From Facebook's 5th Designer to Backing Stripe, Figma & Notion: Ben Blumenrose's Design-Led Bet

Zuck at that time, he had his desk in the corner. He had the designers in front of him. So he could just peek out and he could see the stuff you were working on. It was like fucking wild. And so I start doing stuff. I'm doing like 3D graphics and glossy this. And it looked like— it almost looked like Apple. And I'm so excited to show it to him. And then like a few days in, he sits down next to me. And he looks at what I'm doing and he looks at me and he goes, you know, you can't just come in here and shit on our style. Oh.

Ben Blumenrose was at Facebook in the beginning and was vital to shaping how the world connects today. Then he walked away to back a thesis the Valley dismissed, design-led companies. Today he's built Designer Fund into validation of that bet, backing Stripe, Figma, Notion, and Linear before the world caught on. Today he's revealing what he saw in them that everyone else missed. Super excited to have Ben Blumenrose here. He was an early designer at Facebook, joining in 2007, and he's the founder and managing partner of Designer Fund, one of the first, if not the first, design-focused VC fund, which apparently in retrospect was not a good idea to many people, but it's obviously perfect timing for a fund like this. It's great to have you here.

Oh, thanks. Excited to be here.

So I want to start off with your time at Facebook, because I know you have so many good stories. I read that you were really focused on user education. For new products. And I'm curious, some might say the need for user education is a bug because the product should be so intuitive that people just figure it out. So I'm curious, in your perspective, where do you decide this is intuitive enough that the rest we can take over with tooltips and guides?

Yeah, so I mean, I think it depends on the scale of the product that you're working at.. But we got to a scale— I remember we had a point of view that users should not need any user education. It should be so intuitive, someone should just be able to land there and just go. But we started getting to a certain scale where for me it became obvious that you want some sort of handholding for some people. And by the way, some people might want it at some point in the day, and later on in the day they might go to the site and not need it at that point. And just like depending on mood, and where they were at in the day. So we got to the point where we had hundreds of millions of users, and the idea that you could build a product that for hundreds of millions of people would just be intuitive to me seemed crazy. So one of the things I fought for was at least doing some sort of hand-holding and orienting when we would launch new products. And I remember, yeah, there was— so we had the timeline, the new timeline for profiles, And the thinking was, let's make it super intuitive. I said, look, let me and a team build a little bit of user education. And Mark was kind of adamant that no one would use that. No one would care. And also, even people who would want some user education, they wouldn't actually finish. We basically wanted to build a little bit of a tour. And in the end, we sort of did it under the radar and tested it. And it turned out that 50% of users wanted that tour. And almost 90% of the people who started finished the whole thing. Orientation. And you and I talked about it. It's kind of like a hotel room, right? You get to the hotel room, and they tell you— you might be tired that day, and they just say, OK, you just might say, look, I'm fine. Let me just go to bed, which is OK. But sometimes you want to know, like, how does the handle in the shower work? Where's the gym? Is there something at the hotel that I should know about? Is there something going on tonight that I should know about, right? Just a few basic questions to get you oriented. It doesn't mean that you're like— that the hotel was poorly designed, right? It just means that someone wants a little bit of a way to like orient and get settled. Yeah.

It's so funny that I end up working. We're focusing a lot on the onboarding right now as well.

Yeah. Well, you look, step 1 is at least like watch people. One of the things that we used to do is we used to do user research on new user onboarding at Facebook.. And just to see like what was the aha moment where someone would actually like kind of find the things they were looking for or whatever. And we'd watch the screen and we did this hundreds of times. And at one point I said, can we not watch the screen and can we watch their face? And all of a sudden you basically, there was, you know, we'd ask you what city did you live in, what school did you go to, Where do you live now? All this stuff, right? And then we'd start showing you people from your life. And if you watch people's face, the second we showed you a grid of faces of people you hadn't seen in decades, tears, smiles, that was the moment, like, the second you did that, they were hooked. They were just like, I want to find these people and contact these people. But you wouldn't see that by looking at the screen. It was only by looking at—

It's a hidden piece of data.

