#61: Siqi Chen — He’s here to stop your startup from going broke

I saw my friend's bank account after he raised a couple million and it was pretty crazy seeing a few million in there. And my first thought was like, how do you manage that?

Right?

So that's, I guess, where you sort of come in, right? So what is Runway? How do startups manage their Runway? How were they doing it before and how do they do it now?

When you say your friend's bank account, do you mean his company bank account or?

Yeah, yeah. His, his company's bank account. Don't worry. His, his Mercury account. I guess that was a little A little misleading, but I was at his office the other day and we're young and it was his first time raising money. And he just was like, this is weird. This is a weird feeling for like a 23, 24-year-old. Yeah, it is.

I remember the first time I raised half a million dollars and seeing that even 6 digits in the bank account was weird. And I remember before I raised a round, I was talking to Experience Founder. And he asked me how much I want to raise. Like, I don't know, maybe a quarter million dollars. And that seemed like an infinite amount of money to me at a time. And he was like, what are you going to do with a quarter million dollars? I'm like, what do you mean?

Exactly.

Yeah. So where do we come in? We actually don't come in that early. So, you know, we actually started with a product that helped a very early stage founder understand their burn and their expenses. How Much Runway They Have, which is the name of the company. And what we quickly discovered is that, yes, you want to understand your spend, but at the earliest of stages, your, your spend is actually quite simple. All of your money pretty much goes to people. And from month to month, that doesn't really change that much. And to the extent it changes, it's pretty easy to keep track in a spreadsheet, which is what I recommend most people do when they start. And so much social do, do they have a spreadsheet model where Runway becomes a lot more helpful and why we built it is As a company gets larger and it has revenue, the business becomes more complex. You have many people in the company and many departments and a lot more money going through. That complexity and the impact of the decisions you have to make in your business, just so much higher, right? You know, instead of like making a decision about whether you want to buy more soda for the office, you're making decisions like, should I invest like $10 million into this like new product line? Or should I hire, you know, double the size of the sales team? And what is implications of that? These decisions are actually really complicated because businesses are really complicated and they get more complicated as they get large. And so you need increasingly better tools for thought to think through the consequences of those decisions. And that's what FP&A is about. Like, probably heard the term FP&A and it sounds mysterious. And people, a lot of founders like me, you know, in the past I haven't had a background in finance, this idea that finance is like accounting, but that's like sort of, if you think about accounting, that's sort of the database layer of finance, right? Like, you know, you write code, you have a database, that's the data. And FP&A is actually the logic layer of finance. It's basically a simulation of business and lets you understand where you're going. And that's what Runway is more about. It's making that logic understandable and useful to people.

Gotcha. So I've tried doing that kind of thing by myself where on a much smaller level, right? If I invest X in advertising, here's what, you know, maybe my, my little SaaS or my newsletter will do. But this is on a much larger scale, right?

Yeah. Yeah. So at the simplest levels, like that is a problem that we solve, right? And at, you know, our smallest customer, we have about 50 employees or so, right?

right?

So some of the companies running around we are AngelList and Superhuman and ConvertKit and, um, uh, a Jenner, Kylie Jenner's tequila company. Uh, so those are the types of companies that we're helpful with, um, helpful to. And yeah, at that complexity, because of the sheer number of people and the sheer complexity of the business, right? Uh, it becomes dimensionally highly dimensional. What I mean is like, when you have one customer segment and one product, it's relatively straightforward, but it gets complicated when you have different geographies and different countries and different product segments, different departments, and you have to track all of the intersections of all of those things separately. That's where models get really, really complex. We tried to make a very good version of a business planning modeling and planning tool so that anyone can just come in, make changes, play with it, understand it.

Got you. Got you. Yeah, I was really stoked by the kind of variability and how you could change stuff watching the launch video. Yeah. So most of my listeners are marketers, young founders, so probably not as deep, you know, along that journey. So I want to, I want to jump into the launch video. Your, your launch kind of blew up the internet, which is always the goal. I was, you know, hundreds of thousands to millions of people saw it and it was, it was just really well done. And that's why I really wanted to chat as well, is like from idea to execution of the launch, could you take me through that process of, of how you went about it?

