[Weekend Drop] AWS, Cloudflare, and Techbro Therapy on AWS.fm

Download MP3
I joined Adam for what started off as a discussion about AWS Amplify, my Cloudflare vs AWS article, and then ended up talking about being Extremely Online and how that helps and hurts us.
Listen to AWS.fm: https://aws.fm/episodes/episode-25-shawn-swyx-wang

Shawn joins Adam to discuss Amplify and its place in the developer ecosystem, whether we should care about Cloudflare, yet, and how to cope with the anxiety that can come with being extremely online. Also, it sounds like Adam is a tech bro and he's NOT happy about it.


Adam Elmore: Hey, everyone. Welcome to AWS FM, a podcast with guests from around the AWS community. I'm your host, Adam Elmore. And today, I'm joined by Shawn Swyx Wang. Hi, Shawn.

Shawn Wang: Hey, Adam. How's it going?

Adam Elmore: It's going well. I've been extremely excited. I've said this on a ton of podcasts, that I'm excited to get on with a guest, but this has been a long time because before I took my break, I was going to get on with you. Took a big, long break, and I've finally got you on. You're somebody, and I'm going to say a lot of things, I'm very dramatic, but you're somebody that I really admire in the online space. You have this ability to think about things, and distill them, and put them out there in a way that I admire greatly. I'm so excited to have you on here. It's going to be hard for me to stay on any one topic because I have just a list of questions I want to ask you, basically.

Shawn Wang: [inaudible 00:00:52].

Adam Elmore: First, could you tell everyone on this show who you are, just the short version of Shawn?

Shawn Wang: Yeah. So I'm Shawn, born and raised in Singapore, went to The States for college and then spent my first career in finance where I did investment banking and hedge funds. Loved the coding part because every junior finance person starts to learn to code, and didn't like the stress of the finance part, so I pivoted to tech where I was a software engineer at Two Sigma and then I was in developer relations at Netlify, AWS, Temporal, and I've just joined Airbyte as head of developer experience.

Adam Elmore: Oh, I did not know you weren't still at Temporal. So Airbyte, what is Airbyte?

Shawn Wang: Airbyte is a data integration company, it basically has the largest community of open-source connectors for connecting to any SaaS API source into your data warehouse. So for anyone doing data engineering, the first task that you have to do is to get data from all the different silos of data in your business. Let's say you have a Salesforce being the source of truth for customers, Stripe being the source of truth for transactions, get all of them into a single data warehouse for you to do operations on. So the goal is to have the largest community of open-source developers for connecting all the data and liberating your data from all the silos that you have in your business.

Adam Elmore: And how long ago did you start? How did I miss this?

Shawn Wang: A couple weeks ago. I actually have not announced it on Twitter, which is why.

Adam Elmore: Oh, there you go.

Shawn Wang: I like to slow play it. So when I joined Temporal, I actually waited for six months to really understand Temporal and to practice my pitch before announcing it on Twitter. And that's how I like to do things because, well, partially I want to be fully up to speed before I represent something publicly.

Adam Elmore: Yeah. So I want to talk about that. You get very up to speed in a way that I don't see a lot of people on Twitter. I don't see them understand things in the way that you do. So you obviously write, your blog is a huge source of information for me, and I've enjoyed it quite a lot, but it's not just that you write, it's the way you think about things. Does that come from your finance, your analytical background in finance, or were you like that before, your ability to see the whole forest, take in the way things are trending and the way things are moving, put it all together and distill it into these wonderful articles? Where does that come from?

Shawn Wang: Oh, so first of all, thanks for the very kind words. I don't hear back from my readers that often, so it's really nice when I get to talk to someone like this. So yeah, I would say a lot of this stuff is actually from my finance days. This is the kind of analysis that you would have to do when you do an investment report or investment research on any stock or any industry. You want to get a perspective of what's going on, what the trends are, who the major players are, and form an opinion on where things are going. And I think taking that finance mindset into the bets I have, in terms of technologies, whether or not it's for using them personally in my personal stack or for joining them as a startup employee, I think is extremely underrated. And it's something I'm trying to model and hopefully teach people someday.

Shawn Wang: Although I'm not sure about the teaching part, because if I say like, "Get rich by doing investment analysis stock on early stage startups," I would feel like a hustler. So maybe not that, but I just do like engaging in that. And probably it's an exercise for me to think things through clearly by writing it down. And I also get a lot of feedback from that, so I actually improve and learn a lot by learning in public. And that's the other thing that I am pretty well known for, so this is the application of the general purpose learning in public principle.

Adam Elmore: Yeah. No, and I love your learning in public article. I hope more people see how you break down systems and the world around us and distill it. I hope more people do that because I'd love to have more sources of that kind of information. It's really fascinating and that's a lot of what I want to talk about today is your opinions on the future and where certain things are headed. First, I want to talk, you did work at AWS. How long were you at AWS?

Shawn Wang: A year. AWS Amplify.

Adam Elmore: Yeah. So I'd love to know, I guess what it was like working at AWS, what you took from that, but also more broadly, I want to get into Amplify and where it fits. You sort of live in that intersection. I feel like web, and cloud, and infrastructure, where things are trending, and I want to talk Amplify's place in that, but first, what was your role there like at AWS, at Amplify?

