Mosh Lee 1:18 I snuck out because I wasn't sure if there'd be time to get set up. 1:22 We're all going to be a bit offset. 1:24 Yeah, it'll be fine. 1:25 That's how it always is. 1:27 Yeah. 1:45 Let me move this. 1:57 I'll just stick it over here. 2:08 And then I have this. 2:09 I'll just pass it to you. 2:10 I'm sorry. 2:11 I have the mic. 2:12 Oh, OK. 2:14 Great. 2:14 OK. 2:15 Sounds good. 2:18 Good morning, everyone. 2:21 I'm on East Coast time, so this is halfway through the day. 2:43 Where on the East Coast? 2:44 Philadelphia. 2:45 Okay. 2:47 Yeah, Boston. 2:56 Cool. 3:24 Hey, when your talk starts, sure. 4:47 How are you? 4:49 I feel like I'm watching or waiting for popcorn, like, you know, there's like the last few kernels that might pop or might not, but like, do I open the microwave or not? 4:58 But anyway, it looks like we have some critical mass, so I'm very excited to be here with everyone here in this room today because I think we are building the future of community and collective software, and that is pretty pretty awesome. 5:11 My name is Maashli. 5:12 I'm currently the director of the IPFS Foundation. 5:16 And today I want to talk about feature product business, the framework for building an enduring AT Proto ecosystem. 5:25 So who here has heard of IPFS? 5:29 All right, cool. 5:31 We make a constellation of tools and protocols for content addressing. 5:34 Some of them, like IPLD and Dazzle, are building blocks for AT Proto, so If you're building on AT Proto, you might be using this already, but we've also been evolving the protocol to be more modular, more web-friendly in projects like Dazzle, which is a simple and more adoptable subset of IPFS. 5:52 I think Robin is in the room who kind of masterminded that. 5:56 WebTiles, which are safe composable modules for web apps. 6:01 MetaDisco, which is a system for science metadata publishing. 6:05 There was a short talk about that tomorrow afternoon at 4:40. 6:09 And then the CID Congress is our monthly virtual series where we get together a lot of the nerds and friends evolving protocols for content addressing. 6:18 So come join us. 6:19 That's on Luma. 6:20 I'll make sure that link is out in the world for you to find. 6:24 But we love to meet new friends, learn about your projects, and evolve these protocols together. 6:31 Zooming out a bit, my talk today is not really about my work at the IPFS Foundation quite so directly. 6:37 I am here because I've been on many sides of the table. 6:42 I built Google Forms, I founded a govtech company, I sold it, I watched it get sold again to private equity, I ran dev teams, developer programs, and grants at Protocol Labs and IPFS. 6:54 I think totaling $25 million in grant programs. 6:57 I'm also a small-time angel investor. 7:01 So this talk is really what I wish someone had told me and the conversations I wish I'd had. 7:06 There are no prescriptions here, there are no right answers, but I think there's a lot of information we don't talk about when we're busy talking about building software. 7:14 And I'm happy to make this interactive, so I'll pause every once in a while for questions or hands up or just shout out if you're nice, and I'll make sure we have time at the end as well. 7:25 Alright, this talk is gonna be in 3 parts: the framework, our AT Proto landscape, and some talk about funding models. 7:32 The first thing is the title of the talk, Feature Product Business, and so in a short phrase, a feature is something that solves one problem, often delightfully well. 7:45 I'll give some examples soon. 7:47 A product is a bundle of features solving a clear job. 7:50 It might help you solve what you're sitting down at your desk to do this morning or what you're setting out in the world to do. 7:55 And a business is an organism, sometimes one product, sometimes a collection of products or features that can outlive itself, has competitive advantages, has a team that can be resilient to shocks, exogenous shocks. 8:09 And the reason I chose this analogy, rock-paper-scissors, is that there is no one winner, right? 8:16 All of these can succeed depending on the context, the game you're playing, what other players are putting out there, but they are very different. 8:25 And when you mistake one for the other, I think that's when you enter a world of pain. 8:30 So a feature, like I said, it solves one problem. 8:33 And I think one big test of a feature is can you plausibly think it'll become a pain in someone else's app or someone else's project, right? 8:42 If this is something portable, there's a very low barrier to entry. 8:46 There's no onboarding, often no account required, no logins, no subscription. 