Transcript

Paul Ford The closer you are to a Keurig, the worse the design gets, right? [Paul laughs] When you see those like three glass trays or coffee pods, you better cross your fingers. [music fades in, plays alone for 14 seconds, ramps down]

PF Alright, so, Rich.

Rich Ziade Yes, Paul, how are you?

PF I”m doing okay now but that’s really not why people are here. They’re not here for us to do well, they’re here to talk about…tech.

RZ Oh, no, no. They want us to do well. I think there’s good feelings out there, good vibes towards what we do and who we are. So, let’s not…

PF No, you know what, you’re right about that. It’s nice. We like hearing from people who like the podcast, it’s very positive. [right] So look, during the day, I’m the CEO of Postlight and that means that I help Postlight grow and I help people help Postlight grow right. That’s my job. I think… You’ll tell me if not.

RZ For those that are new to the podcast, you’re the CEO of Postlight. Postlight is a digital strategy design and engineering firm based in New York. But right now we’re everywhere. And yes, I mean, part of your job is to represent Postlight, be the voice of Postlight, to help steer growth for Postlight. Yes. So I just wanted to give that backstory.

PF Thank you, no, it’s important that people know why they’re consuming this delightful marketing product, right. And I think more and more as time has gone on, I’m becoming more and more comfortable with my role and and sort of really deciding that like, okay, here we are, I’m here every day, we’re gonna make this work. But every now and then I still need to build something, like I still need to code a little something. So I feel connected to the world that we’re in. 

RZ You need to do some gardening.

01:45 

PF For real! And these are little projects. We did one for which, we launched under Postlight Labs, which was very fun called Account, which is like just a little interactive calculator. [mhm] Probably we’ll get back to that one. And one of the ways I work stuff out is by coding So anyway, the other day I love archives and there are wonderful, historical archives out there. archive.org has gathered tremendous numbers of books together, the Smithsonian has 3 million digital objects online, so on and so forth. And part of the fun of dealing with digital archives–and this is a very qualified definition of fun–is that it’s messy. The metadata is messy. Everybody will consolidate everything. And you still can never figure out like, Where is the image pointer? How do I get the picture or whatever? [mhm] So long story short, what I did was I built a tool–it’s a web page, you go to the web page, I can’t even give you the URL because it’s one of these cloud service URLs, which it’s just like, it’s literally ephemeralist/ixz4p7mlaq/ue/…

RZ Not useful. Not useful.

PF No, not good, right? And when you go, it loads 10 high resolution images from a variety of public commons. I’m on it right now. And I’ve got literally what looks like a rock from the clearly something very ancient that used to be alive from the Natural History of Paleontology Collection, the cover of Popular Mechanics magazine from 1928, a picture of a donkey from the 1700s, things like that. So mix of like just stuff, just old vases and scientific papers and old books. That’s fun. It’s a nice collage. Anyway, this is all me clearing my throat cough cough to talk about the key part of the experience, which is I built this for entirely for my own use on my own computer. And it’s literally this was a process of hours and then the computer ran overnight.

RZ Let me restate what you’re doing here. There’s this wealth of content, text images, scanned documents that live in archives around the web, stored in these big huge stock files of media. Correct?

PF 100,000 art objects from the Metropolitan Museum of Art.

RZ For example, that’s like a that’s a sliver of it, right? So you went in…

PF A million old books.

RZ And you, you sucked in some of that content and then you presented randomly at some URL, that sounds like that’s what you did.

04:03 

PF That’s right. And a lot of it’s already on GitHub, right? But just I wanted the randomness. Kinda to, it’s like social media, like, I just want to look at something but I want it to maybe not make me feel really awful all the time. [Rich laughs]

RZ Unlike social media!

PF Unlike social media. Just I feel that that’s a real product problem with social media. 

RZ Yeah, it’s interesting. Okay. 

PF So I tweeted out some pictures from this and people were like, you should share it. I’m like, alright, I should and I was like, alright, let’s take two hours. And let’s get this thing up on a cloud hosting problem. 

RZ Okay, so this is an important distinction. You had built this locally.

PF Ran on my computer here and actually in my home network.

RZ Like under your desk. Yeah, got it.

PF I kinda did it just so when I’m going to bed I would have something to look at that would be distracting. 

RZ And it’s not about pain and suffering and insults and humans.

PF Or why one person is bad or you know why the Lincoln Project is destroying society, just whatever it is that people are…

RZ Yeah, yeah. Just get away from the grind. 