Yeah, we knew we wanted to do some kind of coming out party. We, we, we saw it as a marketing campaign, right? The reality is we've actually been live for, depending on how you count, either 4, 3, or 2 years. Um, we've had a V1, um, we've always had customers and users. We've always talked to them and, you You know, I definitely, it would be crazy to build any kind of product for 4 years, just like not have paying customers or users. So we've had that the entire time, but we never, it was never at a point where we said, oh, this is like good enough for us to put a 1.0 sticker on it until fairly recently. And so we treated it as a marketing campaign. We knew that it was like, this is our early access sticker. And so the timing was at our discretion. So The process of planning this, I think end to end, it was probably around a 6-month process. I think most of the work happened in the last 3 months or so. But really the first 3 months was refining our company narrative. And so a lot of people, everyone wants to talk about like, what was the idea and what was the process of coming with these cool ideas? But I think what's way more important are figuring out is figuring out one, what your company is actually about and what your company values in terms of its brand and how you operate and execute on that brand. Like what, what do you value deeply that you refuse to compromise? And that makes you unique in how much you're willing to give up to make it happen. I think everything else comes naturally from those two things. I think if looking for like the process and the ideas is kind of like cargo culting, you're gonna get the surface of it, but you're not gonna get the deeper level of it. And so thing that everything starts with is actually the story. And when you look at the greatest companies and the greatest stories, like you think about Disney, it's a very valuable corporation. They start with a story. And the thing I think about story is that Lulu Chung wrote an amazing article about story and Anderill, right? And the 10 Commandments really crystallizes the way I think story should be a part of a company. And my main takeaway for a founder is that when I was, you know, a much younger founder, my first companies, right? Like your instinct is I need to build a product and I need to sell it to a customer and I need to make money. And this is true. That is the essence of a business, right? Like, and so, but So the implication of that in terms of mental model is that where does story fit in? Story is one of the things that you do in your company is how you think about it. Right. It's like, all right, I built this product and I talked to some customers and I launched this thing. Now I've got to market it or raise some money. Now I got to figure out a story. Right. And I think like that is a pretty, you know, it's a pretty natural way of thinking about like how you should do it, but. That ends up being an afterthought, but a correct way, I think, of looking at a story is the story is a purpose and reason for why you're doing what you're doing as a founder, why the company exists. And so it starts with the founder and it goes out in a series of concentric circles. It goes out to the leadership team, the wider team, to your customers, and to the entire world. And when you think about a story that way, that's how you get a cohesive story where it doesn't feel like marketing was an afterthought. The messaging was an afterthought, right? Where it doesn't fit with like the quality of the product, how you talk about it, what the product even does. A good story should affect who you hire and your values as a company and what you're building. It comes from the story. And, you know, not having that is how you end up with like an Enron where famously one of the values was integrity, right? Like there is no cohesion, right? If they, if the story was like, actually we just wanna make a fuck ton of money, you know, create energy. Like, then we would have had a cookies story, but that wasn't their story because that's not how they thought about it.

Yeah. Yeah. The storyteller rules the world. I love that. Um, I don't know if you got a chance to read Jeremy Stern's, uh, article on, on Palmer last week that, uh, went super viral. Fantastic piece and story about him and his hero's journey. Right. And I think every founder kind of has a hero's journey. Yeah. This, you know, Joseph Campbell idea of you know, a story in Greek mythology, you meet the allies, you meet the mentors, you know, don't want to go on the journey, you get convinced, you go, you have all these like moments, right? And so I've been ghostwriting for founders, VCs for the last few years. And that's been my main thing is telling those stories, the heroic stories, like every founder is a hero, like every company is a myth in some ways. And, and yeah, I think the strongest companies start with that heroic story. And, you know, it's creators too. And when I tell my stories, you know, the strongest ones are my personal stories, right? Or when I write a tweet about the power of cold DMs, right? Like I have one that's doing well right now. I only say that because they changed my life, right? Like, and I say that underneath, like I'm not saying it just because I'm saying it and that's like, you know, my brand. I'm just saying that because, changed my life. Like, yeah, I think everything starts with the story.

It's funny, we only have one advisor to the company and it's a storytelling advisor.

Interesting. Say more on that. Who's the storytelling advisor, if you could say, or what do they advise you on exactly?

Our story. So we were like for 3 or 4 months refreshing our story and trying to get to the the essence of it. And so he's a magical person. His name is Wolfgang Hammer, first of all.

That's a magical name. That sounds like some Greek mythology stuff.

Like, I hope he named himself because he must have the best parents in the world. Um, yeah, but he was executive producer of House of Cards and Hunger Games, and he was a studio head for Lionsgate and Miramax. So he really knows stories. It's his like entire thing and difficulty and the depth and the ultimate outcome and how obvious it is only in retrospect just floored me during this process. Like I thought I was a pretty decent storyteller, but I'm also impatient. And so, you know, when, yeah, and founders tend to be, right? You work on something, you wanna get it done yesterday. And out of all the things you work on, story just feels like a bullshit string of words that, you know, you're writing code, right? Making code and designing a good product, that takes time. Like when you think about writing, you know, the natural inclination is like, just write the words. Like it'll, you know, it'll compile. Like you don't have to compile it. It'll just put on the page and press deploy and the words are done. Right. That's definitely even as much as value storytelling, how I thought about it. It's like, just get it done. But during this process, like I was really blown away by like how deep the art goes and how powerful it can be. Like we refresh our story and the impact it's had on this company is indescribable. So the story we told in our launch video. But it was also affected every word that we have on our marketing site, how we sell the product, but also what the product does. It affected how we hire, what, why people were here to do what they were doing, what features we prioritize over other ones, because finding the essence of it, like defined who we are, what makes us different, and a product needs to reflect that. So all of that came together and that's what happens when you have the right answer and it's deceptively difficult to get to.