8:50 It just does one thing well and stops and it's beautiful, and thank you very much, you've solved my problem. 8:55 The downside, it's really easy to copy or absorb, and if people really love a feature, you might get confused into thinking there's a business. 9:05 I made that mistake, I'll talk about that more later, and I think really the worst case is building a feature and having a lot of people love it, but everyone assumes that love is your business model and that'll keep the ship afloat. 9:18 Okay, so some features we love, just some examples. 9:23 Clippy, right? 9:24 Clippy's a feature, easily identifiable. 9:26 It tries so hard to do its job. 9:29 We actually sort of play— I think Clippy was onto something because now we have autocomplete, which is kind of the same thing but less of a guilt trip, I think, when you ignore it. 9:41 View source is maybe my favorite feature of the whole web. 9:44 I'm a millennial. 9:45 This is how I learned to code. 9:47 It is in every single browser. 9:49 It's really easy to copy across browsers. 9:51 But it's an incredibly powerful feature. 9:52 So I'm not talking about a hobby project when I talk about features, but I'm saying that this is, you know, it's more powerful when it's integrated into something bigger. 10:04 Curl is another example and quite different. 10:07 It powers so much of the internet's data transfer, but it's been maintained by one person for decades. 10:12 It's not a business, it doesn't need to be, and the boundaries are very, very clearly defined. 10:19 A lot of developer tools, like here's one example, just a CID lookup. 10:24 You paste the CID, you get it decoded, you get version, codec, multi-hash. 10:29 A lot of developer tools are like this. 10:31 They're features. 10:33 They do one job. 10:34 They do it really fast. 10:35 You get in and get out. 10:37 And the last one which I'll talk about because I have some personal experience with is Google Forms. 10:43 That is something I created a long time ago now in internet years, and it's It's still around, and I'll talk a little bit about why it's still around. 10:52 But there are also features that didn't make it that were really fun but didn't really survive. 10:57 Does anyone remember Vine? 10:59 Like short-form looping videos, like the next successor to, I don't know, social media. 11:04 And everyone else just copied it because it turns out that that's just a format. 11:08 They didn't really have anything else unique around it to make it more powerful or more sticky or build deeper relationships. 11:17 Another one I think, and it's still around in some forms, but RSS, right? 11:21 RSS was an incredibly capable, powerful capability, still is. 11:27 Podcasts use it, but for a long time this really transformed how people consumed the internet. 11:32 You could build your own RSS feeds, subscribe to publishers or individuals or channels, and Google Reader was the product alongside it. 11:41 Google Reader didn't have a business model, and so I think we lost a lot of what we could do with RSS. 11:48 Let's talk about products a little bit, and then we'll come back to some examples that are blurry and straddle the lines. 11:55 Products are bundles of features that solve a clear use case. 11:59 The big test is would people be upset if it shut down tomorrow? 12:03 A lot of Atmosphere apps are products, including our favorite, Blue Sky, and there are a lot of other builders in this room building some incredible apps on top of this ecosystem. 12:14 So it solves a bigger problem, not just a piece of one, right? 12:18 It solves a problem end-to-end. 12:20 Peter yesterday was using the phrase jobs to be done, so it really solves a big job to be done. 12:26 And they really build loyalty, not just usage, because they can get integrated more deeply into your workflows. 12:32 Products are pretty expensive to build. 12:34 Once you launch a product, every user wants a different feature. 12:37 They hassle you about it, they email you, they skeet you, and you really— and you can't just— stop and maintain what you have. 12:45 You have to keep growing and evolving it to feed that hungry beast. 12:49 So it's quite a bit of software to build, and you want to make sure that it's worth it if you do that. 12:56 Here are some products we love that use very different models for sustainability. 13:01 Wikipedia is a free encyclopedia. 13:03 Anyone can edit. 13:04 It's maintained entirely by volunteers, but the hard costs are funded by donations. 13:09 This is maybe the most successful public good on the internet, and it has survived quite a lot of assault. 