Subscribe
Share
Share
Embed
CRAFTED. is a show about great products and the people who make them. Top technologists reveal how they build game-changing products — and how you can, too. Honored twice by The Webby Awards as a top tech podcast, CRAFTED. is hosted by Dan Blumberg, an entrepreneur, product leader, and former public radio host. Listen to CRAFTED. to find out what it really takes to build great products and companies.
[00:00:00] Brandon Kessler: executives from the top down are saying We need to do hackathons to uncover the value of ai. It is not enough to have a high level AI strategy You need employee buy-in, you need employee tinkering and experimenting and iteration That's Brandon Kessler, the founder and CEO of DevPost, which power most of the world's hackathons... and there are more hackathons than ever. Why??
[00:00:23] AI has changed the landscape completely. Of the 1200 hackathons we powered last year, probably a dozen didn't have some AI component.
[00:00:30] On this episode... we explore what makes for a successful hackathon... why AI and prompt engineering are opening hackathons up to a wider range of people, and why hackathons are still so effective for companies looking to reach developers:
[00:00:46] developers are marketing averse. they prefer to build versus talk. and yet they are building our future.
[00:00:52] Welcome to CRAFTED., a show about great products and the people who make them. I'm Dan Blumberg. I'm a product and growth leader. And On CRAFTED, I'm here to bring you stories of founders, makers, and innovators that reveal how they've build game changing products -- and how you can too.
[00:01:12] CRAFTED. is brought to you in partnership with Docker, which helps developers build, share, run and verify applications anywhere – without environment confirmation or management. More than 20 million developers worldwide use Docker's suite of development tools, services and automations to accelerate the delivery of secure applications. Learn more at Docker dot com.
[00:01:33] And CRAFTED. is produced by Modern Product Minds, where I advise companies on product, discovery, growth, and experimentation. Learn more and sign up for the CRAFTED newsletter at modernproductminds.com
[00:01:45] There's not really an application layer there yet. That's what's so exciting about it.
[00:01:49] Dan Blumberg: our mission is to inspire developers to build. And so, um, when developers are bored, when they're looking to improve their craft, uh, when they want inspiration, we want them to come to Dev Post to find that next opportunity to be inspired, to make something.
[00:02:06] Brandon Kessler: Uh, we power a. Let's see, last year, 1200 hackathons, Those can be in-person hackathons. They can be virtual, um, they can be, um, private or they could be public. Uh, and a hackathon for those that don't know is when developers in these days, non-developers as well are getting together.
[00:02:26] Uh, and. Building software around a given theme, and usually there are prizes in terms of, you know, what's most exciting these days? Without question. It's, it's ai. AI is everything.
[00:02:37] what are some of the key questions that companies are looking to answer, particularly around ai since that's the flavor of the, of the year? Um, and and how do hackathons help them do that? Yeah. So, there's, there's two parties generally in hackathons. There's those that are sponsoring them, and then there are those who are participating in them. Uh, those that are sponsoring them often have developer tools, APIs, um, feeds of data, uh, software development kids. Uh, just developer tools, um, potentially data sets, and they wanna get those out to the world's developers.
[00:03:10] Uh, it's really tough marketing to developers. Uh, developers are marketing averse. They block ads in their browser more than any other profession. Uh, they prefer to build versus talk. They don't respond to marketing speak, and yet they are building our future. And so, and the ravenous for new tools that can make.
[00:03:29] Their lives more efficient, let them do things they couldn't do before. So hackathons are a great way for these organizers, these sponsors, to get their tools in the hands of developers. And then on the flip side, developers, they get to learn. They get to build something cool. They get a chance to get prizes and exposure, and they don't give up their intellectual property.
[00:03:48] Dan Blumberg: yeah. Just a a quick personal story. You and I met in, I think around 2011 at a hackathon, NYC big apps where I met some people I'd never met before, and ended up building something, met some folks from Twilio who were, you know, there and helped us build scene near me, which alerted you when you were near a, uh, a famous movie shoot using New York City data.
[00:04:06] the competition was. To do something clever with New York City data. Um, so I've, I've certainly seen the power of hackathons to, you know, to build something new with people I literally hadn't even met in this case.
[00:04:16] Brandon Kessler: Cool.
[00:04:17] Dan Blumberg: you said new types of people are getting involved in competitions now.