PF And what’s better than old art and ridiculous ephemera? I like a lot of historical nonsense that you don’t to pay too much attention. 

RZ This is good. So this is this is you essentially saying I don’t want to stare at all that negativity, let me just like there’s a wealth that’s on the web. Let me go pull it up. So you did this. You pulled it off in your own house, cool little project, wrote some scripts, but then you said, okay, I’m gonna make this available to the world. [yep!] And then you decided we live in an interesting time in the history of software because there are these incredibly powerful cloud platforms.

PF Sooo many choices. So many. One of the easiest, we have a client Glitch, Glitch.com, Glitch is great. Glitch is the easiest way to build a complicated website that I’ve ever seen, But the paid product tops out very reasonably at like four or 500 megabytes of storage. And I was like, but I needed more than that. So not only is Glitch a client, but it really is easy. It’s man, it’s so damn simple to get if you know a little bit about the web, Glitch makes it really easy to kind of like, take the next two steps. But so that one was off the table for data storage reasons and like, you know, and so I was like I you know what I’m gonna learn one of the big ones I know AWS a little, I’m not emotionally ready for Microsoft Azure. 

06:09 

RZ [Rich laughs] If you can’t pronounce it, you’re not allowed to use it. 

PF I’m just like, I don’t even know what’s in there, man. [fine] Every time I see that one, I’m like, woooo, so I’m like all right, look, Google Cloud Platform.

RZ Another one of the players, one of the big players.

PF So look, they’re all very philosophical, all these hosting platforms, right? There’s Vercel, which I think used to be called Now. And a lot of them like we use a lot at work, which are about the fact that when you push to the main branch of your GitHub repository, and you update your code that triggers a build, and that also will happen when you’re in a branch. So you can always kind of be previewing things in production. [mhm] So those are like the simple ones. And then it was Google Cloud Platform, which we’ve talked about it before. But I mean, if you go to Google Cloud, [Paul laughs] it’s I’m going to right now you hear me type, bah bah bah bah. You scroll down the left of Google Cloud Platform, of all the services, and I mean, it’s like you’re driving cross country. [Rich chuckles]

RZ Okay, there’s just stuff, whizzing by. 

PF It’s also, can I just, I just need to say one thing that’s kind of important, which is that I have two accounts in Google, I have my personal account, and I have my Postlight account. And no matter what I do at Google Cloud, I have to run inside of like a Firefox container in order to log into the right account, like, you know, just, [I know] letting people log into one thing or the other is the great unsolved problem of our era. Like you think climate change is bad. And it is, it’s really bad. [mhm] But Single Sign On might be worse.

RZ So Paul, I mean, there was a day and I’m not I’m old enough to remember when you actually had to lease out chunks of hardware that were located in a very particular geographic facility. Rackspace was one of the big players to host your software that ran on the web. That those days are gone and AWS reset everything effectively and said, just lease out cycles, CPU cycles and storage and etc, etc.

08:12 

PF What I love, you look at this, you look at this left nav, and it’s a history lesson every three to five years everybody reinvents everything. And they still end up with like, it’s kind of a Linux box that you need to log into. So I mean, I’m looking, they’ve got App Engine, Compute Engine, Kubernetes engine, Cloud Functions, cloud run, and VMware engine. And all of those things at some level are just like here will make Linux possible to deploy in the cloud.

RZ Yes, yes. I mean, that’s right. I mean, they they did solve the scaling puzzle. So it’s completely invisible to the buyer at this point. Like you can scale and to the nth, and you don’t really have to do a whole lot. There was a day when you needed experts to scale.

PF I know when you put your mouse over any one of these items on the left nav like another 20 options come up. It’s probably about, okay, so this is, this is the problem they’re trying to solve, which is, it’s not that hard to administer one computer, it’s really hard to administer hundreds, it’s hard to deal with scale, and so on. So what we’re gonna do is we’re going to create an abstraction over the idea of a computer in the cloud. And on the internet, we’re gonna make it easy to juggle those. The problem is there’s no simple point of reference, that everybody can rally around, like there was with the graphic user interface in the 70s and 80s. Right, like the Mac starts to get going. Xerox is doing all this work at Xerox PARC, you know, then you get the Amiga, and then you get Windows and then you get like, like, it just became really clear that like, this is how we’re going to interact with this thing. And let’s all get on board, right? And so then you end up with a sort of whole generation of product thinking around how to make windowing interfaces more powerful, and how to empower people and then it kind of lands on the phone and so on. It’s still dos in Google Cloud Platform or AWS. It is like a set of shelves scripts, but they are trying their damnedest, but it’s still a set of menus that you hit return and it gives you a text list of 20 options no matter what you do.