13:15 Homebrew is another example. 13:18 Homebrew is a package manager for Mac. 13:20 It's maintained by volunteers, used by millions. 13:24 But actually just a few years ago, one of the lead maintainers of Homebrew started a company called Workbrew because people kept trying to make feature requests of Homebrew that weren't appropriate for an open source project. 13:35 They were more about tying into an enterprise environment. 13:38 And they said, all right, well, we'll just give you a different universe that can layer on top of Homebrew That'll be the business model and it'll also fund our work on Homebrew. 13:47 So that's a very popular product, I think. 13:51 And Pinboard is another example. 13:54 This is another single developer project and product. 13:58 They've had a paid model from day one. 14:02 But Pinboard is a very fast, no-nonsense bookmarking website. 14:05 It has survived, it has outlived a lot of other ideas and projects for bookmarking and saving information. 14:12 And part of it is that being a single developer, single maintainer, you can be really nimble. 14:18 So you can make opinionated choices, you kind of live in, as a user, you live inside that person's brain, and that can build really strong relationships. 14:27 It's a way to create software that really makes sense holistically. 14:31 You don't have a committee arguing or disagreeing with each other about certain feature choices, right? 14:36 And so that's kind of the magic of single founder or very small team software. 14:43 And so these are very, very different models for products, but they are all products, right? 14:50 But products are also precarious because we can lose them. 14:55 So Dark Sky was a weather— did anyone use Dark Sky? 14:59 Yeah. 14:59 Amazing weather app, the most beautiful visualizations, wonderful interface languages, really accurate data. 15:08 But it was acquired by Apple, shut it down. 15:10 So bye, Dark Sky. 15:11 RIP. 15:14 Google Reader is another one with a ton of devoted users. 15:18 And for a very long time, it was funded because of Google's overarching thesis that more internet time is more better for them. 15:26 And we know how that's turning out. 15:28 But eventually, they began to zoom the magnifying glass in on specific projects, and Google Reader sadly didn't make the cut, and so they lost interest and killed it, and along with it, I think a lot of the usage and power behind RSS. 15:47 And I think that's kind of what would happen, right? 15:49 The relationship of Blue Sky right now is still pretty dominant in the atmosphere, so if we lost Blue Sky the app, it would have pretty negative effects on the atmosphere, and the sooner we can build against that, build a more resilient ecosystem, system, the better. 16:06 Odeo is a fascinating one, and I think a really interesting story for this particular audience. 16:11 Odeo was a podcasting startup in 2004. 16:14 So in 2004, there weren't iPhones. 16:19 People weren't walking around with computers and audio and video files in their pocket very much. 16:24 And so they just got there a little too soon, and they quit when iTunes also launched podcasting. 16:30 It turns out that podcasting is actually a big enough to support multiple startups, multiple players there, but they quit too soon. 16:38 They weren't really focused on creator or publisher experiences or any help with quality production monetization, anything around the ecosystem. 16:48 I think they just saw that, all right, Apple made a player, I guess we're done here. 16:51 They started running hackathons, and out of that hackathon, one of those hackathons came Twitter. 16:56 So yeah, that is— That is audio for anyone who doesn't remember. 17:06 And there are also products that we really love in extremely obscure niches. 17:11 And I think many of us in this room are extremely online. 17:15 We use social media, we connect to others, we meet strangers online, and then we get together with them in person. 17:20 Sometimes we even start companies with them without having met in person. 17:24 This is a weird way to live. 17:25 Most people don't live like that. 17:27 Many businesses, industries, sectors operate in these very specific niches. 17:34 And if you can deeply understand that niche, if you either come from it or partner with one, or you spent many years maybe providing services to, to these industries, that is an incredible unfair advantage that you can build. 17:45 And I think we underestimate these niches. 17:47 So here's just some examples. 17:49 Zotero, I haven't used it myself, but this is a freemium reference manager for researchers and academics. 