Yeah. And so, again, like, new user onboarding, don't just look at kind of what's happening on the screen. Also, you need to look at what's happening with the person. And so what we ended up doing was like, how do we get to that moment where we show the person the people as soon as possible? Because that's the moment that they can actually like, you know, that basically like sealed it for them. And then the other thing I would say is have people sign up for your product, like your employees, have them go through it daily. Like we basically got to a point at Facebook where Facebook employees were so steeped in it, they had no idea that like when someone could join and they would just see like a blank feed. That was happening to like thousands of people a day. But because we were just like already on Facebook for years, we never experienced that. So there was a product, a PM at Facebook who basically made it his like mission. He would sign up every day, record it, and send it out to everyone. Then do it again and again and again. He's like, it's still shit. It's still shit.

It's still shit.

Fix it. You know? But he was just like relentless. You got to bang the drum until like people internally got it.

Wanna go deeper? Talk to Ben's digital mind on delphi.ai/benblumenrose. Link in the description. Yeah, and so him saying, "It's still shit, it's still shit," this leads into my next question on does he stop because of data or because it feels good? And there's a story where a designer at Facebook moved the add friend button because he thought it would look better. And it actually—

It's not just look, I mean, he actually thought like, that's where all the buttons are and it should go where all the buttons go.

So it was maybe intuitive, but it led to a 3% drop in friendship and apparently the head of growth at— not friendship in general, in friendship at Facebook.

That was Facebook that led to a 3% drop, the platform itself.

And so the head of growth got super angry, and I'm curious how you think about intuitive what looks good and following the metrics.

I think the thing that happens too much is that designers are just not aware of these metrics. Of the things that are critical to the business. And so an example like that where you move the button to where all the other buttons are, right? It's like, okay, that seems fine. And I remember the head of growth basically was like, okay, do you understand what 3% compounding means? I don't remember, by the way, I don't remember if it was exactly 3%, it was enough that it was like significant, right? And to a designer it's like, well, that doesn't sound like that much. Right? And he basically said, okay, doesn't sound like that much, let's do this. Go to Mark and tell him that by the end of this year instead of 850 million users we'll have 600. And then see if he likes that button placement. Right? Now all of a sudden it's like, well, wait a sec, what do you mean? Because 3% compounded daily, that's what the trajectory difference looks like. Right? And what does that mean in terms of revenue for the company? And you as a designer, you have to know that that's what you're impacting, right? If you don't, you just look like an idiot, and probably rightfully so. So I think designers at the very least need to understand the metrics that we're impacting and have a sense of when we make a decision that maybe is at the very least neutral, or maybe even if it briefly makes the metric go down, you better be able to answer for why what you're doing is worth that hit, because if not, it just makes designers look like all we care about is aesthetics. Yeah. And not the final outcome.

Impact-driven.

Yeah.

There's a— you talked about how at a certain point there wasn't really clarity on when to bring an idea to Marc.

Yeah.

It wasn't really like structured process for new ideas.

Yeah.

And so I'm curious, what made you feel like you needed to create a structure and Did you come up with a when-to-say-no list to new ideas?

You know, we— I would say we actually didn't do a great job here. And I think— and this is actually like a learning that I— we tried to do something at Facebook and I don't think we ever really got to it, but it's something that I tell our companies to try to do, which is, do you have a sense of how products go out the door here? So if I had to like articulate the process at Facebook, it would have been like this. First, someone from somewhere at the company sees that there's an issue and does something about it, prototype, design, whatever. Then they go and run around and tell people that this is like a thing they want to build and see if they can get people excited about it. Then once enough people get excited about it and they start to build, then either go to like Mark Zuckerberg or at the time Chris Cox who was head of product and try to get them excited about it. If they're not excited about it, then try to do an end-around and not tell them about what they're doing for the rest of the time. Then if Zuck tells them to stop, then go into a room for 3 weeks and try to do it undercover and get it out the door so that the data will then make it like that Zuck has to say yes. There's so many permutations that we had. Depending on people's feelings, emotions, what was happening that week. It was a cluster, right? And it just made it a little bit of Wild West to get stuff out the door. So I think some of that is fine, but I also think it is really important to have— and ideally you have some sort of visual artifact that says, all things being equal, here's the process that we'd like to follow here. 2 weeks of user— like even just the idea that we do user research. We'd often get to a month or 2 into a product process and someone would ask, how do we know this is a problem? It's like— Fair question. Yeah, it's like, oh, because, you know, well, don't you see that? Isn't that your experience every day? It's like, No, I don't experience that. Did you pull data? Well, no, the data team was super busy. So how do you know? Well, I just thought we all agreed that this was an issue. And all of a sudden you realize you totally skipped the part where it was like, do we actually know that this is an issue for not just for people at the company, but for everyone using the product? So there's all that kind of stuff where it's the checklist. What are we doing in what order, and who do we bring in when? When to bring in— the CEO, and when does that person get the sign-off, right? So that we know like how do things go out the door, and when things are deviating from it, you can actually say, look, you already approved this thing, why are we now having this conversation again? And there might be valid reasons because all of a sudden the legal team in Germany says like, hey, the laws have changed and so we got to rethink this whole thing, and that's fine, but at least we know kind of where we're deviating from the the standard protocol. And it could be lightweight, but I think it at least creates some sort of framework and a plan for the team that they can understand. Like, if I do this stuff, this is how these things go out the door.

