihilated the way people do the media. But other things have come out of it, and me as a person, I can still make a good living doing this, I just do it in a different way. So I’m not an AI doomer in that sense. However, take driverless cars. I’ve had more Waymos than I’ve had real people in Austin, and I’ve been getting them regularly. So that is happening. There’s obviously a business model there working. The one I got this morning was literally seven bucks to get from my place to here, and it would have cost twenty had it been a human. What can you get for seven dollars? Taxis used to cost a lot more than that. So it’s obviously going to happen. I don’t think we’ve yet seen certain parts of the economy getting hit, like your average lawyer. I think the fair thing to say is that at the moment it’s replacing the people who do the jobs that AI does, which in the case of a company will be the new people in the door working their way up, someone down at the bottom. You think it’s going to eat its way all the way to the top, do you?

Liam Martin: I think if I had to bet you a dollar, in a year from now, they’re finished. Everyone. I think the majority of white collar work will no longer exist in five years.

Timothy Allen: So what are you doing then? Oh, you bought that place up north.

Liam Martin: I bought a place up north. It’s got its own water, I’ve got electrical backup.

Timothy Allen: What are you going to do, though?

Liam Martin: I’m never going to tell you where it is exactly. It’s in a very nice place. I’ve got about twenty acres of land. I’m going to hang out.

Timothy Allen: But what are you going to do? Just hang out? You know that doesn’t work.

Liam Martin: I’ve got two kids. So for me, my mindset really changed once I had kids, which is that it’s not so much about going on the offensive, but maintaining the defensive line. But yeah, I agree with you that for me, I need to strive towards something to be content. And if AI is doing it all, just like what I said, I’m never going to produce a slide deck ever again.

Timothy Allen: How did you finish your talk? Because I didn’t hear it, I was interviewing people. What was the end of it? How do you leave a crowd like that with the information that probably everyone’s going to be out of a job? How do you make that something motivational?

Liam Martin: I’m saying this as a warning to prepare people for what’s coming. I could sugarcoat this and say I think everything is going to be fine, and it very well could be. There’s another version of reality, which is that it’s literally like us moving from typewriters to personal computers. I don’t know if you remember this, I remember it very briefly, which was typing pools. You’d have these armies of primarily women who would transcribe documents. You would have a document that would be transcribed by a typist, and that was a big component of the economy. And within years that was gone. It was all replaced by personal computers, because everyone was doing that work directly. That same transition might occur with AI. And if it does, that’s a very utopian version of the future, because I would see everyone being able to do much more of the creative work that they really love and less of the boring stuff they don’t want to do. My concern, however, is that technology has never really just said, oh yeah, okay, we got to the point where all the boring jobs are gone and only the interesting ones are left that are the most expensive, why don’t we stop there? That’s never happened in the history of civilisation. Another really good interpretation, which I love to quote, is the invention of electricity. Electricity was what was defined as a general technology, meaning it wasn’t a better shovel or a better television set, it was a piece of technology that affected everything. And I think that’s the transition we’re currently going through now. But instead of getting rid of manual work, we’re getting rid of intellectual work. And that’s why it’s so much more concerning for me, because I’ve spent the last twenty years of my life pursuing intellectual work.

Timothy Allen: Well, it’s getting rid of manual work as well. Robots will exist. It’s getting rid of work. And to be honest, I know a lot of us complain about work, but without work we are pretty useless. We have to work. I think it limits us, which is a great thing. Abundance sounds like a great idea, but in reality, you could do this with your kids. If you give them abundance they grow up gnarly kids. If you give them some scarcity in their lives and some boundaries, you get great kids. It’s an existential problem that we’re facing.

Liam Martin: Do you know the highest population historically of horses? What was the single largest amount of horses on planet Earth?

Timothy Allen: It’s probably in Mongolia, I would say. Is it?

Liam Martin: No, historically.

Timothy Allen: Genghis Khan?

Liam Martin: No, it was 1910. In 1911, the Model T came out. Now we’re at about one and a half per cent of that 1910 population. Horses didn’t change. It’s just that they have no more value to the economy. So when you calculate it out in terms of GDP per capita, if that capita is mechanical as opposed to organic, meaning Claude Code is now replacing the accountant, or the Tesla robot is replacing your driver, or cutting your hedges outside your house, or cleaning up your house, what jobs do humans do?