17:55 Academics. 17:57 It's a 19-year-old project funded by individual and institutional subscriptions. 18:01 So to use it, basic— the basic mode is free, but if you want to actually store all the documents that you're collecting and referencing, you pay, and you can pay by individual or you can have, you know, the whole chemistry department at your university or research institution subscribe. 18:14 So $5 million a year revenue has been enough to keep it going for almost 20 years. 18:20 ArcGIS Esri is another one, kind of an elephant in the room. 18:26 This is a 50-year-old company. 18:28 That is pretty incredible. 18:30 They've built such a moat. 18:31 They're not cooperative at all. 18:33 But they've done a lot to advance the state of GIS. 18:35 And this is a $5.5 billion valuation company. 18:40 Some more obscure ones that I think are fascinating because they're much bigger than you think. 18:44 I don't know if anyone has little kids here. 18:47 but in the past few years all the daycares and preschools started using apps to communicate with parents. 18:52 That's kind of a baseline expectation now, and most of them are really sloppy with privacy. 18:58 There's only one that gets a lot of the user experience and the privacy, the data security right. 19:04 That's Brightwheel. 19:05 I don't think users are choosing it, or the business owners are choosing it for privacy, but it's really nice that it's bundled along with the nicest customer experience. 19:15 Because they started with a $600,000 investment from Shark Tank and are currently at a $600 million valuation. 19:23 Preschools and daycares. 19:24 So this is a niche I don't think many VCs are really talking about, but it's very real to our economy. 19:31 It's really real to parents. 19:33 And another one is Veeva. 19:35 I had never heard of it before I was doing research for this presentation. 19:38 It's software for managing pharmaceutical sales clinical trials, so you can imagine the data burden, the compliance burden is tremendous. 19:47 Used to be that the consequences for fucking it up were huge, and so it's important enough and useful enough that people are gonna pay a lot to have the problem solved for them, and the team that was able to understand this vertical well enough is doing extremely well. 20:07 So it's not just data there, but they're also managing communications and participant recruitment. 20:12 So there's some element of community there. 20:15 And there are also big communities of patients actually sharing their experiences, especially in rare disease. 20:22 So these are prod— and I say these are products, but they're actually all businesses because they have a deep, deep relationship and very resilient relationship with their customer base or user base. 20:35 They've built big enough teams. 20:36 Figured out some kind of revenue model, whether it's a spectrum of freemium to expensive, selling to individuals, selling to organizations, but they're actually all businesses. 20:46 And features can also become businesses. 20:49 So this is not a linear path. 20:52 Calendly and Stripe, I think are exactly, all Calendly does is schedule your frickin' meetings, and yet people pay for it because they do this one thing extremely well, and they integrate with everything else in your workflow. 21:05 So you can build a business on a feature if you exploit a small wedge and you just really crush that wedge. 21:13 There's another pattern for features growing into businesses. 21:16 Zapier is just 6,000 integrations on a marketplace. 21:21 None of that is cutting edge. 21:23 They've just reduced all of the friction of composing your different services and apps and data sources. 21:29 Mailchimp is the same thing. 21:31 They've got a lot of templates, and the value is accruing over time. 21:34 So the first template is just a product. 21:37 The 100th becomes a moat. 21:39 The 60-millionth template, you start to build some advantages by making this very, very long tail usable. 21:48 And another really common one is just packaging internal infrastructure. 21:51 You build some dev tools to solve your own problem. 21:53 You realize everyone else needs it, and so you upgrade it and expand it so others can can use it too. 22:00 So yeah, I kept throwing the word around business. 22:04 Business can also mean a nonprofit business. 22:07 At the end of the day, it's something that sustains itself over the long term. 22:13 I talk about the team a lot because it needs to be able to survive exogenous shocks. 22:20 It has to have a moat. 22:21 And I think this is possibly controversial here. 22:24 It's an uncomfortable thing to grapple with. 22:27 There are many ways to build moats. 