[00:04:20] Can you share more on that?
[00:04:23] Brandon Kessler: Yeah, I think it comes back to ai. Um, so again, the value of hackathons for, um, these organizers is to, to market their tools to the world's developers. and increasingly, um, inside of companies as well, where, uh, organizations wanna run hackathons. Inside their company to get their employees innovating and collaborating.
[00:04:47] That means learning new tools, that means solving business problems, et cetera. And this is new for us. in terms of non-developers. AI has changed the landscape completely. Of the 1200 hackathons we powered last year, probably a dozen didn't have some AI component. It has just changed what you can do with tools, developer tools, what you can do with your applications.
[00:05:09] Uh, it has changed hackathons specifically, and to your question, non-developers can now play a role. Uh, why? Because. A lot of AI these days is prompts, it's engineering prompts, and even the developers, the creators of these large language models don't even know a hundred percent how these work,
[00:05:29] So really iterating. Experimenting is key to unlocking the value of ai. And so it's not just developers, those are stakeholders, business folks, people who want to improve their own processes. So as it becomes much more low code and no code software moves on that inevitable trajectory. Um, then you have non-developers who are participating more than ever.
[00:05:54] And that's really fun to see.
[00:05:55] Dan Blumberg: I, I'm just, I'm looking at dev post's website right now just to give a sense of like what hackathons are available right now. I see Google AI Hackathon, Microsoft Developers, AI Learning Hackathon, meta Quest Presence Platform Hackathon.
[00:06:09] Um, another me, Microsoft Generative AI Hackathon, uh, Google Vertex.And if I look at, most recently added, uh, we have, the IEE, Ottawa Robotics Competition, uh, college of Staten Island Demo Day, et cetera.
[00:06:23] I mean, it's, it's a real wide range of stuff, but certainly, certainly focused on, um, that generative ai, what are we gonna do here and like, what do we have to understand this, uh, the potential, uh, of this new technology. what is it that makes a hackathon successful?
[00:06:40] Brandon Kessler: Yeah. and those are the public hackathons. Um, we also power these private hackathons within organizations. And if you talk to the average persons, you know what a hackathon is, many of them will say, yeah, my company has them. Engineers get to build what they want for a week, et cetera. Um, and the AI question, applies to everything, private, public, in-person, online, et cetera.
[00:07:02] So to your question, what makes a good hackathon? Um. It, it depends. If it's a public hackathon, then you, it's marketing, getting the word out. It's having a competition that's interesting to, to developers, businesses, et cetera. What does that mean? Interesting. It means there's a, there's a product behind it that's gonna make their lives better and more efficient, and there are enough of them who care about the genre.
[00:07:26] Uh, with ai, they certainly do, uh, if it's public, also prizes, um, which is, are they fair and is there tech support so that if you're building and learning and you get stuck, can you keep going? Will someone help you? so those are the things that make for great public hackathons. And these days we're seeing our highest numbers ever because.
[00:07:46] Really ai, Pivoting a little bit to these private hackathons within your company. I. some of those are similar. Uh, you gotta get the word out and you need buy-in, in your company in order to do it.
[00:07:56] These days, executives from the top down are saying We need to do hackathons to uncover the value of ai. Uh, these are directives from CEOs and CTOs. It is not enough to have a high level AI strategy that doesn't do anything. You need employee. It doesn't do anything alone. You need employee buy-in, uh, you need employee tinkering and experimenting and iteration and engagement to unlock that value.
[00:08:21] I know I sound like a broken record, but that's why, it's happening. So internally, what does it take to make a good one? Yeah, it takes buy-in, it takes a well-crafted hackathon, uh, our product, for a shameless plug it, it Cuts your time and have to organize them and it doubles the participation and helps you report on the outcomes.
[00:08:39] So that's why it's being used, because spreadsheets and wikis, you know, can be a mess. And so that, that's why dev posts for teams work.
[00:08:46] I wanna go back in time and then we're gonna work forward to how you came to launch, how, why you changed the name of the company, how you launched this new, this new product for, um, for companies to run hackathons internally. But I wanna start with, during college, I think you launched a music label, and so how did you go from launching a music label to becoming the CEO of, of Devpost?
[00:09:06] so when I was 13 I read about this guy David Geffen, who had a record label and I realized I could combine business and music. I love technology too, but those were the two things I was really obsessed with as a, as a kid. And I thought, wow, I kind of want to have a record label. And then the, I came to school.