I'm curious, what did Wolfgang do? What was his process? He, I'm sure, talked to you, learned about the company as much as possible, and then did what exactly?

A lot of it is just one-on-one and trying to pull out— great story, especially in a startup, is like personal. And trying to pull out like why I'm doing this and what makes me the right person to do this, combined with, you know, how we're telling the story. Today. And then trying to get an essence of what makes us different and what is the thing that you want to end up with, something that is both obvious. You're not creating anything new. That's the one. So there's a few limits test for like what you're looking for. You don't feel like you're saying something new. Once you have it, you just start saying it automatically. You don't have to memorize it. You already have it, right? But then when, when you use it, it feels like deeply right and highly clarifying for almost everything that you encounter in your company, uh, was what was advertised to me before we got it. And I was like, okay, maybe it'll take 15 minutes or so, let's do it. Um, but, but the process is, you know, he has some very powerful frameworks around storytelling. I'm not going to do it justice, but, um, my main takeaway is that A great story isn't just about emotion and plot. The greatest stories have a deeper layer, a philosophical layer. Some stories have, as a deep— the, on the deepest layer, a conflict between worldviews, right? If you think about like Star Wars, you know, it's a worldview of control versus freedom, of rebellion versus empire. And great stories play out this conflict in worldviews through the eyes of a single character. Um, and in the worldview conflict is usually be between what he calls a global dominant worldview and a global underdog worldview. And so the global dominant worldview is the world as it is, the default state of world, what everyone tends to believe. And the global underdog worldview is a worldview that says there is a better world to be had if everyone can just like see the subtle lies in the dominant worldview. And the key is in every story, the global dominant worldview is always a worldview held by the bad guy. And the underdog worldview is the protagonist's worldview. Worldview and this conflict plays out through the eyes of protagonist is the story. And so that seems like, you know, at a high level, pretty basic, but what's deceptively difficult and my main takeaway in like how I think about getting to the right story is getting the global dominant worldview right. And that is so difficult to do because as a founder, you're, you're, so I'll, I'll give you an example of like what our, like what the school of thought and worldview we effectively started with was at Runway, which is you have— finance is like really complicated to people, right? And that's why you have these people who are super well trained in finance and you have like amazing finance tools to help them do their work. And everyone else can just be a cog in a machine and don't have to worry about business. They can like just focus on their coding jobs or design jobs because it's too much for them. Right. That's a global dominant worldview that we used to have. And our old global underdog worldview was that no, actually finance is super important for everyone to understand because that is important for you to understand the impact for your work, to make it more fulfilling, for you to have more agency and control in your work. And the problem, the reason why you don't understand this stuff is because actually the tools aren't good enough for you to get access to it, for you to, for it to make understandable. And if you can have better tools, then we have a better world because everyone understands finance and has more control of their lives and companies can make better decisions as a result. I think that's a pretty decent story actually. It wasn't good enough. And actually in retrospect, it wasn't even good at all. And, and, but what I will observe is like, that story seems fine. And it seems fine because that is probably like as good as any other startup story. Right. And the reason why it sounds like most startup stories is when you're doing startup, it's so easy to caricature what exists because you need to have an, an event, a reason almost for you to exist. And that's basically what I did. And I think that's what 99% of startup stories do is they have this caricature of the world as it is because. When we, well, and what happened is when we really thought deeper about it and when we really asked ourselves, how can we be as absolutely honest and give as much credit and, and, and faith to grace to what is out there already in the world? Are we doing that? And we asked ourselves this question. We, we realized that we weren't, because if you think about mega corp, like Salesforce or Microsoft or whoever, or Google, Who's saying be a cog in a machine? No one's saying that. People are saying the opposite, right? People are saying like, we should be more collaborative. People should understand how everything works. We should be more transparent. And then we looked at like the incumbent products, like an Anaplan and Adaptive Insights, right? For finance. And I look, we look at what their marketing message is. It's connected planning, it's strategic finance. It's connecting all the different departments. They don't actually do those things, but that's certainly their marketing message. And so what we realized then is that, oh my God, like, because we weren't honest about what the global dominant is, we ended up with the same message as everyone else almost without knowing. And we thought it was a great story. And when you have the same message as everyone else, and that's how you end up there, then your only differentiator is, but we do it for realsies. Which is not a convincing story by anyone. So we had to really dig deeper and figure out like, what is a starting point where we have a believable villain, right? You think about movies, right? The reason why Breaking Bad and Thanos are like good villains is because they have believable motivations. They're not like cackling, you know, mustache twirling, like evil caricature. And a believable villain then makes the heroes and the unfolding events of the story believable. So we spent a lot of time trying to figure out what the global dominant is for runway. And we got that. And that's how we got to our current, like, story that has been pretty impactful. Wow. I love that.