22:29 You can have some extremely anti-competitive practices, use proprietary data formats, extremely aggressive contracts. 22:38 Those are all ways to build moats. 22:40 Other ways to build moats, if you think about it, are— Instagram took the route of we're just gonna make this a happy, beautiful place for everyone. 22:48 And so their moat was really in the accretion of small content moderation decisions over time. 22:53 They built a reputation for creators as a place where harassment would be slightly less than other places in the wild. 23:00 I think this is evolving for sure, right? 23:03 But their moat was their community. 23:05 And I think Reddit has been doing surprisingly well as well. 23:11 And I think that's also because they give each community the tools to manage their own communities, create their own third spaces, set their own rules. 23:20 And so The moat can be the friction of getting a lot of people to move. 23:23 It doesn't have to be strictly something that's aggressive and antitrust. 23:32 I think the biggest challenge with a business is it requires developers to get good at things that many of us are bad at or just really not interested in— distribution, pricing, support. 23:47 Sales. 23:48 I joked yesterday that we're going to have to learn to pick up the phone, probably half of us, and answer unknown numbers. 23:55 I'm really sorry, but some people are going to have to do that. 23:59 And it's really easy to make a mistake where a product and some revenue can start to look like a business, and it does fine for a while. 24:06 You get some base hits, but it doesn't have that enduring value. 24:10 So I want to talk about some examples. 24:12 This is the time I built a feature that got confused for a product, it's really a feature. 24:18 So I invented Google Forms when I was a designer on the Google Docs team, and we kept hearing that people were struggling with permissions on sheets, like who gets to be an editor, who gets to be a commenter, people were overwriting each other in the cells when you had too many people using it, and at the end of the day, it turns out most people Most people don't need to edit cells, they just need to submit their one row of information, and then you've got your 2 or 5 or 10 information managers that actually need to control the structure behind it. 24:49 And so dumping data directly into Sheets was the superpower, right? 24:53 We obviously made it as beautiful as we could in 2000— I don't even remember what year, this is a very long time ago. 24:59 Airtable is a lot more beautiful now, but I think the enduring value here is that we dump the data directly into Sheets, and anyone who manages data is gonna probably say, like, I want the easiest interface for data manipulation and organization afterward. 25:16 So that's the product, the Forms plus Sheets is the product, and then G Suite where you wrap in Calendar, Mail, and then you sell that entire subscription to an organization, that's the business. 25:27 So you can build something that everyone knows and loves and uses a ton, but it might still be a feature, so be careful about that. 25:37 And then another time I built a product and I built the beginning of a business, but we sold it to a much stronger business and you could really feel the difference in the air. 25:48 And so this was 2012, Obama was in office. 25:53 I think we all thought technology could fix government, but governments were also holding public meetings that no one was going to, and so we built a two-way text message chatbot. 26:02 You could program it. 26:04 With different interactive workflows. 26:07 When we launched it, because of the very strong Code for America network, we got 240 inbounds in a week. 26:13 They said, we need this, we need this. 26:14 And so we thought that was our business, right? 26:17 And it was, it was the beginning of a business. 26:19 Knight invested, half of our team switched to sales, I started picking up every unknown number that called my phone. 26:25 I learned about P&Ls, talk to lawyers, contracts, all this stuff. 26:31 And we did get pretty good at local government sales, so city and county and state-level sales. 26:39 We were not qualified to sell into federal, which is where the majority of the customer base is. 26:44 And so I think this started with some reseller licenses. 26:49 So a bigger company, which had been around for 14 years, they were the SendGrid of government email. 26:55 Called GovDelivery at the time, and they were the ones sending out, hey, you have 48 hours left to register for health insurance, to most of Americans. 