Shawn Wang: Yeah, I was a senior dev advocate at Amplify, basically doing demos and talks for Amplify. And the fun thing about working at Amplify is that you are essentially also a developer advocate for all the underlying services. So amplify is essentially a roll up of DynamoDB, API Gateway, AWS AppSync, even file storage like S3. You could do some demos with that. And I did, I made like a DIY Dropbox clone. But it's focus on front-end engineers. And I think that was the first time that AWS had ever made a dedicated arm or products for front-end engineers. And it turned out to be a really good bet because AWS Amplify was one of the fastest growing AWS services, at least during the time that I was there. So I thought it was just really compelling to try it out and obviously everyone has very high regard for AWS. There's a bunch of services that I only experienced on the inside and I only learned about once I got on the inside, and I thought that was really interesting as well.

Shawn Wang: A few things I'll point out. I really loved the AWS interview process, actually. I felt like it was very rigorous and I definitely haven't had as rigorous a process anywhere else. And they really got a good look at every single part of me before they made the decision. And fortunately for me, it was a unanimous, good decision, but I felt challenged. I felt like there was a lot of growth that I took away from that process as well. So I highly recommend going through it, even if you don't necessarily take the job.

Shawn Wang: And once you're in, I think the other practice I really like was the weekly business reviews. Not everyone gets to be a part of, but I was, and essentially you have a P&L from the central AWS finance team that week to week tells you how well you're doing or not. And the PMs in particular, they'll put up highlights, they bring up topics of discussion, and the general manager would be grilling people on. And I thought that was just a fun way to run a business. It was a little bit stressful, sometimes a little bit dramatic, but hey, it forced you to take on the issues head on instead of ignoring them for three months to a year, which I've also seen happen.

Shawn Wang: So I just really appreciated that directness, and everything that you've heard about on the outside about AWS culture applies, like they'll send out the memo and the first 10 minutes of the meeting will be spend in complete silence where you just read the memo.

Adam Elmore: Just read the memo. Yeah, that's real. Well, what about the leadership principle? You talked about interviewing there. Did you feel like you started to embody those? Did those really become something you valued or was it sort of like, you're just doing it because that's what Amazon cares about?

Shawn Wang: There are a few things here. So I think one, people are drawn to Amazon because of leadership principles, like literally is what the interview is for. So you can't really join without already having them ingrained in you. And then second, yes, it gets brought up a lot when decisions are being made or just behaviors being modeled or discussed, especially in the performance review stuff. So I think that is useful, that is helpful, but at the same time I have problems with some of the LPs myself. "Be right a lot." What the hell is that?

Adam Elmore: So what is right?

Shawn Wang: Yes, exactly. What is right, what is a lot? So I think that, for example, what is underdiscussed or just not on the table, just because it comes from so much up high and has so much baggage and history with it, is that sometimes you have to try to be wrong, to take more risks. And being right a lot means that you might be more conservative than you otherwise should be. It leads to very incrementalist thinking, which is like, "All right, what is the most obvious next step? What is the low-hanging fruit? What is the short thing?" You just pick that over something that is more risky, but potentially has higher impact.

Adam Elmore: Yeah. No, that makes sense. I want to, I want to shift gears a little bit and talk about Amplify. Now that you're outside of AWS, you mentioned it was sort of the first example of AWS trying to go to the front-end developer and bundle up more of a developer experience. How do you feel? And you may have information from being there about traction and things like that. How do you feel about Amplify's return on investment and is Amazon doing a good job, I guess, with Amplify in terms of trying to package up their own experience? Do you see that resonating with developers?

Shawn Wang: So I think Amazon is doing a good enough job at addressing the needs of AWS customers. And that's something that is Prime first and foremost, like excels at that. Amplify could be doing a lot better at competing with the other standalone front-end developer focused startups that are out there that don't have the AWS infrastructure, which should help, but actually sometimes hurts it a little bit. So my favorite example of this is, so there's another company Begin, begin.com with Brian LeRoux. It's a four-persons company, and they also do very similar things. They deploy on top of Amazon, they are entirely serverless, they have a smaller set of offerings that they have, but their deploy speeds are in order of magnitude, faster than Amplify. They can deploy faster to AWS than Amplify can.

Shawn Wang: And that's because Amplify doesn't do some of the trickery that they do, like having a cold pool ready or anything like that. When people are not married to the AWS stack, just because that's the solution, that's the technology provider or cloud that their company has picked. When you have free choice, then you come with no baggage and just being from AWS doesn't give you any home ground advantage anymore. Therefore, you have to really, really, really compete on developer experience. And that's something that Amplify still needed to work on at the time that I left.

Adam Elmore: Yeah. I'm glad you brought up Begin too. I'm curious how it fits into the landscape. I've seen you mention Begin within some of your articles, like the cloud distros article I think about, I want to talk about that, but how is Begin doing? I interact with Brian on Twitter, I generally like him a lot, I like what they're building, but it is sort of a thing you have to buy into. It's like a whole different way of building applications. Do you have any sense for how they fit as a player in all of this?

Shawn Wang: They're tiny. I mean, they're not a rocket ship by any means, but they absolutely solve the problem for the serverless full stack minimalist aesthetic that they're going for.

Adam Elmore: Those are all things I like, so.

Shawn Wang: Right down to the API calls, having an inbuilt authentication solution that when you write the serverless function, you just have the user ID and it's all done for you with cookies in the background. That's just beautiful, that's [inaudible 00:12:58] mess with cognito or anything like that. Because it's very straightforward, that is the way that I would want to build serverless applications. If I didn't have some kind of big enterprise thing requirement, which maybe it's a premature optimization to try to glom that on in the first place, which is what you're required to do with AWS Amplify.