[00:09:24] In New York, I started working these org companies, uh, like a dozen record labels and PR firms and this kind of thing. And then started a record label in my dorm room. And that means signing bands, getting 'em in the studio, overseeing the making of their album, promoting marketing, distributing it.
[00:09:41] Uh, getting them out of jail for cocaine possession and babysitting them and all kinds of, you know, cra crazy stuff. after doing that for 12 years, uh, and I loved it.All the creative people were, were moving into technology and that's what makes me, um, most excited is, is helping people make stuff, helping creators, uh, make stuff.
[00:10:05] And so there's a, pretty clear transition for me that I was going from helping musicians, but now I started seeing these developers getting together at these amazing hackathons with these. Competitive spirit, um, where they could get prizes and stuff. And, uh, I started calling, all of these grassroots contests that I started seeing Sprout up online.
[00:10:26] And there was no platform and there was no community. And so that's why I started the company at the time. It was, it was called Challenge Post, which is to be a platform and community for the skill contest, mostly engineering, but. Not just engineering. So that switch on how I did it, I just decided that it was time and worked really hard to kind of get it off the ground and, and raise money and convince our first
[00:10:52] Dan Blumberg: I, I know the, the Netflix competition in 2006 where they offered a million dollar prize for anyone that could give them a, I think it was a 10% improvement in their movie matching algorithm Back in the early days of Netflix was a, was a big one. That was certainly one of the first competitions that ever caught my eye.
[00:11:07] What, what do you think the ROI for Netflix was on that million dollar on, on how they're doing today?
[00:11:14] Brandon Kessler: Ye Yeah, a lot more. And, and that was an, uh, I don't know the exact percentage, but,
[00:11:19] Dan Blumberg: Many multiples. Yeah.
[00:11:21] Brandon Kessler: many multiples, because. You know, you put out this problem, you have lots of people who work on it. In their case, they pay, a big prize to the winners. Um, and, you know, competitive spirit gets people moving with hackathons, to be clear, it's not spec work.
[00:11:35] Um, you get to keep your intellectual property regardless of whether you win a prize or not. So that was key to what we're doing. But that prize model was exciting. I started seeing contests not just being about, you know. Scratch a lottery ticket and win a trip to Disney World or something like that. It was like solve real problems that matter, uh, that use technology and, you
[00:11:58] Dan Blumberg: Yeah. I'm curious how you got the courage to launch challenge posts and then what, and then what the first version of it looked like. I.
[00:12:07] Brandon Kessler: Yeah. I really admire people that can do 50 different things and be happy at every one of them, but I am not. That person. Like, I'm, I'm, if I'm not sort of doing the thing I'm, I'm really obsessed with usually for long periods of time, then I'm, I'm pretty miserable. So, when I decided that I wanted to pivot from the music industry and we were profitable pretty much every year and, and, and proud of what we, what we did there, um, I.
[00:12:31] I just knew that this was a cool idea. I, I was really obsessed with this and I, I was, let's see, I was in business school on nights and weekends, uh, here in New York, um, sort of incubating this idea and I would apply to business plan competitions and things like that. And I met investors that way and just kind of networked.
[00:12:51] I mean, I just remember it was 2008 also when the sky had fallen. There was, nobody was raising money. And I just, I mean, I remember. Taking trains to Boston to track down investors who were ignoring me. and just, oh, you're gonna pass on this investment, no problem. Do you know anybody? Can you give me two or three names?
[00:13:09] Uh, and you're just grinding, uh, and doing everything you can to kind of get it done. We had a minimum amount we wanted to close. We were able to do that. And then from there, fortunately we started getting customers, including the one you mentioned, New York City was
[00:13:21] Dan Blumberg: Oh, that's
[00:13:22] Brandon Kessler: much our first real paying customer.
[00:13:24] Dan Blumberg: Um.
[00:13:25] Brandon Kessler: Yeah. Yeah.
[00:13:26] Dan Blumberg: So move us forward. I know the Obama administration and challenge post worked together, and then it was, if I'm correct, it was not long thereafter that you pivoted again and sort of changed the name and the, the, the, the new exclusive focus on developers.
[00:13:37] I'd love to hear more about that.