10:09 

RZ I think what you’re touching on here is this world is the exclusive domain of engineers. [yes!] You know, the laundry list of links, and AWS is just as guilty. Amazon Web Services is just as guilty, you’ll go on there and you just get hit.

PF That’s the thing. AWS is much worse. They don’t even bother to name the things. It’ll just be like, this is called squid factor, and you’re like what’s it do. And they’re like, I don’t know.

RZ It’s actually really, really interesting to see. Because what it tells you is these are products and wildly successful products. AWS is a monster, like it is a huge, it’s a meaningful part of Amazon’s business and Amazon is a huge business. But what I think what happened here is design the idea of thinking or worrying about design is not even in the conversation like we’re talking about it. Because we’re kind of outside, we have like one foot in design, one foot in engineering, and then their third and fourth foot in other places. So we get to make comments about it. But I guess the assumption is that engineers don’t care. They want the power, right? I mean, this is not is this an upgrade? From terminal? Is this an upgrade?

PF It’s like anything. It solvescertain problems. I think it’s more complicated than what you just said, because I’m looking at Google Cloud Platform right now. And I’m looking at AWS, there’s so much design work that’s gone in, but the product platforms keep building, they keep growing, and new services keep showing up literally faster than people can contextualize or understand them. Like it’s just what happens if somebody is like, Oh, you know, be cool and…and then they launch it. And then it’s really complicated, like everything else. And they just sort of put it on the left and cross their fingers. I mean, I’ll stay with it. If you look at Google Cloud Platform, if you look at the left bar, you’ve got this section, which is like Marketplace, Billings, okay, so that’s my actual interface to it and tutorials. Then I’ve got Compute where I have like six products that are basically kind of do the same thing like let me run software, now I have got Storage, which is confusing because I guess that means I’m not supposed to store things inside of the computers I buy but the storage just like databases and SQL and big table and Spanner and I have no idea and and memory store, then I’ve got Networking, okay, so that’s those are things I want to I want to connect computers to each other and be in VPN, and all sorts of stuff. Then I have Ops, operations, monitoring, logging. And then I have another section called Tools, which lets me like, put my containers places and do all kinds of stuff. Then I have another section called Big Data, which now all the sudden I’m in a world where it says like, data catalog, healthcare, Life Sciences, AI, and there’s like 10 of those, including one for like, talent solution for help wanted ads, Other Google Solutions, which includes game servers and maps and then partner solutions, which is third parties. I said, I think this is the future of computing, right? Like, I think we’re all going to be in this thing if you’re doing tech in any meaningful way, you’re going to be inside one of these environments, and the browser tends to be your big interface. Help me make sense of this, dude. Where do we start?

13:14 

RZ Oh, I mean, I don’t know if this is that different than complaining about the UX of AutoCAD.

PF Oh, you think?

RZ Well, this is for the experts. I mean, you’re a tinker. And you went in, I’m a tinker, I’d go in, you know, I’d go into 3d rendering tools that are like huge industrial strength tools that are all about knowing the keyboard mappings. And I think there’s a couple of things at play here. One is, is that there is this expert cohort that actually likes the really small font, just absolute chaos and mayhem of it, because it reinforces their expertise. And it’s saying this is not for mass consumption. This is for the experts, right? I think there’s, not to say that that you know, there, you know, engineers are a collection of insecure people. I guess what I’m saying is, we’re fine. We don’t want ease of use, we want power. And what AutoCAD is about is power. And you’re not going to make me feel more powerful by making this really easy to use. I just want all the power. Look, what inevitably ultimately happens, almost always, is someone comes along and decides to break barriers and carves little pieces out and makes them available to the masses. Right? And that happens inevitably, with almost everything right here. You still have a world of complexity that is made available to people who are used to typing in into scripts right and or managing consoles. You ever seen you think do you think point and click solves this? Have you ever seen like the systems administration for like Windows NT2030. Have you ever seen it? You ever seen like the…

15:01 

PF Oh yeah. I have. 

RZ Like the rights delegation stuff. It’s one of the most complex things I’ve ever seen.