Shawn Wang: So I don't think I have enough experience to really judge, are they the right technical choice in all aspects? But I think there's just a certain aesthetic that you try to optimize for. And if you have full stack needs, if you like serverless, if you like one of everything, essentially one story solution, one queuing solution, one database solution, then Begin is the right curation for you. And then Amplify is sort of the more fully loaded solution if you want an easy way to access, let's say API Gateway, even like the... Actually just before I left, they actually launched support for serverless containers with a AWS Fargate, which is also super interesting.

Adam Elmore: Oh, I didn't even know Amplify supported that.

Shawn Wang: Yeah, exactly. They're just different trade offs in the spectrum, like Begin is way more opinionated than Amplify. Amplify is way more opinionated than the full set of AWS services that are possibly out there. I think they serve front-end developers well in all different respects. Yeah. I think Amplify is definitely hitting its goals and probably exceeding its goals for adoption internally. Begin could do a better job at marketing and something that I should probably try to help them on just because I'm a friend of the company and so, I mean, I just really like the philosophy, but at the same time, there are other competitors out there, like CloudFlare Workers is essentially trying to become a Jamstack or a backend-as-a-service platform, because they have Workers KV and Durable Objects. And that's a very compelling solution for a particular type of audience.

Shawn Wang: And it's weird because you have to be much more specific now. Like that's the thing, you have to figure out which part of the population you are in, in order to figure out which provider is best for you. There's no such thing as one provider fits all. It's really about like, "Okay, do you like the minimalist approach? Go with Begin. Do you like the edge-first approach? Maybe go with CloudFlare. Do you like the little bit more full stack, scalable, cloudy service? Maybe go with Amplify." There's a lot there. Like, "Do you like to self-host containers? Maybe go with Fly.io or Render.com. There's just a lot of options out there, but all of them happened to be built on top of AWS, which is why we had the cloud distros thesis.

Adam Elmore: Yeah. And I've consumed a lot of your content on that front, like hosted back ends. I do wonder where it's all headed. Maybe the answer is that there's just going to be a lot of options, and because there's a lot of different use cases, I guess maybe narrowing it down. Like if I really don't care about enterprise stuff or big teams, if I just care about building stuff with small teams, startups, that's where I live. Do you have any predictions, I guess, for where ideal product building is headed? Is it hosted back ends to go with your hosted front ends on Vercel or whatever else? Is it learning AWS primitives and just good and good at building stuff? How do you see that forecasting into the future?

Shawn Wang: What's the alternative to hosted back ends?

Adam Elmore: I guess what I do right now is build... Like I kind of use all the Amplify services, I just don't use Amplify. So I build a lot of bespoke APIs with AppSync, and Dynamo, and whatever.

Shawn Wang: So because you have that knowledge, that's the best thing for you, because you already have that knowledge. Like it's not a big deal for you to spin up another service, but for others it would be, because they would be new to that and sometimes a more friendly layer that abstracts it away for them would be helpful. So it's really hard to say which is going to win just because they're all going to win in some way, but some will be more winning than others. That's kind of how I view it.

Adam Elmore: Yeah. Yeah.

Shawn Wang: Because at the end of the day, like cloud is such a big deal, it's such a multi decade thing. It's going to take the rest of our lives to play out. That means that the vast majority of users of cloud haven't adopted it yet, still. This late into the game, they still haven't adopted it yet.

Adam Elmore: It's so hard for me to wrap my brain around. It seems like it's been so long. And when you say the rest of our lives, I don't put it in that kind of perspective. I need to calm down trying to figure out what's going to happen in the next three years. Like it doesn't matter.

Shawn Wang: Yeah. Yeah. Lambda is like seven years old. This is so early. The way that this looks 40, 50 years from now is going to be so different. AWS has like a million-something customers, imagine it having 10 million. When you have order of magnitude, when we start to think in terms of orders of magnitude, you start to really sweat the small details a lot less because you're like, "Whatever. Everyone's going to win."

Adam Elmore: We all win. Yeah, I guess it's true. I don't know if you've talked about this, I'm sure you've thought about it, and maybe you have written about this, but it's the idea of scarcity versus abundance mentality, I guess. It's weird because all at the same time, I agree with the sentiment that if you're on Twitter or you're very online or whatever, you should have this mentality that we can all lift each other up and we can all succeed. But then on the other hand, you've got the climate and how much can the earth sustain in terms of everything can only grow so much. I just had that thought, that sort of raw stream of consciousness. So I don't know if you've got any refined response to that. Is that sort of totally different concepts that I shouldn't conflate?

Shawn Wang: What, the limits to growth thesis?

Adam Elmore: Oh, yeah. I guess that's what it's called. See, I knew you'd have a name for it or something. Like the idea that we can all succeed, but at the same time, we all need to do a lot less because the planet can't succeed if we all...

Shawn Wang: I mean, this is about the offline-online shift. So we can still do a lot less and cloud can still grow because the mix of what we do in-cloud versus off-cloud is still very much imbalanced. So when you do things like pay attention to an Andy Jassy Keynote, and he'll talk about like, "Oh, cloud penetration is whatever, 20%, 30%." That is how low it is and it still takes a long time for people to adopt for whatever reason, institutional or just generational, or maybe our technology's not there yet. There's still a lot that needs to be developed to serve all kinds of markets that it hasn't penetrated. My favorite stat was that online shopping went from 10% to 20% in COVID.

Adam Elmore: I can't believe it's only 20%. That's actually...

Shawn Wang: Exactly, right?

Adam Elmore: That's bonkers.

Shawn Wang: So there's some version of the future where that is 70%, which means that you still have a long, long, long, long, long way to grow for every part of e-commerce and the planet can still win by maybe more efficient sorting or less retail outlets. I don't know. I don't know about that. I think I'm much more shakier ground there, but yeah, often the online transition, I think it is a very positive thing for the planet, especially because a lot of the major clouds are committing to net zero carbon footprints. I'm not sure if AWS has actually done that yet, but definitely Microsoft and Google have done it, which means AWS will eventually do it.