I love the idea. Like, I think every founder, or maybe, maybe it's especially creators and writers. It's nice to have an archenemy, you know, even if they're somebody famous, like somebody that you are trying to beat, an enemy, or even just an idea that you could put out there, right? Like, My enemy for a while was like AI writing, right? It still is. I don't use AI writing, but I love shitting on it. And, you know, my audience resonates with that, right? They want that real authentic, uh, goofy cursing memes. Like my enemy is also very boring marketing, right? It's, it's the crappy marketing, the also paid ads. A lot of the time paid ads work, but like I focus on organic and I try to try to push that. So I think. Yeah, the more you know yourself and your story, the more clear you are and stand out. It took me a while too to realize what kind of my story was. I love that though. Global dominant. I always think of it as just like, you know, the, it's a David and Goliath story at the end of the day. It's a biblical story as old as time, right? Like that's, that's all it is. I was reading that Palmer piece on, on VR and, you know, his enemy, right? For a while, was, is now, or still is, Zuck, um, I guess, uh, for better or worse. Um, and, and you, uh, you started a VR company yourself, uh, Sandbox?

No, no, I was an early investor in a company and I joined.

Oh, early investor, my bad.

I was CEO for a brief time, but I wasn't a founder. No, Steve Zao was a founder.

Okay, what's kind of your thoughts on VR right now? Um, VR, uh, Vision Pro, anything like that?

Yeah, well, I joined Sandbox because I've been a VR superfan. I've owned every developer kit, and I thought it was going to be super interesting. It just took quite a while for it to be interesting. Arguably, it's not even interesting yet, but I think it will be. And the reason I want to join Sandbox is when I really thought about the experience that you want in VR, people imagine like this, the Star Trek holodeck. And what you get even today is either a seated experience at home that's kind of disembodied, or you can move about in a relatively small space because most people don't have that much room. And the kinds of experience you can have is like fairly limited. You can like fly an airplane, get in a spaceship, you know, pilot a mech. But otherwise, like, you know, like playing a real first-person shooter You know, it doesn't really make sense. The locomotion, like Gorilla Tag is the biggest game on VR and it's like a very specific way of moving. So yeah, you can be creative, but it's like still fairly constraining creatively. When I saw Sandbox, I thought it was really interesting because combined motion capture, you're on this real-time motion capture stage of your entire body tracked with VR and it's in a retail environment. It's not in your home, it's out of home. So you can, you have like a fairly large room, like about the size of a full empty living room or two in most apartments. And I thought that is an experience that's way closer to what people imagine VR should be. And if I were to project out the change of technology, even as all the technology gets better, you know, headsets get better and resolution gets better and tracking gets better, What's not gonna change is locomotion. Like the Ready Player One, like treadmill that violates lots of physics. That's not a thing that exists. You know, there's like Disney's HoloTiles also is like, you know, if you watch someone actually go on it, it basically causes you to trip or like get motion sick. So it's not a thing. And so you really need to have wide open space for that to work. And I thought, okay, like real estate isn't gonna get any cheaper. So this is a sustainable experience and it's gonna become more sustainable as technology improves. And that was my thesis for investing my own personal time in joining the company. And I think that's playing out, you know, there was a, you know, it almost died during COVID but now it's like, you know, on track for like $100 million annual revenue and, you know, 1,000 employees. So it's on an amazing path today. But at the time, you know, it went from like 400 employees to 10 when I was running it during COVID But yeah, like form factor experience is how I think about this. I do have a Vision Pro and like, it's just really sad to see that it's become a new Google Glass. You know, it's become this punchline. You know, it's like, who uses this thing? I love my Vision Pro, man. I use it like almost every day. Wow.

What do you use it for?

I use it as a giant monitor.

Has it improved?

From anywhere.

Yeah. So I got the Vision Pro, but I found using it as a monitor, uh, was like, there were just enough errors where it was very annoying. Um, like it was, uh, with typing especially, I guess maybe that's what it was.

Uh, but yeah, 1.3 will pass through the keyboard visually so you can see it.

Okay.

I don't, is it, is it because you can't see the keyboard?

Uh, it was just with like clicking and typing, it would be like just enough off where it like get me out of a flow state. It'll almost be like annoying.

Well, you don't use a real keyboard?

I need to try that again. I mean, I've hooked it up with my laptop.

Yeah. But you use your laptop keyboard, right?

Yeah, I've done that. Yeah.

Huh. Yeah, I haven't had any issues with typing, so that's kind of weird.

Interesting. Interesting. Well, let's keep talking investing. You invested your time in Sandbox because, yeah, that very cool thesis of real estate is just not getting any cheaper and people want somewhere to hang out and do fun stuff. And also, you know, the running around with VR. What about investing in things like Mercury, ElevenLabs? Those are both pretty awesome companies that I'm fans of. What was kind of your thought process around those?

Those weren't difficult.

They weren't difficult decisions. No, no.

I mean, I'm just, I was lucky to get any allocation in all of those deals.

Yeah.