Timothy Allen: I don’t know. I don’t think there are any in that sense.

Liam Martin: Okay, so then you agree with me.

Timothy Allen: No, I think I might agree. I disagree on the outcome. I think at that point we will limit ourselves, because we do that already in many ways. I’ll give you an example. You could buy anything. If you’re rich enough, you can buy whatever you want. But some people still buy old cars, because they want to buy an old car, they don’t want to drive around in something new. So they’re deliberately limiting themselves. Same thing, I think I told you this before, I used to run expeditions to remote locations. They were very hard and very uncomfortable, very cold. It was a horrible way to live. But people paid tens of thousands of dollars to do it. They were literally paying to put this hardship in their life. So I’m almost certain that as humans we will find a way to insert the hardship back, to insert the limitation back. And I think that’s how we’ll move forward. Because I think without it, we’ll turn into brats. We’ll turn into horrible kids.

Liam Martin: I agree. I’m so hopeful that that is the vision of the future that happens, because I think the other one could be absolutely disastrous for everybody.

Timothy Allen: It is. But hey, let’s not leave on a bad note. Give me something. I hate leaving on doomer notes, give me a positive outcome. Are you going up on stage again at any point?

Liam Martin: Yes, later today.

Timothy Allen: Give us your best positive keynote ending.

Liam Martin: Listen, I think there’s a thirty per cent chance that we’ll enter a utopia, a sense of abundance that you won’t even be able to recognise in comparison to where we were two to three years ago. I think everyone will be able to get whatever they want at any time, because labour will be free. And in an economy where labour is unlimited, where that capita, the GDP per capita, is mechanical, and I can print out a ten-thousand-dollar robot that, by the way, also prints itself out for other people, the economy is unlimited. So that’s an incredibly exciting future. And that’s a future that I hope exists, because that’s a point where we can really strive to pursue who we are, and there will be no must inside of what you do. But at least for me, there’s no must in what I do, and I still pursue something. I still make sure that I can be pushed up against something, that I can try to push myself to the next level.

Timothy Allen: Don’t go doomer on me.

Liam Martin: I don’t want to go doomer on you. I hope that we have this version of the future that I just described. I’m just really stressed out that it may not be the case.

Timothy Allen: I will take the thirty per cent bet. Liam, thanks, mate. Nice to see you again.

Liam Martin: Nice to see you too.

Timothy Allen: Wonderful event, well done for this. And thanks for coming, thanks for speaking again.

Liam Martin: No worries. Thanks.

Part two: Nick Francis

Timothy Allen: I didn’t manage to work out why it was called the second wave of remote work. What was the second wave bit? Can you elaborate a little? I’m interested in the macro view of all this stuff.

Nick Francis: I can explain it very quickly. The first wave was the pandemic. It was when everyone was sort of forced into remote work. If you could work remotely, you started to work remotely. So that entire paradigm shift was a big wave for remote work. Before the first wave it was niche. What I said on stage is that I was just one of many entrepreneurs who got made fun of for having a remote company. But the first wave normalised it, it made remote a normal thing. Now ninety-eight per cent of workers say that they want to spend at least part of their time working remotely. The second wave is when remote work is now optional again. We spent six years learning the best practices. Now remote work is optional. There are a lot of return to office plans, but there’s a high percentage of the workforce that still wants to work remotely. And so as somebody who runs a remote company, I’m looking at an exponentially larger talent pool than pre-pandemic. So in this second wave there’s a much larger talent pool, but the stakes are much higher for remote excellence. Just doing the basics, having a remote-friendly work policy, is not going to cut it. So most of the talk was really about what it means to be remote excellent, to go beyond the best practices and to really embrace remote first as an operating model.

Timothy Allen: But you were remote before the pandemic, right?

Nick Francis: That’s right.

Timothy Allen: Almost a pioneer, I suppose, in a way.

Nick Francis: Yeah. Just no wave yet. It was just flat water.

Timothy Allen: So compare that to now. What’s the difference? Because in a way you’re seeing in this wave some of the same things you saw before Covid, because Covid was a big push that made people do it. Now they don’t necessarily have to.