Adam Elmore: And I know AWS, they've launched sustainability insights and stuff recently, where you can start to see the emissions impact of the services you're spinning up. I know Google's done that for some time, but AWS is now doing that, I think.

Shawn Wang: Right. But we're actually measuring it now versus not measuring it before, so whatever. This is peanuts compared to like, "All right, are we moving to electric vehicles or something?" That is way more of an interesting concern than this stuff. Like invent a better battery and that will drastically accelerate the move to solar, and that will be much more meaningful than choosing paper straws. Sweating over the carbon footprint of your EC2 instance is the developer equivalent of choosing a paper straw. Really, look, I appreciate the effort, the spirit's, the heart's in the right place, but really if you want to make an impact, go work in the big things.

Adam Elmore: I'm glad you said that because this is not on my notes, this is not something I planned to talk about, but this is the thing that I feel like to make an impact, I've really struggled, I'm 15 years into my career, I've been like a software engineer mostly early in my career, then I did a startup, and then I've mostly just been doing consulting. I feel like there are more possible things I could do with my time than ever. And it's so hard for me to decide what is worth spending time on.

Adam Elmore: And I guess, do you have any thoughts on senior engineers, when you get to a point in your career where you have more flexibility and more opportunities, what is the most impactful thing? I've thought about making courses, I've thought about building products and just continuing with consulting. Is there a way to split your time that you're ever going to feel good about?

Shawn Wang: Probably not.

Adam Elmore: Okay. It's good to know. I can stop trying to find it.

Shawn Wang: Yeah. The menu options is so high. I think just figure out what gives you energy and then try to spend more of your time and day on that than stuff that takes away energy from you, so it was just a very hippie thing for me to say.

Adam Elmore: Yeah. No, that seems much simpler than I'm making it.

Shawn Wang: There's a concept here that I do like to share about leverage. There's an inherent tension between productivity and leverage. I think we are trained from basically our days in school, that high productivity is the goal, which is you want to have a packed calendar, you want to be doing eight different things at once. You should feel bad if your efficiency went down 10% compared to last week or whatever, and you're not meeting your OKRs or whatever. And the exact opposite to that is leverage where you want to have one thing, you want to do one thing and just have a lot of impacts come out of that.

Shawn Wang: And I think there's a movement, at least in VC circles, but also in sort of tech bro circles of waking up to the idea of slack in your life, and having peace and not having so much going on, and just doing high leverage activities that help you extend your reach without you necessarily putting more hours in or being super productive. Like being unproductive is fantastic. It's actually people who cannot figure out leverage who have to try to be productive. If you can figure out leverage, then productivity doesn't matter at all.

Adam Elmore: Yeah. No, that's good stuff. I think I intuitively knew that. I just have a really hard time. I feel like I'm much more seeing the tree versus the forest, so I really appreciate talking with people like you that see the broader picture. I think I have a lot of thoughts and then I read an article of yours and it helps me put words to those thoughts that I couldn't really formalize in my head.

Shawn Wang: I should really write about this more, but I feel like I haven't got it yet. You see me out there, you see me doing all sorts of random crap. So I haven't internalized it fully. I haven't let go of the sort of productivity mantra. Part of that is me being very risk-averse, part of that is me being doubting myself. Definitely, the stuff that you see from me has extremely high leverage. I think, okay... The other thing is I also have second thoughts or doubts about this whole leverage thing, that's why I have a very divisive tone about VCs and tech bros, because everyone wants to be high leverage, everyone wants to do the 80-20. Nobody wants to ship stuff, they just want to tweet thoughts, and then they think they're done. Right?

Adam Elmore: Yeah.

Shawn Wang: That's what they think high leverage is. But really the people who get shit done, swipe to find details and take things to the finish line. And guess what? Doing that last 10% is super low leverage. Like, "Oh man, I got to fix this stupid SEO description or the OG image isn't right, let me go fix that." That kind of small little details matter for the quality of the products and for shipping things, but all the high-leverage people feel like they're above that because it's not a good use of time.

Adam Elmore: So are they the high-leverage people or you're saying the people that want to be high leverage, is that the VCs and the tech bros?

Shawn Wang: Yeah, exactly.

Adam Elmore: What is tech bro? I feel like I probably am a tech bro, and I don't want to be a tech bro, but I feel like I'm a white male that has a podcast, so I can't escape it.

Shawn Wang: Yeah. Yeah. I'm a tech bro guy. I'm sort of reluctantly in that demographic. Yeah, the tech bro is a bro that's in tech.

Adam Elmore: Okay. Yeah. Well.

Shawn Wang: That is fully aware. Okay. I do like to have this mis-metric. If you're fully up to speed on the latest news, the gossip, you know all the new launches and new products, you're definitely a tech bro.

Adam Elmore: Okay. Okay.

Shawn Wang: If nothing surprises you, you're a tech bro. If you know what AUM is, if you know what ARR is, if you know all these acronyms without even blinking, you're a tech bro. Well, the real people who get shit done out there are wonderfully blissfully ignorant. They'll be like, "What is this whole Twitter kerfuffle, what's going on? I don't know. I just completely stayed out of the loop." But you being a tech bro, you would know the blow by blow of like Elon did this, twitter did that, Elon did other thing, twitter did other thing. It doesn't matter, the stuff doesn't matter to some extent and tech bros are so involved in their own filter bubble that they don't see their own forest for the trees, so.