[00:13:39] Brandon Kessler: one of the many, many, many, uh, product and business lessons that we've learned the hard way, you know, was that, If you're all things to all people, you're nothing and that you need to focus. So when we first started, it was clear that contests were cool, that people would pay to sponsor them, to solve cool problems, and people would enter them.
[00:13:59] But we weren't really sure, What would be the The key types of contests? And actually we proved that there was interest when first we're doing a tech competition and then the Obama administration decides. 'cause of those economics, you were saying that where people will, create a lot more investment than the prize money to solve real problems.
[00:14:20] They said we wanna have a platform for these prizes, these prize contests across all 80 agencies. And they put out like an RFP. And we worked really hard and we won that. So like within our first year, we were chosen. Maybe 18 months by the federal government to be the competition platform for all federal agencies.
[00:14:39] So we were like, forget focus. This is huge. Like we're, we're across every agency and it was great. We worked on many, many hundreds of these competitions, but they were business plan competitions and photo competitions, recipe competitions with Michelle Obama These were awesome, but they were all over the place. And so from a product standpoint, if you're all things to all people, your product's gonna be super generic. You can't delight software developers if you have to support photographers and chefs and business plan competition entrance, um, and your community is gonna be all over the place too.
[00:15:14] Imagine getting a newsletter with the next competition that's. And you're a software developer and it's a recipe competition. So, really all of our pivoting at the beginning was just saying no to people. So even after we had done this, we actually gave them back this platform. The federal government said, we're not doing this anymore.
[00:15:32] This is too big to do all things government. Um, we had already said, we'll just do software and government. Government
[00:15:39] Dan Blumberg: And it's a
[00:15:40] Brandon Kessler: is, it touches everything. So we said, okay, we'll do, it's a beast. It's not a, you're not narrowing down anything at all. If you say go, if, just say government, they're 82 different agencies.
[00:15:52] So we said, alright, just software. That's what we're doing. At that point, we changed our name to Devpost. You could be a government agency, you could be a nonprofit, you could be a business startup, big whatever. But we're gonna just do competitions that involve software. It's big enough.
[00:16:09] Dan Blumberg: let's talk about the, the, the, the private side. 'cause that that's the more recent product launch. I'm interested just to get into the weeds a little bit of like how you built it, how you tailored it for a super specific user.
[00:16:19] Brandon Kessler: the inspiration for this product was really the pandemic. we wanted a subscription model, uh, for, you know, revenue stability purposes and for growth. But, we get four or five inbound requests, every six months or a year, and it turned into 60 inbound requests from companies.
[00:16:40] For software to power their internal hackathons, uh, during the pandemic 'cause they were remote in there or they were a hybrid. And that just continued. And we realized that we could have a subscription product that would help support your internal hackathons and, that could support more than just the all hands hackathons.
[00:16:58] So if you have three or four designers, they want to come up with the. designed for their a t-shirt for an event. You can just with one click, hit a des a button and have a design, you know, ideation hackathon that companies can use and they can run as many hackathons as they want. You pay per seat, et cetera.
[00:17:16] In terms of the product, um, a totally private and secure. B, we don't have to worry about all the rules, the contest rules and all the marketing that you have to do with public contest and international contest law and these other things, and cheating and all of this kind of thing. It's within your organization and your employees.
[00:17:35] So we just made it very open. Anyone could launch a hackathon. Anybody could click one button and have that hackathon. That was a dream of ours. Um, to just be able to make it so simple. So I think from the, there are many, but from the product standpoint, the biggest thing is just. Giving people much more power.
[00:17:51] so there are some we can talk about. 'cause we have case studies They're, they're almost all large public companies. So SAP, Okta and Estee Lauder.
[00:17:59] You know, on Estee Lauder's standpoint, it was a top-down directive. Like, what does AI mean for our company? Let's get everybody involved and come up with ideas on how it can save time, save money, uh, delight our customers. Uh. It's really, really smart of, of them.
[00:18:17] And you might not think of Estee Lauder, for example, as a, as a tech company, first and foremost, but they employ a good amount of engineers, and they're one of the first companies to be pushing hard on what AI means across the organization.
[00:18:30] SAP is an interesting use case.
[00:18:34] Um, they're doing customer hackathons, so when customers are messing around with their tools or prospective customers are thinking about using SAP, um, they'll run hackathons so that those customers can get their hands dirty, uh, versus say a workshop where you're just. Watching a screen or watching a video, uh, they're actually, getting their customers or prospective customers to build.