Nick Francis: Now remote work is normalised. If you work in tech, you have remote working experience. Whereas you may have been sceptical prior to the pandemic and thought, no, I want to go to the office every day, it’s just what makes me comfortable, now a lot of those folks are like, I’m not going back to the office, I prefer to work remotely. So not only is the talent pool exponentially larger, but pretty much every company has the remote best practices dialled. So if you want to separate yourself and acquire the best talent, you’re going to have to be operating at a different level than just remote friendly.

Timothy Allen: And did you go remote originally for that reason? Was it for the talent pool?

Nick Francis: A hundred per cent. I actually love going into offices. I love a beautiful corporate office as much as the next person. But at the same time, I love working with the most talented people who challenge me and bring out my best more than anything. So being limited geographically to a specific radius around this beautiful office that I want to have just didn’t really make sense to me at the time. So in 2011, when we founded our company, we just felt like, look, the way we differentiate, the way we win, is on talent. We’re going to make it so that we can hire the best and brightest talent from anywhere on planet Earth, and we’re going to make it work.

Timothy Allen: I see that’s a great selling point, and a lot of people, especially yesterday on stage, this seems to be one of the big talking points, the talent pool. Obviously there are downsides as well, otherwise everyone would be remote. What do you see as the main downsides?

Nick Francis: I would maybe reframe it as trade-offs, because I think there are some very specific trade-offs that you have to make in order for remote to work. Because I understand the value of being together in person. You can still do that in a remote company. You can still do retreats, you can still do small team offsites, you can still spend time on Zoom, which is about ninety-five per cent as good. There are ways to create the sort of connection that you’re looking for without having to be in the same geographic location all of the time. So one of the trade-offs is that in a remote environment I’m going to have to spend a lot more time on team connection, because there’s nobody roaming the hallways, there’s no water cooler. I’m going to have to overcompensate and invest proactively in making sure my team can feel connected to each other and to the business. But the other side of that is, when people get to work at home, believe it or not, they’re a hell of a lot more productive.

Timothy Allen: And there’s data to back that up.

Nick Francis: There’s data to back that up. So that’s why I think of it as a trade-off. There’s some good that comes of it, but there’s also another side of it. And so if you understand both sides, and on the negative side you overcompensate and ensure that you can keep your team connected, then they’re going to be way more productive. It’s a win-win.

Timothy Allen: It’s true, actually. I’ve been quite jealous listening to people talk about the kind of retreats they go on here.

Nick Francis: It’s awesome. Our last one was in Rome. We literally crowned the new Pope while we were there. It was awesome.

Timothy Allen: I put two and two together on this particular trip and realised that that is a massive part of the remote work thing, these incredible gatherings. Do you think it’s enough to keep your team going?

Nick Francis: I think the ideal situation, because we’ve done both, is that you do two all-team retreats annually. So every six months you’re spending a week together. And then on the small team side, for our leadership team, we would get together another two or three times throughout the year. And yeah, I do think that is sufficient. I’m seeing them every couple of months, and not to mention I’m seeing them every day on Zoom, every day on Slack. We’re able to stay connected for sure.

Timothy Allen: How does that rate against a company that has less of a remote work team, as in the gatherings? Because that still happens in most companies, right? They still meet once a year, often at least.

Nick Francis: I think the office cultures will have little offsites for sure. But I just think it’s so important for remote companies to do well that we’re just so much more focused on it. We had three people on our people team in the first thirty employees, because we were so focused on compensating for team connection. In an office environment you’d have one people ops person, maybe for the first hundred or even hundred and fifty people. So by way of overcompensating for that side of it, we tend to get more value out of these times than an office culture that might take that experience for granted.

Timothy Allen: As far as the second wave thesis goes, tell me a little bit about what you view as actually happening right now in the remote workspace. I’ve heard a couple of people with slightly differing opinions. After Covid, a bit of a push back to work. Now, in somewhat of a grey area, a lot of people deciding, do we go back, do we not? What’s your view from someone from within? What’s happening? Is it on the up again, or has it plateaued, or is it on the decline?