Adam Elmore: You said Twitter. I think I've been on Twitter actively for a year or so and I don't know that I'm better for it. I don't know that like... I know that I'm very influenced by that sphere and sort of feeling like, I think that's why it's so surprising to me when I hear about cloud adoption or I hear about online shopping. It just seems like everyone lives in this little community and it's very easy to just not really remember the people that are actually around me in my local community and what life is actually like. Is there a way to balance it? Is there a way to balance being very online, being a member of this Twitter community and still keep a good grasp on the real world?

Shawn Wang: I don't think I personally have figured that out a lot, but I think it's basically the developer equivalent of go touch grass, which is go outside.

Adam Elmore: Yeah, yeah, yeah.

Shawn Wang: Have hobbies, have kids.

Adam Elmore: That I was going to say, I've got two boys and they make me be outside a whole lot, so that probably helps, I guess, somewhat.

Shawn Wang: Yeah, yeah, yeah.

Adam Elmore: I think the biggest thing for me just career and in terms of the always online, the tech broness, I think giving my wife the opportunity to set some boundaries around the time that I am working, I think this stage of my career, I've been able to say I'm going to work less and just seeing her role and what her life looks like and realizing how it shouldn't be this different. Like we shouldn't have such a, I don't know, huge chasm in terms of our daily life. Like I get to go enjoy what I do all day. Yeah, that's helped. We've carved out a lot of time that's like, "This is time for family." I think yeah, but my online, my work life feels very homogenous, I guess. And it could be better.

Shawn Wang: For me, it's like, "All right, figure out what is probably going to make your money and focus all your attention on that. Ignore everything else. Try to stick to, okay, what can you reasonably explain to your non-technical relatives? If you can't really justify it to them, then maybe have a second thought about like, 'All right, what am I really doing here?' Am I really making the world a better place by inventing a better form of infrastructure as code? Probably not." Unless you become a billionaire by creating HashiCorp, right?

Adam Elmore: Yeah, I guess it happens in that very rare instance. Yeah.

Shawn Wang: Right. But it can happen. You just have to be super clear on what you're trying to do here. And just like, yeah, be super intellectually honest about like, "Look, you're you're in this for the money, whatever you work on is probably going to be irrelevant in 10 years anyway. It doesn't matter, but you're at least going to have fun, you're going to build some relationships, you're going to make some people happy, create some jobs, whatever, and then spend the rest of your time with family and friends."

Adam Elmore: That was a very succinct way of wrapping up a lot of the things I needed answered. So I don't know if anyone that listens to this podcast cares about any of this. I really appreciate the conversation we just had.

Shawn Wang: No, no. I think yeah, this is very real and I really appreciate you bringing it up, because I don't get a lot of chance to talk about this.

Adam Elmore: Yeah. No, I live in the Ozarks, so tech literacy here is super low. I think that's where getting into the Twitter community, it was like, "I have friends now that I can talk to about technology and things I care about." But yeah, finding that balance. I think it's really very practical of you, very wise of you to point out that ultimately this stuff doesn't necessarily matter in a decade, that whatever I think I'm working on that's so important is probably more about the people, more about what I'm kind of enjoying the process along the way and that it's making a living and that we're moving a little bit forward whatever parts we touch and what other people we can be involved with. That was very nice for me to hear.

Shawn Wang: I will point out one thing. So humanity is kind of moving onto this metaverse. If there's anything that's actually real about the metaverse is that you have your community online that is dissociated from your physical community. You're so into AWS, or cloud, or anything like that, and no one else around you physically is, and it's fine. And this is something that actually the crypto bros, they probably got right. So I think Balaji Srinivasan, who is one of the crypto investors at Andreessen Horowitz, he released this book recently about building a digital nation, which is really compelling, which is like, essentially there's the world of physical nations, like the ones that country that've boundaries, but then there's the digital nations, which are formed online, and you're a member of the digital nation of probably tech Twitter, whatever.

Adam Elmore: Yeah, yeah.

Shawn Wang: Or AWS Twitter. And I kind of liken it to the difference between friends being the family that you choose versus the family that you have is the one that you're born with.

Adam Elmore: Yeah, yeah, yeah, yeah.

Shawn Wang: So where you're physically located is just the nation that you're born with or the nation that you have to live in for your family reasons, but the one that you do online, that's the nation that you choose, so you're member of a different nation online. And that nation is global, it's ephemeral, it's virtual, whatever that is. But it's something that you prefer to spend your time in as compared to your physical nation.

Adam Elmore: Yeah. So I feel like since getting really active in Twitter and being involved with the AWS community, even outside of Twitter, it is so global. It's helped me see the perspective of America, where I live, so differently. Just getting all those other points of view and just knowing that when I interact with someone, it's not this base assumption that they understand the world through the lens of America like I do. I very much appreciate that. I feel like I'm, if anything, becoming more and more dissociated with the country I physically live in, because I just don't interact much with people outside of these walls. I don't know if it was COVID and being in all the time. I always have been kind of an at-home person.

Shawn Wang: So that is dangerous. Right? That is dangerous.

Adam Elmore: Yeah. It feels dangerous. Yeah, tell me why.

Shawn Wang: Well, because if you don't care about the physical environment that you're in, then it's going to degrade, it's going to diverge away from your preference.

Adam Elmore: Yeah.

Shawn Wang: I don't know if that's inherently bad to me. Like there's definitely a physical element to humanity that we should keep around. We are not just brains plugged into the matrix. Essentially this leads to the matrix, that we might also just be plugged into something virtual online and spend zero time on a physical environment. Most people would not like to live that way, and that means we should care about what's going on around us. And we should try to have some physical presence that we're actually proud of and enjoy. And I think that there's a tension there that I think is sort of the modern humanistic existentialism, which is like, "How much of my life should I spend online versus how much should I spend in person?" And the fact that you have to choose is just nuts.