What was it? I think investing helps you be a better founder and being a founder helps you be a better investor within some constraints. Part of like what you have to believe in order to be a founder is there are these things that seem obviously bad in the world and for some unreasonable reason, you think you can actually change it and make it better. And I think that is somewhat of a rare, like, trait. Maybe the, like, there's a few things there. One is like, you have to look at something and like say, this sucks. And I think that's not that uncommon. And then you have to imagine a way it can be better. And then you have to like think that you can do it. And then you actually have to go and start doing it. That funnel gets pretty small. Um, if you, get down there. But the reason why I invested in Mercury is I used banks as a founder and they were terrible. Like, this is not what I want. You know, it's not a good experience. It's annoying. And you can build something way better that people find pleasant. And when I think about it too, this is interesting. Like, you know, occasionally I talk to investors who want to learn about Runway. And for a long time, I shied away from this answer. And I think a lot of founders have learned to do this. And the question is like, what, what makes it different? Like, like, well, how are you solving this problem differently?

Right?

Because for our industry, it's been around for quite a while. And the answer I shied away from is it's a better design product. And like I felt insecure and sheepish in saying that's the answer, right? Like what's Mercury? Mercury is a better designed product.

Yeah.

That's what it is. And it's incredibly valuable because of that. And as investors, what I noticed, like people are looking for like, what is like the business model strategy, enterprise go-to-market synergy, whatever hack it is that's gonna like help you win. And like, I don't, I mean, the complete discount, like those are real things, but like the most obvious thing that doesn't get attention is like, what is the product? And is it good? Is it just really good? And how do you make it good? And my deep feeling in interacting with a lot of investors is that design is still so undervalued in a world where everyone has iPhones and Apple is like the most valuable company in the world. It's still undervalued. Which is like still really shocking to me, but I feel like that's such a huge arbitrage opportunity even today. Whether you look at Linear or Figma or Notion and, you know, what we aspire Runway to be, like there's a lot of low-hanging fruit there. So yeah, I love that.

I love that arbitrage opportunity. I'm, I work with jam.dev, been working with them for about a year and a half. Incredible design. And, you know, proud to, proud to have a little bit of equity there and love that company. And like, it's, you know, you just stand out so much more. People have seen it with my branding, you know, everything is this purple and blue cyberpunk design from my book to the jersey, to my profile picture, to all my newsletter covers. You know, it's, uh, it's a lot of time and craft went into this. And I think as AI and no-code products get easier and easier to make, The, the real craftsmen will stand out more and more. The people who really care about the software that they're building and the users that they're building for and just want to build something delightful. Uh, I think it's why like Arc Browser, personally, I don't use it. I find it very confusing, but it is a beautiful piece of software and I can agree with that. Like it is beautiful. There's a lot of delightful features. It's delightful. I will say that, but very confusing. It's a vibe. They've, they've achieved a vibe. They have 100% achieved a vibe.

I used it 4 years ago and I churned after a month or two.

Yeah.

I started using it about a year and a half ago. I stuck with it. Yeah.

Should I go back?

It's interesting. I don't love it, but it's, I don't not like it enough to want to switch back to Chrome. It does solve some real problems for me in terms of workspaces. But it's also just still weird enough where I'm like, this is not quite doesn't quite feel good.

Yeah. I really respect what they tried to do of creating new words and like phrases like easel and like all this kind of stuff. I was actually, I'm actually reviewing Arc for, for Product Hunt, uh, this week for like a SaaS review we're doing. And I was like, oh, I love it. Like creating phrases and slogans and all this. Like they really thought through from the ground up, but what that does is it actually like creates a very steep learning curve, uh, for a new user. And like, you kind of got to meet users where they are sometimes. Like, they give you way too much. Uh, yeah, it's kind of frustrating and confusing. It is.

It's a little too maybe opinionated for me. Um, yeah, a little too like indulgent. But you know, like, people disagree. I mean, they are like, for a new browser, massively successful. That's not easy to get the kind of distribution attention they have. I think their like greatest strength is storytelling, right? Like they have like almost a million followers on a YouTube channel for a product, you know, which is like wild. And yeah, I, I, that strikes me too. Like we can, you know, critic, critique the products and that's totally fine and fair to do. I have like a lot of gripes with it, but I think, you know, my takeaway with, from that is that reminds me of like Runway too, right? Like when we started on this current version of Runway, we started this maybe like 3 years ago, we kind of reset on, started on the, a new version of it. We start, you know, version 1, as I mentioned, it's for like early stage founders and V2 was like, okay, we need people who with like more significant pain who's willing to pay us like $2,000 a month, which is not a founder because they have very different problems. And the thing that's most memorable is one, it was really, really painful and took a long time for us to get any traction at all to just like build a complicated product because, you know, it's valuable because it's complicated. And those are hard to build. And so for a long time, it wasn't very good and it didn't do much. What blew my mind is when we were working with these design partners, very busy people with real problems, and they were using the piece of shit that Runway was at the very start. It was not good. And, you know, they would have these bad experiences and they would say, it's okay though, we so believe in what you're doing. And like, I think that in some ways is a lot of like what Arc has captured at scale. It's like people are willing to put up with a lot of things if they think they're, they support and they feel good about joining your movement and why you're doing it.