Nick Francis: The data says that work from home has stabilised. The public narrative in the media would say that everybody’s going back to work. That’s actually not true, according to the data from Nick Bloom and some other folks. So it’s stabilised for the last three years. I think it was roughly twenty-seven per cent of work days have been remote and the rest have been in office. So that trend is stabilising. The concept of the second wave is really about, if you want to attract the best talent, which is the number one priority, it’s the whole reason and thesis behind why I built a remote company. So if you want to attract the best talent, the bar is now much higher in order to attract that talent that wants to work remotely but wants to feel connected to their team, wants to actually do their best work in that environment. Being remote friendly, having a document saying you can work from home every once in a while, doesn’t cut it.

Timothy Allen: So in the landscape of normal companies, your Googles, these kinds of companies, what’s the current state of remote work there? Are they insisting people return on the whole, or are they allowing people not to? And what do we see happening as a result?

Nick Francis: I think in most cases it’s two to three days they want you to be in the office. And so teams will end up coordinating and they’ll come into the office and do their thing. But even two to three days, even one day in the office every week, limits geographically the talent that you have access to. So it already breaks the rule where I want to have access to the best talent regardless of geography. So even one day required in the office per week, I get it, it’s great for a team to be able to be together in the office, but you’re creating that geographic perimeter that you can’t go outside of if you want to actually see the true benefit of hiring that sort of talent.

Timothy Allen: So you personally would never even have a hybrid approach to it? If you were to analyse it purely pragmatically, for you, having a completely remote team is the best possible way to do it?

Nick Francis: I think there’s a world in which different hubs can work. I think you’d have to have a big company. Help Scout was about a hundred and fifty people. I don’t think it was big enough to truly make a hub culture work. But for a little while we had an office in Boston, for instance, and a high number of people in our team were concentrated in Boston, myself included. My apartment was too small to work remotely, I had to go in somewhere. So in those cases I think having a hub is not a detractor, as long as the lens through which your operating model works is all remote first. You can still have a hub where people go into the office at their leisure and hang out. And I think a lot of companies are figuring that out. The people leader from Cloudflare was up on stage yesterday talking about this, and they have a hub approach. Distributed is effectively a single hub. I know Stripe does this as well. So they have geographic hubs and then they have a distributed hub, and there are subcultures in each one of those hubs. But there are benefits to having an office that people can go in and work at, that they really enjoy coming into, where they can see some of their teammates and break bread with them, have a meal. But it’s not full on, like you’ve got to be in this geographic area to participate.

Timothy Allen: Interestingly, I think we see that system evolving in the digital nomad world. Hubs around the world, people choosing to go and live in a specific place because there’s a co-working space, because there’s cheap accommodation or whatever. And I suppose that’s a pretty good strategy for a remote team if you’ve got enough employees to create these places.

Nick Francis: That’s what I was saying earlier. I love a great looking office as much as the next person. I think that’s awesome, to go into an office, especially one that belongs to your company and is branded in such a way, and you can actually see your teammates and break bread with them, have a meal. I think all those are really positive things, and they don’t have to detract from the remote first strategy.

Timothy Allen: It’s a good point, actually. That must have quite an effect across the board, not having a head office, so to speak. When you invite new clients to come and see you, how does it actually work in reality when you’re fully remote like that? How do you deal with those kinds of issues? I would call it almost a branding issue, not having a head office.

Nick Francis: Yeah, you can always rent space. So that’s what we would do. We have small team offsites. I’m based in Boulder, Colorado now. I would just rent space, my favourite office space, for the week. We’d set up shop there, have a board meeting, do a team offsite, do all the things. And it felt like a real hub, it felt like a hub for the week. And then that goes away, because frankly we want to pocket that budget and invest it in our team instead of our office.

Timothy Allen: Fair enough. Last question, I’ve been asking everyone this. It’s a rather broad one, but I’d be interested to see how you answer it. What is the future of work?

Nick Francis: I don’t know.

Timothy Allen: You should know.

Nick Francis: Let me think on this one.

Timothy Allen: The future of work. It’s touted a lot, it’s something that’s spoken about a lot in these circles.