Adam Elmore: Yeah. And I think my problem, like if I'm just being honest with myself and just thinking through this, I spend about as much time, I think, in the real world, but it's just with my family, at home, it's with my neighbor, I got a neighbor that I go for walks every week with. It's like my very, very hyper local community. But what's going on in the City of Nixa? It's like 10,000 people where I live. What's the local government doing? I don't know. I have no idea. What's the State of Missouri doing? Probably stuff I don't like.

Shawn Wang: Exactly. And look, this has a very real impact on us because these people are making the laws that we have to follow. And we don't have a voice because we choose not to have a voice because we choose to not care. But hey, is it really our fault when the Supreme Court or the Congress makes a law that we don't like? Well, yeah. I mean, what did you expect? You didn't spend any time investing in that part of the world. It's like, "When are we going to have a software engineer in Congress?" That's really the big question.

Adam Elmore: Yeah. There's not a lot of tech representation, is there? In government in the United States.

Shawn Wang: No, because everyone hates politics, they love to dunk on it, they don't want to do a thing about it, but that's kind of the problem. I don't care which side of the bench you're on, like just the politicalness because you feel like you're not a member of the physical nation, you're a member of the digital nation. That is a problem for the physical nation, because at the end of the day, that's basically a reality.

Adam Elmore: Yeah. Oh, I think of that, there was that Netflix documentary. I don't even know if it was just on Netflix, but there was that social. Well, I don't even remember what it was called, it was about social media and had all these people from Facebook and other places, or ex-Facebook, talking about just this impact that the very online nature of our generation, what it's doing to our brains and all that. This all sort of ties in my mind. Like I definitely need to do some more things that are yeah, going to impact my life, my kids' lives, sort of being more involved, I guess, outside of... Like I divide my time into I'm at work and I'm on a computer all day or I'm with my family and we're out in the yard playing. It's those two things. And I make no time for anything else, but that's probably not good. Not a good, long-term solution.

Adam Elmore: Okay. Now I'm getting way off the rails. AWS FM, people literally listen to this for some good AWS bits. They've turned out long ago. I do have a couple more questions here, getting back to like I'm a developer, I like building full-stack web applications and I happen to like leveraging AWS. I'm going to ask you a few things. When should I care about CloudFlare? They announce all this cool stuff and it really is genuinely cool sounding, but there's so much of a barrier to adoption, like for me to change my day to day and start using a new thing. When should I care about CloudFlare?

Shawn Wang: I have the article on this, about how CloudFlare is playing Go while AWS plays chess, so I highly recommend reading that up. Essentially, CloudFlare is a really good CDN. AWS has its own. I would think you can do up comparisons of CloudFront and CloudFlare all day long, but I would say that CloudFlare probably has much more of a security focus than CloudFront has, and that by default wins you the majority of the business and it happens to be very easily adoptable because you just need to configure some DNS, just is carrying a lot of weight there and it comes to DNS.

Adam Elmore: If you're asking someone in the Ozarks around me, then what's DNS, first of all?

Shawn Wang: So I think it basically starts from the outside in. You want to think about CloudFlare, you think about where your user's traffic is coming in. Maybe you want to protect those with CloudFlare and then you want to come in a little bit. CloudFlare has this S3 wrapper called R2, that basically reduces a lot of your outgoing bandwidth costs. And that seems like basically a Pareto optimal win. Pareto being you're no worse off in any dimension and you're better off in one dimension, which is cost. And that's just a function of CloudFlare.

Shawn Wang: Like how many points of presence does AWS have? I think in the hundreds, maybe 100, 150, something like that. CloudFlare has tens of thousands, right?

Adam Elmore: Oh, okay.

Shawn Wang: It's just a much better edge network than AWS has. And so they just have a fundamentally different business model. And I think once you understand that from a fundamental physics and points of presence perspective, then you're understanding, "Okay, this is what I'm getting that AWS doesn't do." It's not a straight up one-to-one competitor, it's trying to tackle the cloud problem from a different way.

Shawn Wang: So you do the cloud traffic protection, then you do the sort of egress charges, which are sort of the main sticking point of AWS. Then you get into the extra stuff that CloudFlare offers for application builders. And I focus on this because I'm an application builder. CloudFlare's other offerings for security that I have no idea, security and networking that I have no idea about, particularly if you need to wire a building or an office, they have a box that's pretty sweet for everything I heard. CloudFlare One is the name of it if you want to Google it.

Adam Elmore: Okay. Yeah, I do.

Shawn Wang: But for application developers, CloudFlare Workers, that team is the sort of primary team that's working on that. And that is, there's edge function service that would be a big leap to adopt because they don't run Node.js, they run V8 isolates, which are taken out of the Chrome V8 engine.

Adam Elmore: Is it similar to like Lambda@Edge? Like the same kind of...?

Shawn Wang: No, it is not.

Adam Elmore: Oh, is Lambda@Edge node?

Shawn Wang: Yes.

Adam Elmore: Oh, it is.

Shawn Wang: Yes.

Adam Elmore: It is. Now, what is it similar to? It's similar to, I guess like Middleware and Next.js, that's that same kind of a limited runtime environment?

Shawn Wang: I think so. Yeah, exactly, exactly. I would say it's more limited in Lambda@Edge and it's got different costs and criteria. Basically, there's just more of the open source ecosystem that it will be incompatible with CloudFlare Workers than it would be with Lambda@Edge. And that's the thing that you need to know because you're going to use...