Yeah, yeah, I agree with that. I think that goes for creators, founders, whatever it might be. Like, you don't have to be perfect, but if you nail the story and provide some, you know, provide some value and are genuinely you, like, I screw up all the time. I delete tweets all the time. People still, you know, love my stuff, hopefully. And I've screwed up on my blogs, you know, errors here and there. And yeah, like, I think people are very forgiving if they could tell that you're genuine and are really trying your absolute hardest to build something awesome for them.

Yes. I think that is a pretty undervalued part of storytelling, but also marketing, right? Our website, people like often comment on like how memorable it is. And I think one of the big inspirations and through lines was our belief that finance people and executives are people too. And if you look at how people try to sell these people, like, you know, it's basically trying to sell like a generic guy in a suit, right? So you have a website with generic dudes in suits shaking hands with each other. You don't show the product and it's like, it's business. And I think what's underutilized is like, they're people too. You can talk to them like people and it's okay to speak to them like people and people want to be spoken to them like to them like they're a person. And yeah, that, that is probably like one of the big sort of brand principles we have, um, as a company in terms of how we market and what a website looks like and how we write and all those things.

Yeah, I was just about to tweet this earlier. I have it in my drafts. I wanted to edit a little bit, but if you assume your customers are smart, you're going to make smart content, smart products. If you assume they're idiots, you're going to make stupid stuff. So always write and assume people are smarter and will figure it out and understand and are smart people with good taste. And it pushes you further. Yeah, I really try to do that with my pieces and hopefully it pays off. Some other questions I got for you around investing as well, if that's cool.

Sure.

What, what, uh, industries or niches are you particularly bullish on right now and, and, uh, looking to invest more in?

Man, it's changing so quickly, and I, I try to not have industry or niche theses. And the reason why is an exceptional company is always like different and unique and unexpected. So generally, like If you have some preconceived notion of here is a thing that's going to work, that's going to blow up, I don't think you're going to capture the outliers. So I'm looking for unexpected and weird things. I think Wordware was pretty unexpected and interesting where there's a lot of companies that are making AI more accessible by helping you do visual programming. But they're like, why don't we just use LLMs and make a document the program? You could just describe what a program does and then we chain LLMs together. I thought that was like particularly interesting. What else has been a surprising thing? Yeah, that was a surprise. Like GoodFire is a fairly recent investment as well. They just announced funding, but you're familiar with GoldenGate Claude from Anthropic?

I'm familiar with Claude. What's, what do you mean GoldenGate Claude?

Okay. So Anthropic did some research. Around mechanistic interpretability. So it's based on earlier research where OpenAI used GPT-3 or 4 to label every single neuron of GPT-2. And it was able to correlate what each clusters of neurons correspond to different concepts like lying or self-preservation or the concept of a Golden Gate Bridge. And so what was really interesting is Anthropic took that research a step further and did that for Cloud. But what they did is they released a version of Cloud for like a couple of days where it's the same Cloud, but they jacked up the dial for the one cluster that corresponded to the Golden Gate Bridge. And it was crazy.

So, yes, I remember that on Twitter. That was so weird.

It was so weird, right? It's just like, and it was kind of disturbing too. So you ask it to like, give me a, a recipe for baked chocolate chip cookies, right? And it's like, okay, like take 2 red beans of steel, melt them and put them over the water. And then you get the Golden Gate Bridge constructed in like 19— and then it would record, it's like, oh wait, that's not a chocolate chip cookie. Nevermind. Okay. You take some flour and then you put it into the Bay Area and then you like construct. So. What GoodFire does is that, but for all the open source LLMs. You can just uncover the dials and move them around, which is really cool. You can now say, "I want to create the Star Wars cloud." And it just can't help but be a Star Wars character. It's extremely fine-grained. It's not even fine-tuning. It's discovering the correlations and just being able to dial it up. Which is a very different way of controlling LLMs than prompting. I thought that was particularly interesting. But yeah, those things are what would have been things that I would have guessed that someone would have worked on. But when it's weird enough and new enough, it gets super interesting.

Yeah. Have you been using Claude a lot? I can't believe how good Sonnet is.

It's so cool. It is very good. Sonnet is very good. I still have ChatGPT open. Out of habit, but yeah, Sonnet is clearly better.

Their marketing also has just been incredible. I think I'm so impressed. Like OpenAI, like I love, you know, Altman. I love that whole crew, like Roon and whatnot. And it feels almost a little clinical, whereas like Anthropic and Claude feels like your grandpa that you're asking questions to. Like I think they, it's just so well designed. Like you said, like design, it's super undervalued.

They do a really good job. There's no question. I mean, I think that is the zeitgeist of like, you know, and OpenAI is like falling off a bit and Anthropic is like outperforming, is the best foundation model and is outperforming their product. I think there's a good amount of truth to that. I do think it is like unfair to OpenAI. Only because I think, like, in terms of foundation capabilities, I think they're— my sense is that internally they're, like, considerably ahead. It's just very hard to deploy it safely.