PF What I love in Windows is when the tabbed window interface, those tabs just start piling down. And there’s like 10 layers of tabs. 

RZ Yeah. Yeah. Yeah. And and so, and again, that’s for like an IT administrator that they’re supposed to understand how all that works and how rights delegation works and all that. I don’t think anybody’s gonna apologize for this is what I guess what I’m trying to say.

PF No, they shouldn’t. Right. So I mean, what is this really, right, this is Google, the best thing that Google does, like the all time best thing it does or did before spackled ads all over everything is Search, like you put words in a box and you get everything you’re looking for in return. And the actual power necessary to pull that off is stunning. You need datacenters powered by hydroelectric plants that usually ran aluminum smelters in order to do that well. And this is all the stuff that they use to get that work done, right? What is in here? And what’s left for everybody else to figure out is the kind of like, what’s your search box? How do you make something that simple or powerful like Google Maps or like Search using all this infrastructure? And it’s a really interesting proposition, right? Like, hey, we’re gonna give you literally big table, which is something we use or Spanner, which are things that we built in house, just like Amazon is. Amazon gives you their internal tools that you know, that’s where a lot of the stuff I prototyped and then we’re gonna we’re gonna give it to you in the world. Here’s what I think, when I look at this, I just see, yeah, oh, my God, it’s probably hundreds of billions of dollars, all told. And then there’s like the trillion dollar opportunity. And it’s more than dollars where all of this power could just get simplified in some way that I could open up my web browser and have a little computer that lets me do interesting things online and share my little computer with other people. Like, that’s all in there. But it’s not taking shape because I think there’s just so damn much money to be made right now. And everybody’s like, all the clouds services competing by just looking at this. I’m like, I could have 30 little computers doing my bidding right now. And they can be running all night. And I would just need to figure out what the hell I want to do with them. 

17:09 

RZ Yeah, here’s my argument against that is you are a market of $20. The tinkerer market who wants ease of use, and the market of like this, I am compete, Google and Amazon will go toe to toe to host Disney+ right?

PF Oh my god, right. Yeah that’s true.

RZ Like they go to war over it. Right. And it comes down to the decision making process against a grotesque set of requirements that get competed against and…

PF Why not have a team of 20 or 40 and you know, HBO Max or wherever, who are going to know their cloud platforms really, really well.

RZ You know, I mean, that those are the customers they want, they don’t want, you and pictures of like bones from the 1200s or the 4th century, are not interesting to Google Cloud, with respect, Paul Ford. 

18:01 

PF No, utterly fair, right, except, and this is I’ll advocate for myself a little bit. It’s going to go this way, there are going to be weird consumer layers on top of this world of giant cloud products in which people are like setting up and throwing away little computers because the interface is easy and it’s convenient for them. Minecraft servers are a great example. You’re gonna set up a Minecraft server because you want to play Minecraft. So whatever makes that easy and easier is better because it otherwise it starts to take away from your time playing Minecraft.

RZ When you lower barriers, you open up markets, right? Like I mean, that’s a real that’s a real thing, right? I mean, I you know, it’s interesting that you see the trajectory of Salesforce and Salesforce started simple. And then it became messier and messier because Salesforce said, you know what, I’m going to sell to the people that want all the power so I’m going to give them everything and evolved from a GUI interface to a software platform of the most, of the worst order. I mean…

PF It’s much better now. But that early Salesforce interface was, wooo.

RZ It’s just grappling with all the asks. Right? It was, I mean, it was a war crime. The whole thing was just this, this just collection of buttons.

PF I gotta say, as a CEO, as like a tall chubby guy, and what you need to do in order to succeed in this role. I look at Benioff more and more and go like yep, yep, yep, that’s right. Okay. Okay. [Rich laughs] But he’s just, yeah, he’s very positive. He loves Salesforce. He believes in his, he’s got his politics out right in front and it never, it just never stops. And you’re like, alright, I guess that’s how you do it.

RZ It kind of speaks to where you are and what you’re seeing peering in, and you’re seeing opportunity, right? Like this can be so much better. And I think their agenda is just driven by, just it’s just war to win the large swaths of land rather than the little guy who wants to play Minecraft. 

19:55 

PF I gotta say ’cause I will hear from people like how come things aren’t this way, right. Like We know what a really great computing experience would look like. And here’s Google Cloud, and you could do so many things. And it could be so wonderful. And it’s because they have to kind of find the growth that’s in front of them before they can invent the future. And they kind of forget about the future when you’re just trying to keep the server’s up. 