Adam Elmore: CloudFront Functions.

Shawn Wang: Ah, okay. Yeah, that's the one I keep forgetting.

Adam Elmore: I don't know who's using it, but that's what I was thinking of.

Shawn Wang: Right. So I used to use this only for smart redirects, like looking at the headers of a request and saying, "If you're coming in with a header indicating you're from a certain region, certain IPS, certain language, then I'm going to route you to a different location than I would normally." Only for route, but now Edge Functions are becoming so capable that you might be able to do rendering on demands instead of just routing. And that actually is unlocking a few new things because on top of that, CloudFlare also has persistence solutions with Workers KV, which is their eventually consistent store, and Workers, and Durable Objects, which is their strongly consistent store. So either one of those combined with the ability to render, means that you can actually just host a site full stack with Front on the Edge. There's no origin server, there's no region, you just have everything everywhere all at once, which is a favorite phrase that I try to sneak in.

Adam Elmore: Yeah. That's super compelling.

Shawn Wang: So yeah, your latencies go down from like 300 milliseconds to nine, just because you're just pinging near a cell tower or something.

Adam Elmore: Yeah, that's incredible. And they've just announced, I don't remember D1 or whatever. I don't know, I can't keep track of their product names, but they have like a distributed SQL offering as well that's coming or...

Shawn Wang: SQLite. Yeah.

Adam Elmore: Yeah. SQLite at the edge.

Shawn Wang: I mean, everything's just built on top, it's just clearly built on top of the original persistence primitive that they have. And so once they got strongly consistent and eventually consistent, those are the two dimensions that you really care about. You can build any sort of solution on that, so the SQLite offering is just built on top of that.

Adam Elmore: Yeah. Okay. So I don't know if I'm going to like jump on this stuff yet, but it does sound like there is a world where I could build side projects just on CloudFlare, like stuff runs all at the edge and I don't have to build up, I guess, is the interop, like if I want to still stand up a GraphQL API in AWS, like AppSync or something, is there interoping between the two services? You said their durable storage sits on top of S3, so it's actually, you're using an S3 bucket, you're just wrapping it with a CloudFlare thing?

Shawn Wang: It's a proxy.

Adam Elmore: Okay. Are people building hybrid CloudFlare, oh, I know they are, hybrid CloudFlare and AWS back ends today? I think I know of a couple at least. Is that a thing you recommend?

Shawn Wang: I would say yeah, there are. I'd say this is definitely on the cutting edge. You do it because you feel like [inaudible 00:42:35].

Adam Elmore: It's like Twitter, where you do it and you talk about it on Twitter and then everyone thinks...

Shawn Wang: It's theoretically possible, it's just like probably not in any size.

Adam Elmore: Doesn't make sense yet. Okay. So I'm going to say, I don't need to care about CloudFlare yet, that's what I'm going to say based on this conversation. I mean, I'm going to keep reading the articles, but.

Shawn Wang: The only thing I'll point out is don't stop there because this is what they've achieved in the past three, four years, they clearly have a roadmap, they clearly are going to keep going, and just eating the cloud from outside in, which is the name of the article. What else of the functionality can be replicated in an-edge-first way? CloudFlare is probably going to do that. And so there's a whole roadmap that just consists of looking at the AWS console and just going, "That first, that first, that first comes [inaudible 00:43:17]."

Adam Elmore: Yep. Yep. Yep.

Shawn Wang: And then there's a question of just what kind of application are you building and do you really need the full set of AWS services, or can you just start from the edge first? That's how disruption happens. Disruption happens by taking a section on the market that nobody cared about and making that your entire thing, and then making it so capable over time that people see no use to use the old thing, but it takes a course of what, 10, 20 years to do that because AWS has just spent the past 20 years doing that in the first place.

Adam Elmore: I just don't keep those time frames in mind. Like Twitter has warped my sense of when things are coming. And when you say 10, 20 years, it's like, I don't think about anything that's coming 10, 20 years from now. I think I'm thinking what's coming in the next 18 months.

Shawn Wang: Right. But that's a problem for us, because that short-term mentality stops us from betting on big trends early. And I think to build anything of significance, you have to do it for 10 years.

Adam Elmore: Yeah. I got to get off Twitter, that's what I'm coming to here.

Shawn Wang: I think so. I think I'm going to do it in healthy amounts. So I actually, one of my longstanding wishlist projects is to actually build a Twitter client that has a time limit.

Adam Elmore: Oh, nice. Yes.

Shawn Wang: [inaudible 00:44:25] Client with a time limit. If you're going to have more time, you're going to have to pay to donate to your favorite charity or something.

Adam Elmore: Oh, I love it.

Shawn Wang: And that's in my wishlist.

Adam Elmore: Yeah. I will use it. You've got your first user if you build it.

Shawn Wang: I'll just say the only reason I don't do it is because nobody trusts the Twitter API.

Adam Elmore: So one more, should I care about it yet or not? Because I see Brian LeRoux talk about this quite a bit. Deno. Should I care about Deno yet?

Shawn Wang: I think so. I think it's there. I think it's there. So what is Deno? Dino is sort of the new runtime that the original creator of Node.js is saying, "All right, I'm going to do this over. Node.js has been around for 10 years. I see all the flaws of it, now I'm going to start over from scratch." I was very skeptical of Deno when it first came out, but it's been two years and it's really shown a lot of progress. And I think the governance is right, the funding model was right, and the adoption is growing. What is really compelling to me about Deno, just not from a technical perspective, from a business perspective, which feeds into a technical, the business side. There are companies so Superbase and Netlify, both launched edge functions powered by Deno, which means that their biggest products shipping capability announcement of the year of 2022 was someone else's product. It was a startup that's way younger than them, but they just have the right abstraction and the right cloud service that is already functional that they're launching. So it's weird.