Yeah, yeah, I, I could definitely see that as well. I feel like there's so much fear around that now, and people are so hard on them. Um, yeah, I feel like they definitely are gonna have some Sonnet-like feature, very Or they probably already have it internally, like you said, where like, yeah, I mean, like if Sonnet has Sonnet 3.5, right?

Like, yeah, there's no way. I mean, to be fair, like Sonnet, the existence of Sonnet 3.5 implies the existence of OpenAI 3.5, which is like, okay, they already have another thing that is going to leapfrog a thing if OpenAI decides to just like incrementally beat 3.5. But like my take is if they have OpenAI 3.5, OpenAI definitely has something that'll be that. Like, they have it. They just, it's just very difficult to release it.

Yeah, top secret right now. How do you kind of plan out a roadmap when you're building out Runway? Like, do you have any tips for that for young product builders?

I mean, I really think the actual most important thing is to build in a way that makes sense to you. You should read a lot about what are best practices. So, so, okay. So there are just foundational things that you've heard advice about. And the thing about advice that I think about a lot, one school is Paul Buckheit's school, which is advice is limited life experiences plus overgeneralization, overgeneralizations equals advice. That's certainly a true perspective, but the thing I think about a lot is, you know, there's advice you hear and it's like, holy shit, that is like so insightful and clarity creating, right? And then there's other advice that you hear and you hear it over and over again from everyone and it's like so obvious. I think the second category of advice is the more important category of advice.

Yeah.

And the reason why is because, not because like everyone says it. Well, kind of. The fact that everyone says it and everyone knows it and it seems obvious and yet it remains advice implies that it must be deceptive, deceptively difficult to actually execute on.

Yeah. Talk to users.

You wouldn't make it advice.

Yeah.

So the foundational advice of talking to users is very good. Like you should. Do that, but like how and with what mentality and what are you trying to learn from those users? That those subtleties are actually fairly difficult to get right. But with the baseline of talking to your customers and being flexible in the path, I generally think you should like execute in a way that's like most comfortable to you. Because the truth is like outside of talking to customers, I really think that's like the one thing you don't want to change. Every piece of best practice, the opposite of the best practice is also best practice.

Yeah.

Yeah. Just build these. You might as well do what works for you.

Yeah. And what about like kind of managing engineers pushing, you know, product direction? For reference, I'm working on a, basically like a meme SaaS right now. So it's Google Trends for memes. I've got, you know, a couple engineers scraping. Internet for new viral memes, and then I'm alerting people. Have, I guess, 4 figures MRR, just hit 1,000 MRR for this. Um, and so I'm like, it's my first time managing engineers and I'm trying to communicate and it's, you know, it's, uh, and trying to like plan out a roadmap. I don't know how to do the timing. I don't know how to like, you know, what, what, uh, any advice there around that is like, you know, Getting to—

How many engineers?

So there it's a, it's 2 engineers, kind of a no-code agency, but it's 2 engineers on the team, I think.

So I would not really worry too much about like spending a lot of time on planning. I mean, 2, you just get in the same Slack room and adjust in real time. I think that's the fastest way to do it is when a team gets larger. Communication complexity between the people becomes more complex. But the real turning point is when you as a single person can no longer follow everything that's happening at all times because the sheer amount of people. That's when it gets really difficult because it's not just that you don't have the context, but more that in that world, the decisions that every person has to make without you multiplies. Significantly, and they're going to make decisions by default very different from you. And that's where things get really messy. And so a lot of this is about alignment and holy shit, it's so hard. Like there's— I am— there is an answer, but the way to get that answer just requires a significant amount of experience and time is honest truth to this. What it deeply comes down to is self-awareness. And you can think you're self-aware and you can still be more self-aware. So one of, and I 120% include myself in this, right? Like I'm gonna litmus test for alignment and self-awareness around this is everyone has experienced this. You're working with somebody, one of your engineers, and some idea or thing was said or proposed and you're taken aback by like, what the fuck? Like, how could, how could you think that? Like the, the, the right answer is obviously the opposite of what was just said. So then you're like, I must have deeply fucked up and we're now not aligned. And you're thinking, I've not done my job in like communicating, or this person's actually just too stupid to understand it, or you're just so different that I can't work with them. Have you experienced that before?

Yes, uh, definitely.

Okay. So everyone has, right? And that used to be my default reaction. And I think like one of the things that has really helped me as I've gotten older is the increasing understanding that the way, particular way in which I think and my tastes and preferences is unique and rare actually, because your default, you know, my default when I'm younger is like, everyone thinks like me. You know, everyone, this, you know, this is obviously good and that's obviously bad. And like, of course it would be. And accepting that that's actually not the case. And that's okay because that is just a reality of the world. Means that when you encounter that situation where it's like, there's just such this, this is a really obvious thing that there's a disagreement. What it means is that there's The reason why you feel that way is intuitive to you. It's like super deeply ingrained in how you think and who you are. And the implication of understanding that is incredibly important because if you know that a reason why you feel differently about a thing is something that you so deeply believe and so intuition-driven, what that almost certainly implies is that you didn't think it was necessary to make it explicit to people. Like, and, and because it's so deeply ingrained and you haven't had the practice of making explicit to people, it can be sometimes even hard to articulate in words because you didn't realize you ever needed to. And that realization was like a big unlock for me. It's like, now I'm going from this person may be an idiot to, oh, actually, I was not consciously aware of how deeply my preferences go. And I reflect on, I was like, I actually never talked about this. And it just seems so obvious. I can't imagine why I would even need to talk about this, but now I know I need to talk about this. And then when I try to talk about it, I'm like, fuck, how do I even like describe, how do I even find the words to, it just, it just, of course. And then I have to find the words. And once you have that, people look at it and it's like, oh, I get you. I, I know what you want.