Yeah. Before we move on from Facebook, one more craziest story with Mark Zuckerberg.

So I came from, like, the agency world, and I was— because I think Saleh was on here before, right? So the designers at Facebook, I would say, when I joined I was the 5th designer and none of them looked like designers to me in terms of just like went to design school, like went through that rigmarole. Like it's like they had backgrounds in film and HCI or engineering. They all actually looked more like engineers than designers to me. I was like, wait, you guys can like code in C? Like where is the designer here? And so I came in and I was just like, and I don't know if you, you're probably not old enough because you were like 4 at this time, but like the site looked almost like a wireframe. It was like undesigned. And I came from the agency world where it was just everything was like glossy and beautiful and whatever. So the first project I get in and I'm just like, I'm going to blow the doors off this aesthetic. Like I'm going to make this look like, 100 times better than what it looks like right now. And Zuck at that time, he had his desk in the corner and you would design on 30-inch monitors and he had the designers in front of him. So he could just peek out and he could see the stuff you were working on. It was like fucking wild, which is why so many of us would like find nooks and crannies to like escape. And so I start doing stuff, I'm doing like 3D graphics and glossy, this like stuff that actually probably right now would be really cool. And it looked like, it almost looked like Apple, right? And I'm so excited to show it to him and I'm trying to like hide it, whatever. And then like a few days in, he sits down next to me and he looks at what I'm doing and he looks at me and he goes, "You know, you can't just come in here and shit on our style. I was like, dang. I was like, what do you mean? Like, what did you hire me for if not to come in here and push the thing? And he said, come in and understand why we do the things that we do and then earn the trust and then move the needle. And it was, you know, as a young— I was At the time I was like, what is this? What's this guy's deal? Like, what the fuck? Like, you know, like. But later I found out that actually, as an example, speed was like critical to Facebook's success. So much so that they literally tested, this is an example of like a designer not knowing metrics. We actually fake throttled certain users to see what it would do to their usage. So like you would be in San Francisco on high high-speed, and we would throttle you 10%, 20%, 30% in speed just to see the impact of your usage on the site. And it would drop you. Like, you would use it less if it was slower. And so the site was actually designed— like, the wireframe-y— it was an aesthetic thing too, but it was also like— it happened to be just the fastest way to get a site to load because it was like very few graphics needed. And that was a big part of why it won.

It's a good segue to— learning more about how you think about design. And at Facebook, you had to design for the individual, but you also had to design for the system. And you've talked about the difference between human-centered design and humanity-centered design. And so I'm curious if you have an example of something that looks good for the individual, maybe a design decision, but maybe doesn't really work when you zoom out and think of a network or a system at whole.