Nick Francis: Well, our species loves to pontificate about the future, just generally. I love to live in the here and now and learn from the moment and iterate on a day-to-day basis. But I’ve been doing remote work for twenty plus years and I still feel great about it. I still think that’s the way to attract the best talent, and talent gives you the best chance to be successful and build a great business. So I think it’s both the past, present and the future, at least in terms of my perspective. But can you make the hub thing work? A hundred per cent. Can you make co-located work work? Apple’s a good example, where they really just want to be an office culture, and I think they can totally make that work. At the end of the day it’s about the work culture that you create, and ensuring that people are able to do their best work day in and day out. That is the future of work. Creating and cultivating a workplace where people can do their very best and they feel excited about coming into work every day.

Timothy Allen: Sorry to add one more question on there, but as you were speaking I completely forgot to ask you about something which everyone has mentioned continuously, which is AI.

Nick Francis: Oh, we’ve got to talk about AI. It was going so well, Tim.

Timothy Allen: It was going well. But I do need to hear your opinion on it, because once again I feel like there’s a rather grey area with AI. And you’re the eighth person I’ve interviewed. If I was to take a survey of the other people, fifty per cent are doomers and fifty per cent are, we’re all going to be fine and we’re even more productive. What side of the fence are you on with AI?

Nick Francis: I’m an entrepreneur. I love technology and innovation. I think the tools that are now available to us are beyond extraordinary, and I can’t wait to build products and businesses using these tools. I think I’m an incredible optimist as it pertains to AI. Is it a powerful technology? Do we know exactly how every little thing is going to play out? No. But again, the media narrative is disconnected from the reality that we’re seeing. The media narrative is that jobs are going away. The actual reality is that more software engineers are being hired than ever today.

Timothy Allen: Is that right? Why haven’t I seen that data? Where did you get that data from?

Nick Francis: The media doesn’t like to talk about it. But I think the reality is that companies are hiring. Companies are excited about this innovation and they want to make the most of it. Jensen Huang, the CEO of NVIDIA, said a few weeks ago, and I totally agree with this, if you’re looking at AI purely through a cost-cutting lens, it really lacks ambition and creativity. Those that look at AI and see the opportunity for incredible things are going to realise its full potential.

Timothy Allen: Wow, there you go. You’ve just swayed the rest of the data in favour of people that think it’s a good thing. So really, you had how many of the seven so far, and some people are quite doomerish about it. How many doomers? Four and four. Well, no, three and five now, sorry. But I don’t know, you do hear the doom scenario, I find, from people who are in the AI world. We’re consumers of AI. I assume you would consider yourself a consumer of AI.

Nick Francis: Yeah, I’m building around it too, but I’m certainly not building foundational models.

Timothy Allen: Right. When I speak to people who come from the other side of the fence, they tend to be a little bit more doomery, a little bit more like, do you know what, I don’t know what the future holds, but I’m making these plans. And that includes buying a house in the middle of Boulder, somewhere in the mountains or something.

Nick Francis: So the doomsday scenario, that’s just not something I spend much time thinking about. But I do think about the future of software, and I do think about the future of SaaS, and I think software as a service is going to continue to be the world’s best business model. I don’t think that’s going to change. Will the way that software companies are funded, built and operated evolve drastically? Absolutely. But the fundamental structure of software as a service will continue. Even if you can clone a company or a product like Help Scout tomorrow, good luck. You still don’t know how to scale it and earn trust with customers and build a team that can support it. There’s much more to it than just the build. There’s a reason why ChatGPT and Anthropic and these companies have actually been focused on coding for the last twelve to fifteen months. It’s because it’s binary, it’s black and white. You can build tools, predictive models, which is basically what an LLM is. You can build really great tools to do binary work. But what happens to all the subjective work? What happens to all the other tasks that are done in an organisation? The honest truth is that an LLM is just not going to be nearly as good at operating in that ecosystem relative to coding, which is just maths. So I just think they’re focused on the very best and most impactful use case right now. Some of the other use cases are not going to be so straightforward.

Timothy Allen: Can I ask what your favourite use case is, then, for you in your life?