27:06 They had 97% of the state and federal agencies as their customers. 27:10 They bundled that into— bundled our product into an add-on. 27:15 They immediately started selling it for a lot more than we were, especially at the high end, and then we were kind of the expert expert consultants that came in and helped them figure out their integrations. 27:24 So I think we got to close to a million-dollar run rate for the year, but it was really palpable when we went on these ride-alongs how much of a head start this other company had in building their business. 27:40 And I think the takeaway here is that to survive and really thrive as a standalone business, probably needed to one, more deeply integrate into like 311, Salesforce, all those existing workflows so we could become part of their system, and also go multichannel. 27:58 So we were text messaging. 28:00 We were like the hot new kids who understood text messaging. 28:05 But we needed to come alongside handling 311 calls, emails, other constituent services to really become like an all-in-one suite. 28:13 So if we'd wanted to stay independent and continue to grow, we needed multiple products to bundle into a solution. 28:24 Okay, we're gonna start part 2. 28:27 Any comments or questions or stories? 28:32 Yes. 28:32 [Speaker] Do you think that with the AI landscape that SaaS businesses are gonna get eaten up? 28:40 Are people leaving software as a service I always think of these revolutions as a spectrum. 28:49 If we pay attention, things like Airtable and Zapier already took the place of a lot of that software. 28:56 I think it's going to be a little bit slower for these enterprises to adopt. 28:59 Right now, they're still all stuck on chatbots. 29:01 And an MIT study showed that 95% of enterprise AI deployments are a waste of money. 29:09 The 5 that succeed, the 5% that succeed are deeply integrated into workflows and done by domain experts. 29:16 So if you have the right people telling the AIs what to do, I think there's huge potential there, especially in really constrained environments. 29:23 You know, one of my passions is public sector software. 29:25 And so if you can imagine a crack team, instead of paying like $115 million to McKinsey, being able to manage a much smaller contractor software project on their own with the help of coding agents. 29:38 That's really powerful, and I think, yes, the SaaS market will really collapse, but there will always be innovators on the edges that understand how to connect components in a way that the AIs don't see or the big incumbents don't see. 29:56 Okay. 29:58 I think our app proto landscape in a completely unscientific review, it looks a lot like this. 30:03 We have a lot of amazing features, we have a lot of fantastic products, we have a handful of businesses' solutions that have real customers. 30:12 It's really fantastic to see that we have all of these. 30:16 But this is lopsided, and we've got a little too heavy in the first two categories and a little light in the third. 30:26 Let's see. 30:28 Jim Barksdale from Netscape has said, There's only two ways to make money, bundling and unbundling. 30:34 And what he's talking about is like making something easier to access or grow by putting pieces together, or once that gets too bloated, you can unbundle and innovate in a very specific slice to do something novel and exciting and really meaningful and powerful. 30:51 And so his thesis is that our technologies go and our markets go in waves, so there's like the great unbundlings and the great rebundlings. 31:02 And Ravi Mehta wrote an essay that highlighted like the unbundling is appropriate when you want to disrupt and you want to bundle to grow. 31:11 I don't know if we're fully on one end of the spectrum, the pendulum swing or other. 31:16 I think we're somewhere in the middle, but just think about it as ways to evolve what you do if you want to do more disruption or more growth. 31:25 So think about a customer like NBA. 31:29 We've always said that brands are so hungry to take control of their own customers and relationships and communities. 31:38 And what they need is a managed community experience with moderation at scale. 31:42 They need brand-safe content controls. 31:45 They need analytics. 31:46 They need video infrastructure. 31:47 They need some boring enterprise checklist of compliance and SLAs. 31:52 And they need a single person to call when something breaks at halftime. 31:55 We actually have a lot of this in the atmosphere, right? 31:58 We have community moderation. 32:00 I hear it's coming. 32:02 Content controls. 32:03 We've got Fetica in the house. 