Yeah, I think we need to, you know, it's not that human versus it should only be humanity and not human. Ideally these things go hand in hand. But I think too often people basically say, let's give people what they want versus let's give people what they need or help them understand what they might want. And so the newsfeed is sort of a good example of this and we're always, you know, I still say we, but like, we're always like tweaking that algorithm. But the way I kind of describe some of the way it works is imagine you came over for dinner and I said, hey, I have, you know, whatever, salmon and like roasted broccoli, or like I can just give you like a scoop of ice cream. And you're like, really, for dinner? It's like, yeah, if you want. And you're like, yeah, I'll have a scoop of ice cream. That's kind of crazy. Next time you come over, I'm gonna remember, oh, you know, he was really into ice cream. So you come over and instead of salmon and broccoli, I'm gonna say, I have 1 scoop of ice cream or 3 scoops. Which one do you want? And you're like, ah, 3 scoops, great. And then the 3rd time you come over for dinner, it's gonna be ice cream sundae or 3 scoops. And you go for ice cream sundae and now you're like, dude, I'm going to Ben's every time. You know what he gives for dinner? Fucking ice cream sundaes. And so the newsfeed is like often that way where it's like we basically give you the thing that you want to see more of. But it might not be the thing that is like nourishing to you, right? But like we always took the point of view of, well, we shouldn't be in the business of like dictating what's good for a person, okay? And the thing that we never got to, so I have a good friend who, started this like leadership academy program in Israel. Thousands of thousands of young Israelis now go through this thing. And when I was, I was in Israel many years ago and basically I was sitting in the desert with him and earlier that day I got an email that said that Facebook had just surpassed television in time spent. And I told him this, and his head exploded. He was just like, "That can't be. That just cannot be." And I'm like, "I'm telling you, I know the numbers. I know the number of people who use it, and I know how much time they spend on it. Like, it's true." And he said, "How many people are responsible for what goes in there?" I was like, "You could probably fit them in a little room. There's like 15 of them." Yeah. And he's like, "How old are they all?" I was like, probably most of them in their 20s. Most of them are guys. Most of them have a similar background. And he said, who are the historians, sociologists, ethnographers around Mark to help him understand what does it take to build a good and just society? I was like, what are you even talking about? Like, what are you talking about, for real? Do you know how these companies work? A council of elders? Like, what are you talking about? I was like, there's no one like that. You know, he might have mentors, but like, it's not like those people are sitting in a room helping us understand, hey, historically, when power was put into these people's hands, or when this kind of information was disseminated, or the printing, you know, all this other stuff. There were— look, we were smart and curious people, so like, we— some people went and sought that out. But man, we could have been— it's like, it's the same thing like a designer who doesn't know the data of the thing that he's impacting. It's the same thing. It's like you have to have a much more holistic view of like the history that you live in, right? And understand the impact that you have, and those people need to be in the room. And I remember when the first kid passed away who had a Facebook account. I was like, what do we do with a Facebook account of a person that passed away. And it's like, well, take it down. That person didn't want— like, that person probably doesn't want his account still live when they're dead, right? And it's like, we don't know. Maybe they do, maybe they don't. We actually never asked. And then we started to notice— so we, you know, there were some accounts that we just kept up, and we started to notice a week or two in after someone dies, people would continue to post on their wall. To like manage grief. And you're like, well, that's kind of crazy. Like, what are they doing? Why are they doing that? Well, it's like, that's what people, when they go to graves, that's what they do. It's like, oh, hmm, well, that's kind of interesting. Should we be in the like grief— Grief business. Business. Well, I don't know, it's like part of humanity, and like, yeah, maybe these things actually should live on in perpetuity. And so we started letting people keep them up. And then we'd see a year, you'd see these spikes on someone's profile a year after. And it was all their friends, hey, it's been a year, still miss you, Dara. I was like, really? It's just like, oh my God. And over time as Facebook got to be, you know, being used by older people and all this stuff, like death became a very common thing on the platform and we had to design around that. And now it's like, I think we even, now Facebook will ask you like, what do you want to happen after you die with your profile? So what do you want to happen as an individual, right, is an important question. Now another question is, how healthy is it for your friends and family to have this thing that exists past your death, right?

Is it good or bad for them?

Yeah. Religion deals with this a lot, and religion does take a very strong point of view around what to do and what not to do. So as an example, like in Judaism, we basically say we don't want you to visit the grave, I think, for the first month. Well, why? It's like, we want you to— so you have your period of grieving, it's 7 days, right? You sit shiva, and then you have to get on with your life. Like, living is about living, and the religion takes a very strong point of view around that. Well, like a lot of tech companies and tech products are like, we don't want to dictate that. Like, you know, it's like, what, what, like at least have a point of view around, you know, what is like a healthy process through this and maybe let people kind of navigate some changes. But we do know that probably holding on to someone, you know, for years or decades after they've gone is probably not super healthy, right? There's some, there's some point where you have to kind of like, go through some process, right? So these are the kinds of things I think companies need to be asking more, and AI, oh my God, there's so much of that stuff, it just needs to be part of the conversation.