Nick Francis: Oh my gosh. So I learned to code when the internet was first coming of age, so mid-nineties, and I just thought it was the greatest thing ever. And it feels a lot like the early internet today. So I love building with AI, because whenever I don’t understand what the code is doing, I just ask. I don’t ever run into a blocker as a coder any more. I’m never like, I can’t figure this out. I’ve just got this copilot right next to me, and it’s like, oh, here you go, I figured it out for you, I can help you figure that out. And when it comes to just building and creating products, like little websites for my family or for a side project, I love working with these tools now. So those are the use cases. And there are a lot of day-to-day use cases where I’ve got skills that operate every morning that help me understand my day, give me coaching on how I could show up better on my calls, things like that. But those are little productivity hacks, to be honest. They’re not actually changing my life fundamentally. Some of the other use cases I actually get more excited about, where the quality of my output, whether it’s building a product or designing something, is just better because I’ve got this copilot next to me.

Timothy Allen: I’m surprised you actually talked down the self-reflection part of AI there. I found that super important. I do vibe code websites and stuff like that, but on the whole that’s what I use it for most, to steel man ideas, to ask it about myself, to give me some kind of direction moving forward.

Nick Francis: Well, that’s what I would associate with higher quality of output. So the talk that I just gave this morning, I used AI way more than I’ve used in any previous talk that I’ve ever given. And it’s hard to get the data on this, but I feel like that talk was objectively better than previous talks, because I continue to get this coaching on how the story is landing. And it’s really hard to bring a human being along for every aspect of the journey like that and get that sort of high quality coaching. So I do love that aspect of it.

Timothy Allen: Have you noticed any of the slop creeping into your work, though? Certainly in what I do, it writes show notes, for example, for these kinds of conversations. And I’ve stopped using the AI to write my show notes now, because I can spot it immediately. And I think that undermines the trust that I have with my listeners, actually. Because they’re looking at it, same with when you get a bad email and it’s got telltale signs of an AI.

Nick Francis: I agree. Well, I was on a panel yesterday and one of the questions that came up was, what is AI not going to be able to do, or what shouldn’t you do with AI? And I think it’s the last ten per cent of anything. So the last ten per cent of your podcast involves the show notes, and that’s where you really want the quality and the effort and the craft to show through. So what makes me so excited about this is not the fact that you can clone anything. Anybody can clone anything, who cares? It’s that we’re elevating the craft. People who care deeply and care about quality and craft, this is their moment. Because the last ten per cent is going to be everything. So I’m very excited about this cycle for that specific reason. People that care about craft and quality, the bar has just gone up tenfold.

Timothy Allen: And actually, in my experience, AI is terrible at the last ten per cent anyway.

Nick Francis: Always will be, because it gets subjective. Whenever it gets subjective, these tools are just not nearly as good as everybody thinks they are.

Timothy Allen: Okay, well, great. Thanks for stopping by.

Nick Francis: My pleasure.

Timothy Allen: Thanks for giving me a bit of your time. And yeah, I’ll run into you somewhere else, hopefully.

Nick Francis: Cool. Thanks, Tim.

Notes on this transcript

This transcript was produced from the audio recording of the episode, which contains two separate interviews recorded back to back at the Running Remote conference in Austin. The host’s opening introduction, the sponsor messages, the value-for-value segment and the conference announcement have been left out, along with the sponsor message that sits between the two interviews and the closing housekeeping remarks. The text therefore begins where the first conversation begins and ends where the second one ends.

Speaker names are shown in bold. Three people appear: host Timothy Allen, who conducted both interviews, Liam Martin in the first, and Nick Francis in the second. The two guests were interviewed separately and did not hear one another.

The transcript has been lightly edited for readability. Filler words, repeated false starts and verbal tics have been removed, and a small number of overlapping or fragmented exchanges have been smoothed into complete sentences. The wording, meaning, order and character of both conversations have been preserved throughout, including the disagreements between host and guests and between the two guests’ positions.

Names, products and specialist terms have been rendered in their standard spellings. Among those that appear are: Time Doctor, Running Remote and Peppermint; Claude Design, Claude Code and Codex; Google Slides, Slack and Zapier; HeyGen; Anthropic and OpenAI; Waymo and Tesla; the Model T; Help Scout; the economist Nick Bloom; Cloudflare and Stripe; Boulder, Colorado, and Boston; Jensen Huang of NVIDIA; and ChatGPT. Figures, dates, predictions and characterisations throughout are those of the speakers.