Yeah. Yeah. I, find it so hard to do that with taste. It's very hard to scale taste. At least it's, it's been a challenge for me. Like I thought about starting a ghostwriting agency many times and, you know, I had a couple freelancers working for me, but scaling my taste in writing was so hard that I, I, I was like, I, I can't do this. Like I, I'm, I'm, I'm just not doing it. So instead, just continuing to work with solo and then have a couple employees working with me on the side on other things. But for writing specifically, my taste is just like my taste and people are paying me for my taste. And that is—

I get you.

Yeah.

I work with a ghostwriter for LinkedIn and, you know, she's in-house and she also does our help center and a bunch of other content, but also ghostwrites for me. And, you know, you probably have the same like procedure where You know, we talk, we talk to each other and then we extract out things that could be blog posts and like getting my tastes offloaded is ongoing process. And it's quite difficult to do. She's an incredible writer. Yeah. You know, she's like, you know, quantum physicist, you know, founder and LVP and writes content for us. Right. Like incredible. Very difficult. The thing I'll say about that is like, it 100% is difficult. And. I think you kind of need to do both. There is no choice. If you're small enough, or if you just want to be one person, then you don't have to figure it out. Right. But as you scale, if you don't figure it out, then you can only be in so many places. And like Brian Chesky does as well. And if you're personally involved, that's great. And you want to be as personally involved like Zuck and Chesky as much as possible. But even then, practically you can't be involved in everything still. And so it becomes incredibly valuable to be able to make it explicit. And what I found is that working with my own ghostwriter in breaking down the specifics of my taste has actually not made the writing just better, but made me more self-aware. Of what makes me different and how I think and what makes it good and bad. And so it's actually incredibly rewarding and it is no doubt super difficult, but it's as an ongoing process, it is super worth it.

Yeah. Yeah. It's a very cool process, I will say. So I've been ghostwriting for Danny, um, Danny Grant, uh, over at Jamf for the last 18 months. And at first, you know, it's— you start and it's very hard for a writer to capture the voice. But as you start talking more, like, I could hear Dani's voice, her tone, the way she talks in my head.

I'm Dani, I'm one of the founders of jam.dev, and I'm based here in San Francisco.

And also why I think it's very important, like, a ghostwriter talks with their client frequently is to have that, or just listen to them on a podcast or whatever. It's like, I could hear Dani's, like, bouncy, happy voice in my head and what she wants. And, you know, she's very so specific about the style of like, this needs to sound that way, that way. We never want to come across this way. Like having that distinct set of beliefs around your writing and style is like so important if you want to scale that. And being able to communicate that is another completely different challenge. I'm even thinking the other day, you know, somebody was offering to make some art for some of my accounts and, you you know, I told him make it cyberpunk, right? And like come back and it's like totally different colors and like gray. And I'm like, wait, you don't know like what I mean by that? Like purple and blue? Like isn't that obvious? Like, and it's like, oh, you know what cyberpunk means a lot of different things, honestly. Like there's plenty of gray robots that are cyberpunk. I'm just talking about specifically vaporwave, synthwave cyberpunk specifically, right? It's like explaining my taste specifically. Because I'm in it for so long, like, it's, it's a challenge. Um, yeah, yeah, it's hard.

Managing is really hard. It's definitely like a learned skill, and like, I don't think you ever stop learning it. Like, I think one of the more powerful things too is just like— and this is the most impactful and hardest thing— is just like telling people what you think is so hard. Fine with that is telling people what you feel. That is probably the most impactful thing to practice and get right.

Yeah, I, it, it, it's really quite helpful to just kind of say the hard thing, especially with like a writer, creative, our feelings are hurt. We're sensitive people, but I'd rather get the hard truth and I'd rather give the hard truth to like a creative and be, you know, I'd rather just give it to them than, than say the dishonest thing and make them feel good.

You know?

And I'm, I don't know, I, you know, feelings are temporary, artists forever, right? And it's, uh, It's that, what's his name? Spielberg line, pain is temporary, film is forever. But it really helps to just get the hard truth sometimes, especially around writing art, around products too, around design. Like if it's not good, like you just gotta say it, right? That's right. Yeah. This was an awesome call. I really appreciated it. I found it very helpful and I think my listeners will enjoy it too.

So I hope so. All right, cool.