[00:18:58] So we do customer hackathons with them. So, um, ultimately it's about inspiring people to build in an experiment, whether it's inside your
[00:19:07] Dan Blumberg: And, and, uh, we actually un crafted feature the former chief analytics officer of Estee Lauder. So it, it may be a slightly less unfamiliar to, to some longtime crafted listeners
[00:19:16] Brandon Kessler: Yeah. Very cool.
[00:19:17] Dan Blumberg: I'm curious on the, on the public side, over 15 years you've been running competitions. Are there any particular favorites that come to mind? I.
[00:19:27] Brandon Kessler: Oh yeah. the early data, big apps, competitions with New York City and federal government was really cool because, people are creating applications that help their city, like yours, and many others. Uh, really cool to see that. Love those,
[00:19:42] One of my favorites is, uh, Alexa, uh, owned by, by Amazon. Um, they have these things called skills that are like apps, and they had one called Kids Court, which was, uh, for those of you that have kids, when your kids are fighting, you can call up Alexa and say, and in each side gets to speak before a judge, and the judge actually makes a determination and there's a cross examination.
[00:20:07] It's incredible. So it's basically for kids to resolve. Difficulties between them, moderated by the judge that's called Kids Court. That that is a very, very successful app slash skill that we're, we're happy, that we're a part of, cool hackathons. Um, we do a ton of nonprofits. Worked with like the, the World Bank, um, and a bunch of companies to address things like the sustainable development goals.
[00:20:30] Um, those are definitely among our favorites because hackathons to do well, to do good. Um. Are, are, are quite powerful. And of course on the business side, we're fortunate to work with a lot of companies that have really cool tools and just to see companies get formed and launched, um, at hackathons.
[00:20:47] And we see it all the time. So it, it's rewarding. It's rewarding. It's nice to go show up at a hackathon and, we only have 25 people at our company, but.people think, oh, dev posts, you must have 2000 employees. I'm like, no, I'm actually coming from the post office, like carrying 500 packages on my back and,
[00:21:02] stickers and, that kind of thing.
[00:21:04] But it, it is rewarding to see the
[00:21:06] Dan Blumberg: yeah, I was gonna say the, the,
[00:21:08] Brandon Kessler: and the engineers and
[00:21:09] uh, the last hackathon that I participated in, I think I just joined the Slack community as a kind of, to answer questions if anybody had, but it was during the pandemic, I forget the exact spec specifics of it, but it was a pandemic. There was COVID-19 Global Hackathon. It was one of the largest hackathons in the world. Yeah. So Mark Zuckerberg announced it and it was, but it brought in Apple and Microsoft and Facebook and many other companies where they gave their engineers time to, um, come up with solutions that could in some way address the pandemic,from lifestyle to other things.
[00:21:41] And that was a really, really fun competition to
[00:21:44] Dan Blumberg: Over the years, I imagine that the hackathons have kind of mirrored whatever the top tech or the most, uh, future forward. We're not sure how to use this type of tech like we've been talking about with generative ai. I'm curious what some of those inflection points have been over the last 15 years and how they compare to today's.
[00:22:02] Uh, seemingly all-encompassing need to, you know, get smart about gen ai.
[00:22:09] Brandon Kessler: That's a such a good question. most of the time hackathons are about some emerging technology, something, it could already be big, but it's new and it matters. Um, and so certainly just cloud products initially, AWS, um, versus Microsoft Azure and Google Cloud and many others. Um, all with sort of, um.
[00:22:33] Amazing product, uh, improvements. Um, that's been a big one throughout, but I remember when it was really a, a big war. I think it still is for sure, between these cloud companies, but basically like whom you're hosting your, your apps and your products with your websites. So that, that was ver you know, big, um, certainly, uh, xr.
[00:22:54] So, you know, mixed reality, virtual reality, all that stuff is huge right now. Um, it ha it, you know, started with Oculus. Um, so a bunch of Oculus competitions, meta has continued with folks like Snap, and their lenses, and others. So that's certainly a key emerging technology.
[00:23:12] crypto and, slash Web3. there are a ton of hackathons. Uh, for a while it was, you know, close to half of the competitions we were doing, had some crypto Web3 component because there. All of those products are dependent on developer communities extending and building on top of whatever layers they're working on.