Shawn Wang: Deno's go-to-market strategy is just waiting for other people to wake up and go, "I need this. Deno's the only supplier in the market for this. And yeah, let's just bring it on and ship it as our thing." Where it really is Deno's thing, but they're just letting other people white label them. It's that's fantastic. So I mean, from that perspective alone in the past six months, I've really changed to, from like, "Okay, Node and Deno will coexist for the foreseeable future because there's such a huge install base of Node into every incremental app will probably be built in Deno."

Adam Elmore: Well, that's... Yeah. No, that's what I needed to hear. I think I there's a lot of excitement. I see it all, but it's all Twitter, so I needed to hear it face to face that it's worth digging into.

Adam Elmore: One last question. We do have a couple more minutes here. Do you have thoughts on the whole macro venture capital situation and how that might impact the next 5, 10 years? And I don't know if we're entering into some tightening cycle that we've never seen anything like the last 10 years, 13, whatever years, of government injecting so much capital into the system. And if that starts going away, do you have opinions or thoughts on all these startups that are making our lives better? Like I think of DevX startups where I don't know how financially sound they are yet, they've been living off the VC. Do you have thoughts on all that?

Shawn Wang: Not fully formed ones, but I can give you a quick hit.

Adam Elmore: Yeah. Yeah.

Shawn Wang: So how bad did it get? It got to the point, so the average price of sales ratio of a publicly traded company would be in the range of 10 to 50. That's a very wide range, meaning your market capitalization, the total value of a company is 50 times your sales. In private markets, the price of sales ratios of funding rounds, series A and B, and all that, got up to 1,000 times.

Adam Elmore: Oh my God.

Shawn Wang: We had 1,500 at one of the startups that I was at and I heard of one startup that was 2,500.

Adam Elmore: Wow.

Shawn Wang: So that was the peak in November of last year. Those days are gone, people are now asking for 100X, which is very like 10X fall, like very, very big. That's why almost nobody's raising money. So that VC market is right up, I'll say it has different impact on different stages. And this is all to do with like, "Okay, would you invest in Stripe at 95 billion when Shopify used to be 100 billion and now it's worth 20 billion?" You probably want to buy the more quality asset that's already publicly listed than the very stable asset that is at a high valuation.

Shawn Wang: So this is the deal making has just gone off. Like I think at the seed stage, people are completely unaffected. I think people are cognizant of the fact that economic cycles repeat or like, this is not going to... This is a recession. We are probably already in a recession right now, we are in a tightening cycle right now, but this is probably not one of those that's just going to drag out super long. And startup take 10 years to build anyway, so why should your early stage investing be affected at all by what the current level of the S&P is? It shouldn't.

Adam Elmore: Yeah. No, it's true. I mean, so much of this conversation just echoes your bias towards long term versus short term, and I should have known that coming in. I'm asking all these questions that are very much like, there's a clear answer if you just think outside of the next year.

Shawn Wang: Oh, I love training people to do that.

Adam Elmore: Yeah. No, it's really nice.

Shawn Wang: Take a long-term perspective in the history and then project it out to the future as well, and try to make decisions on that, so.

Adam Elmore: Yeah, it's sort of refreshing, especially in this sort of anxiety-ridden digital space. I feel like when you zoom out things feel a lot less pressing or anxiety-laden, I guess. I don't know. Yeah, I appreciate that.

Shawn Wang: It's weird because I think that's true, but at the same time, you're only here on this earth for so long. When you zoom out, that actually reduces the available number of decisions that you can possibly make, which means that each decision goes from being a two-way door into a one-way door because you want to make more substantial decisions. Therefore, for example, when I changed jobs, it took me like two months of agonizing to finally land on something, because I could have done any number of things and I think you have to really examine your beliefs as to what the long-term trends are going to be and trade that off versus being happy in the short run.

Adam Elmore: Yeah. I'm going to be trying to do that. I think I'm in the middle of the agonizing stage right now, trying to figure out what's next, but I'm going to try and think a little more long term.

Shawn Wang: The thing I'll point you to, you're talking about courses and stuff like that in leverage, I'll say definitely check out Eric Jorgenson, who is the book writer for Naval Ravikant. He wrote the Almanac of Naval Ravikant, and he's trying to build up a thesis or a body of knowledge around what leverage is and what leverage means. And then the other thing I'll point you to is Nathan Barry, who's the founder of ConvertKit who talked about the letters of wealth creation and how some things are more high leverage than others, so.

Adam Elmore: Thank you so much for that. Again, this podcast may just be for me, but that's okay because I got a lot out of it. Thank you so much for taking the time, Shawn.

Shawn Wang: [inaudible 00:50:58].

Adam Elmore: I didn't know how much I'd get in on my... I think we covered half the things I thought about talking to you about. You're just a wealth of knowledge, you're sort of a wise sage in this community and it's been so great to pick your brain. Thanks for coming on.

Shawn Wang: I think we're the same age.

Adam Elmore: Oh, yeah. Well yeah, you've been using your time better, I guess. You've been doing more high-leverage things or something.

Shawn Wang: Yeah. Thanks for having me around, but we can talk anytime. I really enjoyed this conversation.

Adam Elmore: That sounds good. Thanks, Shawn.
[Weekend Drop] AWS, Cloudflare, and Techbro Therapy on AWS.fm
Broadcast by