AI has kind of commoditized just like MVPs, spinning up the average of designs. So I'm curious, where do you think the edge for designers today, where does it lie? What do you see your 100x AI-first designers doing that make them ahead of all the rest?

Yeah, so I would say we're in a very shifting sands moment with product building and design and all that stuff. I think one of the themes around this is AI versus craft and basically what's the impact of AI to building good products. And to me, I think the way I want to see this evolve and the way I think a lot of designers are using it is basically like AI is the enabler, is the supercharger, but is not the replacement for good high-quality output. So I actually think like the best use cases I've seen is AI actually letting you get to higher quality craft and higher quality output, where you're like, well, that doesn't make sense because it sucks at design. So even for a relatively small feature, you can ask, what have we done here that has worked and hasn't and why? What are other companies doing like this and why? Where does it fail? OK, great. Now go through my product as an 18-year-old and tell me what you notice. Go through my product as a 40-year-old, tell me what you notice. Go through it as a 75-year-old who's not— tell me what you notice and synthesize that. And all that hap— and these are things, it was not even tenable to think this way 5 years ago. Now it's just like, you know, it's almost like imagine you have endless resources, what's the information you would take in before you start your work? Then once you start the work, it's like, well, how much time do I have a week? Okay, I'll try 1 or 2 ideas and maybe 3. And the best of those 3 ideas should be good enough. Well, maybe there's like 100 ideas I could try, just quickly. Ch-ch-ch-ch-ch. But then you as a designer is like, okay, like of those 100 or 20 or whatever it is, like this idea seems to work really well. And the way you know is because again, you're putting these AI agents or some other methodology through these things. That plus your taste gives you a much better place to start doing like the craft work.

It allows you to do fewer reps to get to the local or global maximum.

Yeah, yes, though I would say like in some ways it's like there are reps that happen before, so it's like you probably have to do fewer like high-fidelity reps that take a long time, but you can do dozens of like low-quality reps to get there. So I don't think it's like, um, We just, we really need to understand there's like different parts of the process and AI is like gonna let you supercharge different parts of that process. But it doesn't mean that you have to AI-ize the whole thing and now it's like the end product looks like something that an AI machine spit out. 'Cause at least right now in terms of like a final output, it's just still not that good as far as design. Like engineering is quite good, but to Design-wise, like, it doesn't hold water to, like, really good designers yet.

Do you see that on that final mile, like the finesse, the interactions, or is it more in the UX and being data-driven?