[00:23:33] And so, um, that was huge. It's certainly gone up and down, with really the price of Bitcoin and Bitcoin and Ethereum
[00:23:41] Dan Blumberg: those are some of the big ones that come to mind. I would say that AI and ML have been big for a long time. We've been doing, AI ML competitions for about a decade, but it was, you know, PyTorch and tensor flow. It was about optimizing, you know, uh. Specific point solution. Uh, and then of course gen ai, which, um, is, you know, completely different and has, as I said, you know, it's just, it's just really affected everything from how your company operates internally to how you can delight your customers.
[00:24:12] Brandon Kessler: But it's still in such early stages, like, there's not really an application layer there yet. That's what's so exciting about it. Yeah. There, there are. Prompt boxes and you can access them programmatically. And that's cool. Uh, and it's getting better and better. But we, you know, I think we all, we all see it coming and know that there's gonna be application layers, um, that do all kinds of things that haven't been done today.
[00:24:39] And that it's not just gonna be the big incumbents is certainly they will be involved in and, but
[00:24:44] the applications will be delivered by others.
[00:24:46] Dan Blumberg: What's next for Devpost?
[00:24:50] Brandon Kessler: So for us, so much of our focus is on dev post for teams, which is helping enterprises, organizations,drive innovation and collaboration. So it's DFT, this subscription product and getting it out there. Um, what's more challenging about getting the word out for DFT is that it's not public's like our others, where if you can go to dev posts, you showed it earlier, you know, you could just pull up any number of public hackathons, look at them.
[00:25:15] Say I want one of those, but with the fact that they're private and, a hundred percent private and secure, it's different. Um, you can't promote them in the same way. And there aren't hackathon managers, you know, that's not a title within an organization. These are, uh, people working on it who have other jobs, you know, the office of the CTO or innovation or other things like that.
[00:25:36] And so, you know. Growing it has been, it's rewarding, uh, and slowly but surely, you know, getting really, really terrific customers and impact. But that's the big focus for us.
[00:25:48] Dan Blumberg: I like to ask this question of, of, of all guests, um, and you're, you were just talking about a lot of emerging technology, so I'm really curious for your take. Um, what, what's something that sounds like science fiction now, but that you think is gonna be really commonplace in the near future? I.
[00:26:03] Brandon Kessler: I mean, this isn't, this is not original in any way, shape, or form, but, and, and it's here in some ways, but building applications by voice. Uh, I, I want an application that does X, y, or z. It pulls from a library of UIs, takes you through a little decision tree. Can can show you the demo. No, I want it to be this way.
[00:26:26] Okay. Here's three more options. Which would you like? I. And then Boom, can publish to the app store. Um, this is possible. Now, we actually had a really successful competition with AWS, uh, party Rock, which allows you to sort of type, create an app through prompts. So you can think of it's got a few prompts, like, oh, what is your app want?
[00:26:46] Do, I wanted to find, bathrooms in New York. Uh, great. And then you, you say a few more things and then it will sort of help give you that framework, but that becoming just. Commonplace. So the, everyone will be creating software. That doesn't mean that software developers will go away. That means there'll be a lot more of them, a lot more opportunities, no question.
[00:27:07] Brandon, thank you so much. I really, this is, I love what you guys are building. I, I love, I love looking into the future with you.
[00:27:15] Thank you so much for having me. I, I really appreciate it.
[00:27:18]
[00:27:19] That's Brandon Kessler and this is CRAFTED.
[00:27:22] Crafted is brought to you in partnership with Docker, which helps developers build, share, run and verify applications anywhere – without environment confirmation or management. More than 20 million developers worldwide use Docker's suite of development tools, services and automations to accelerate the delivery of secure applications. Learn more at Docker dot com.
[00:27:42] Special thanks to Artium, where I launched CRAFTED. Artium is, a next generation software development consultancy that combines elite human craftsmanship and artificial intelligence. See how Artium can help you build your future at artium. ai.
[00:27:57] And CRAFTED. is produced by Modern Product Minds, where I advise companies on product, discovery, growth, and experimentation. Learn more, subscribe to the show, and sign up for the CRAFTED newsletter at crafted.fm
[00:28:10] And please share CRAFTED with a friend... right away!
[00:28:14] forget focus. This is huge.