RZ Enterprise software is a shitty wasteland, top to bottom and what you’re talking about with the cloud platforms, it is effectively the fertile soil that sits underneath that shitty wasteland. 

PF And and they’re better than most enterprise like Google and Amazon and I’m sure Azure too are much better.

RZ They are! Once the consumers in the mix, then Google will mobilize the materials, what is it called? Material? Is that their design?

PF Oh yea, Material? Their UX system? Yeah, Material Design. 

RZ Oh, there’s 300 people that are rounding corners at Google right now for the consumer, for the consumer who has dried spit on the corner of their mouth, because they’re scrolling through their Android phone, there are designers that are making sure that you do not look up and make eye contact with your child. That’s their job. Their job is to minimize human interaction, right. But when you’re talking about the enterprise, right, once you cross that threshold into the workplace, where there are Sun Chips and bananas for the common good. For the commoners, all bets are off, you are going to use that tool and you’re going to eat that shit. That’s just how it works. 

PF Look, I mean, the story I’m, so let’s be clear. First of all, I tweeted about this, I had a good time. It took me two hours to go from learning enough Docker to get a container up and get it into Google Cloud and run it and all the regular stuff happened. I ran out of memory and just like but you know what? I have to admit, as messy as it is, and as confusing it is. I’m going to be hosting something for about $3 a year.

22:00 

RZ [Rich laughs] Right. And we will share the link to this thing once…

PF Yeah, sure. Why not? 

RZ Why not.

PF It’s not even a product. It’s a couple hours of goofing off. 

RZ It’s an experiment. Yeah, yeah. Cool.

PF Yeah. And so like, so there’s stuff to learn. I still, I cannot look at these Cloud platforms, Richard and look away, because I just keep seeing the future of the operating system and how it’s all going to go and I can’t quite get my hand around it. But damn it, I know it’s in there.

RZ Look, there is value in it. We’re mocking it that nobody cares. Nobody wants it. We just did a beautiful redesign and rebuild of MailChimp developer experience. And it was great working with them because they got it. They knew where they wanted to be. And we worked closely with them to get there. And it makes all the difference in the world. Giving developers a good home to educate themselves and feel empowered around what you’re doing is a pretty awesome thing. Right? I think where it gets a little murky here is that this is the world and we need an echo effect when I say these words of Dev Ops, where it’s not just engineers…

PF Dev ops, ops, ops, ops, ops!

RZ Right? And and and you’ve got administration, you’ve got all kinds of other people in the mix that have different interests and different drivers of what they think is good.

PF This is what’s real about tech man. There’s the 20%, about design, making the product work well, you know, doing the research, shipping that initial MVP, and then there’s the 80% ongoing of keeping the server’s up. Yeah, upgrading Red Hat Enterprise Linux, like all of that.

RZ And that, you know, this is a world that’s still young. That’s the only thing worth mentioning here. We’re at like Windows 2.0 here. [music ramps up]

PF And what’s happened is the 80% world and the 20% world are like they’re eventually going to meet up.

RZ So we are Postlight, Paul, I don’t think we even need to pitch, I just snuck in that we just redid the developer experience for MailChimp. We partnered with them on that. That’s a wonderful example. It’s a wonderful example of the kind of work we’ve done. But there’s all kinds of other nice examples if you visit Postlight.com. Check out our case studies and then there’s a wealth of articles by lots of good writers, including Paul Ford, Gina Trapani, and many of…

24:09 

PF Also Rich Ziade!

RZ Rich Ziade, he’s pretty good writer for US English as a second language. He’s not a bad writer.

PF Yeah, you came over when you were five. Come on now,

RZ No I know, but I’d like to show like I you know, I kind of like pulled myself up by the bootstraps and etc.

PF We should actually pause for a second and just note like we have we’ve talked about we have a team in Lebanon, everybody is safe and healthy. A lot of people have reached out. We’re taking some steps to support the team and the country. And really, we’re thinking a lot about Beirut and what we can do.

RZ Yes, absolutely. Absolutely. 

PF Thank you people who reached out.

RZ So reach out, talk to us, hello@postlight.com. We’d love to chat. 

PF Alright, let’s go.

RZ Let’s go! Bye! [music ramps up, plays alone for 3 seconds, ends].