Look, design is this thing that you can apply to so many lenses. So I'll give you a great example. So when I joined Facebook as a product designer, one of the first things I did was I was like, how does this company like even work? Like what's going— who's making decisions and why? And the company I think was about 150, 160. So I was like, I want to have an image in my head of this company and who— and what are we doing, right? And who's like kind of holding the strings? So I ended up setting up meetings with every org and I was just like, look, I'm a designer here, Forget what you know about designers, that if I could design for your team, and I met with head of legal, head of HR, head of customer support, what would you do with that? And almost universally, all of them were just like, I never even thought about it. I never even, I can't even imagine having a designer, so I don't even know what I would do with that. With that resource, right? And so one of the things that happened out of that process was like customer support. We had 50 customer support reps that were all 21 to 24, all graduates of Stanford, Harvard, and MIT. And they were the judge, jury, and executioner on Facebook. And the stuff that they were being put in charge of to decide is insane, right? You're talking about someone saying, "Hey, my friend is threatening to murder this guy on campus. You need to do something about it." And they need to know to escalate that to police, right? Or like, you know, there's just so many things that were like societal that these people were— this wasn't like an Emma, like, I'm returning a set of— a pair of Nikes. Like, this was like real fucking things. And dealing with all those things at scale was hard as all hell. And so I went and I— you know, this was— I talked to you a little bit. I went and I sat down and I just said, let me spend a day to understand what's going on with customer support. And I sit down and they basically all had these common responses that they were told. Like, when someone's like locked out of their account, use this common response, right? And I'm sitting down with the first person and he says, "Okay, here is a common response I'm supposed to reply to. It's not that good. Not a good response. Here's what I do." And he pulls up this like the sticky notes app and he had all his common responses in a sticky notes app and he was using that. I was like, "Oh, well that's interesting. This guy's gone rogue." But like, you know, he seems to know what he's talking about. Then I went to number 2. I watch her and she's like, "Okay, this person," they wanna report something, there's a common response for this. I don't use the common response. I was like, you don't? She's like, no, no, no, they're not that good. Here's what I do. She pulls up a Microsoft Word doc. She has all hers in a Microsoft. I'm like, well, okay, that's 2 people out of 2 that have gone rogue. That's interesting. And then I go to the 3rd person and same fucking thing. He opens a doc and he's like, I have my, I was like, everyone's gone rogue. This whole fucking place is rogue, right? And no one knows, right? And by the way, all these kids— and I call them kids because they were in their 20s— all of them were using these tiny monitors and they were like flipping back and forth between tabs to do all sorts of research and blah, blah, blah. And when you looked at the managers, they had these 30-inch screens and they were doing similar work but on 30-inch screens with like way faster because they had they were able to do side-by-side research. And I was like, this tool stinks. Everyone in this fucking room should get 30-inch screens like tomorrow. And basically in a month and a half, me and 2 engineers rewrote the whole system that— and literally overnight we like, we took the 50 people, we cut them off and put them on this new tool, and overnight their ability was 100% increase. So instead of supporting 100,000 users per person, it went to 200,000. So it was literally like you hired overnight 50 new reps that were expensive, and now boom, they're there. And by the way, guess what? All the 50 now are enjoying their job, are collaborating because like they were able to write customer responses that they could share, best practices. So the system got better and better instead of people hiding their customer responses in a corner. And that was all because a designer just said, I want to see what the hell's going on in this little part of the org. And so when you think about like the impact of design in a company, you start thinking about applying this to every part of an organization. And that's what it means to be like design-driven, design-led as an org. So for me as an investor, like that's what I want to see. I want to see our companies not just doing it on product or brand, but basically bringing a design process to everything. And when you do that, it just supercharges. So there's so many areas where design is just gonna be like a big needle mover that most companies don't look at it, and the ones that do are just gonna like outperform, I think.

Yeah, I didn't even think about design for sales. So to talk about Designer Fund, you told me this, but I'd love to go deeper into You were starting the first ever design-focused fund. Yeah. And why did people not— what was people's reaction to that? And what gave you the conviction that this would be something that feels obvious now, but in 2014 maybe it wasn't so obvious?

So it was a confluence of things. One is the success of Facebook basically all of a sudden, if you were like a lead of some sort, especially a designer engineering lead, all of a sudden you started getting emails from people outside of the company. And I was getting a lot of like, "Hey, can you help us with design?" And it's like, "Well, tell me like what do you need help with?" And it'd be like a company that's 3 years old, raised tens of millions of dollars, and they're like, "Here's the product." And you're just like, "This is criminal. Who let you build with 27 engineers and a part-time designer from Eastern Europe and expecting that that would work? Like, this is crazy." And so I just got so many of those kind of things that it made me wonder, like, why is this happening? And so then I went and I followed, just like a user researcher, let's like follow the tracks, right? So it's like, well, who's investing in all these people? And I started interviewing all these investors, and they were like older, usually white, usually dudes from engineering or finance. And to me it was so obvious when I was at Facebook, I was just like, it's so clear that when you marry good design and product and good engineering, that's like the winning formula, right? And you started like, the more I talked to other investors, design was like this total afterthought. And I was just like, how do so many investors like not get it? So that was one. Two was like the importance of getting it became bigger and bigger because it wasn't just like, 'Oh, okay, they're just designing social networks. Who cares? We have a good one. That doesn't matter then.' But it was like healthcare software, education software, it's like, you know, we were software, we were turning everything into software and I'm just like, I'm starting to think like we're building a whole city and architects are nowhere to be found. This city is going to be shit. You know, you're talking about humanity-centered design, like this is not good. So I started getting more and more concerned about just that as a thing. And then the other thing that was happening was I started to see like my peers who were designers who were starting companies, really struggle to get anyone to believe in them. So like Joe and Brian, who started Airbnb, they are RISD designers, and like early on, they would stop by Facebook and be like, here's what we're doing, but like investors don't get it, and we're just like, how do they not get it? It's just so much better.