32:06 We've got some amazing video infrastructure. 32:08 Thank you, Streamplace. 32:09 So a lot of these parts are here, but a brand or a customer like NBA doesn't know how to navigate this. 32:18 Like kelp forest of magical symbiotic experiences. 32:24 They need help. 32:25 They need either a tour guide or they need like a Blue Apron meal kit. 32:31 They just need some help, right? 32:33 Let's get them some help. 32:35 And that help doesn't have to come from outsiders. 32:37 That can come from cooperatives or revenue share opportunities or integrations from within the community. 32:44 But let's think more about how to work together to bring this shared streamlined benefit, right? 32:51 Because today we have a lot of technical options for folks. 32:55 I think we have a lot of building blocks, but they're, like I said, they're kind of like meal kits you have to cook yourself, and not every user is ready to do that or capable to do that or wants to do that. 33:09 But a more interesting one, because like the NBA is a little bit aspirational and they are probably incredibly picky and difficult to land, right? 33:17 Let's talk about something super boring. 33:19 Like I said, I'm kind of obsessed with these niche markets of experts. 33:23 The American Society of Anesthesiologists is 53,000 members in a professional membership org, and they make about $52 million a year in revenue. 33:34 Half of that is membership fees from their members. 33:37 The other half is in educational products. 33:40 And they have a private verified community for credentialed professionals. 33:44 You can imagine their education involves a lot of video or interactive discussions or community building or knowledge sharing. 33:50 They have a ton of compliance. 33:51 They probably have a terrible forum or a listserv or really bad conference websites that we can do better in the atmosphere. 34:00 And so if we think about these midsize customers and what they need, it's similar combinations of these building blocks, but possibly more attainable by us in the next year before the next ATComp. 34:15 So we are at part 3. 34:18 I wanna talk a little bit about the model. 34:20 And like I said, there is no one winner, right? 34:22 We just wanna match our model and align incentives. 34:25 If you have a feature, you can gift it to the world. 34:27 You can charge a small subscription. 34:29 You can use it to get hired. 34:31 You can fold it into a product. 34:33 With a product, you can sell the thing using a few different models. 34:36 And with a business you can chase revenue, equity, strategic investment. 34:41 And I want to highlight here— yeah? 34:43 I lost your audio, so I'm afraid they won't be able to hear you. 34:45 Oh, OK. 34:48 OK, are we back on audio? 34:51 Streaming. 34:51 Thank you. 34:52 I'll just try to project in the room. 34:54 When it comes to funding, I hear grants, I hear VC, I hear a lot of builders and founders saying I need to pick a path, right? 35:01 It's a fork in the road, I need to pick a path. 35:03 And I'm here to say that there are a lot more models, especially outside of tech. 35:08 Inside philanthropic models, you have mission-locked for-profit models. 35:13 You have a lot of cooperatives. 35:16 Some folks from the Haifa Worker Cooperative are here. 35:18 Some folks from Egalia are partnering with Blue Sky. 35:21 There are huge cooperatives. 35:23 There are billion-dollar cooperatives. 35:24 Look in Spain, especially. 35:26 This is what DAOs wanted to be, but actually they got beaten to the punch. 35:30 50 years ago in Spain. 35:34 There are concessionary loans and financing available. 35:40 The real estate sector has done a lot of this. 35:42 So look into those. 35:45 In the community in OSS, there's a few different kinds of models. 35:49 And then under the capitalist box, there's also— bootstrap and venture capital is what we talk about the most. 35:56 Strategic investment can be really interesting, especially as you get started, because it's a smaller player— a larger player funding you because you make their business stronger. 36:04 It can be gateway to acquisition, partnerships, revenue-sharing models. 36:08 So I just encourage everyone in this room to think a little bit bigger. 36:13 You don't have to reinvent your model. 36:14 I'm especially calling out to the funders in the room to consider some of these alternative models, because what we're building I think is more generative than some of the extractive models we've been talking about, but also maybe more creative than just straight grants. 36:33 That is it from me. 36:35 I hope everyone in the room got one thing out of it.