So it's 2026, we have agents that are using interfaces for you, we have personalized software. What is Designer Fund excited about in 2026?

Yeah, so for us, I think it's where are these like hairy problems? And more and more so, I think it's like software and the physical world where AI is going to be an enabler for better products and services, and we think there's like a 10 or 100x product design improvement for it. So I'll give you like a couple examples. So we're investors, and some of the stuff that we do is like people just— it's not like— we do do some of like the sexier stuff, but there are things where 99 out of 100 investors look at it and they're just like, what, that? So I'll give you an example, headache clinic. So we have a company called Maven. And if anyone knows, like if either you suffer from migraines or anyone you know that suffers from migraines, it's like an incredibly debilitating thing to suffer from. There are so few experts who can help with this thing. So the wait times are like 9 to 12 months long. And then when you finally get to see a headache expert, it's like, let's try this, let's try this, let's try— it takes months and months and months until you find something that's, sort of works, right? And so these guys were like, felt the problem firsthand, either them or like, you know, family members, and they basically created this experience where it's like we make headache practitioner, like headache experts way more accessible. Once you get seen by someone, which is in days instead of months or years, the experience after the fact is like AI-driven, right? So like the more that people get treated by them, the more data they have, the better it is. So they're going to slowly start outperforming— I've already seen this— outperforming the original data, the original— or like the experts that are now at Stanford or whatever. And you should see like the feedback they get. It's just like— I mean, literally, yeah, it's like they're changing, like people's like lives, they're just like, I can finally function again, you guys listened. It does tend to be, it's like 80 to 90% is more, of their customers are women, and the feedback is like, do you know how often I go to a doctor and they're like, are you stressed, or like, maybe it's your period, or maybe, you know, it's like, it's still to this day, right? And they're like, no, we're gonna actually like listen to like what you're feeling, why you're feeling that thing, and here, and based on the hundreds of people that we have already seen, here's the medication to try, here's the data input. They can get data input way better and faster than like these traditional clinics who are going to see you 3 weeks after the fact, right? And it's messy, and it's atoms, and it's emotional, and that's where it's like, to me, is like you build a better product and service there, and people are going to flock, and they're going to tell 100 people, and that's what's happening right now. Everyone that's been treated by Maven is telling dozens of people like, I found this, like, I can't believe it's real. I can't believe it's actually, right? And it's like OpenAI is not going to do this shit. Like, it's way too fucking messy and crazy, right? So OpenAI can enable some of this stuff, but when you actually need the medicine, when you actually need someone who's— you need to feel supported, you actually need the treatments physically, right? So that's the stuff for us. Like, we get really excited where there's a lot of user experience around it, a big meaty product. I love when someone tells me, like, 'Even with AI, this is going to take like 2 years.' I'm like, 'Great.' Yes. Yeah, yeah. Because the things where it's like, 'Hey, in 2 months we now have like millions of dollars and hundreds of thousands of users,' I'm like, 'You can lose them that fast then, and another company can build this just as fast.' So I want things that take forever to build. I love that.

And what's one question do you wish you were asked more?

What's something about our product that should be improved. So I think it was—

I'm gonna ask you that after this.

Yeah. But I was talking to this guy and then it turned out he was like the guy that started Substack. And the second he started, and then his second question to me after I found, I was like, oh yeah, we use it. He's like, what would you improve about it? I was like, how much time do you have? He's like, I'm all ears. But it's like, you know, it takes like a humility, like Substack is big. And plenty of people use it, and they're doing a lot of things right. And his second question is, how could we be improving the product? And man, I wish more of our founders and designers just seek out— just seek back feedback. Who wants to hear, especially if you're doing well, I don't want to hear someone be like, well, this thing doesn't work so well. It's the last thing you maybe want to hear, but I think The more that you can ask that, the more you can understand ways to improve it. And I just wish more people would just like bite the bullet and just have that be a muscle they flex.

Thank you so much. This was awesome.

Yeah, thank you.

So many good stories. I feel like I need more access to you. Thanks for tuning in. For personalized advice from Ben, head to delphi.ai/benblumenrose and ask his digital mind.