Paul: Hi, I’m Paul Ford.

Rich: And I’m Rich Ziade.

Paul: And this is the Aboard podcast about how AI is changing the world of software. And guess what, Rich?

Rich: What?

Paul: Usually we talk sort of big picture. AI—whew! What’s happening?

Rich: [mocking] Ethics.

Paul: [laughing] Fair. Okay, great.

Rich: [mocking] Government.

Paul: That’s literally what people think business people do all day. [laughter] [mocking] Ethics! We are going to talk about AI in the world of software in a very specific context over two episodes.

Rich: Okay.

Paul: It’s gonna be interesting, don’t worry.

Rich: Okay.

Paul: And it’s about our software, what we’re shipping, how we made decisions, and how we’re gonna accelerate the future of business software development, like, fivefold.

Rich: It’s about frickin’ time.

Paul: It really is. It feels good to launch. Let’s—

Rich: Yes.

Paul: Let’s go.

[intro music]

Paul: Okay, my friend, we did it. First off, high five.

Rich: High five. [high-fiving noise]

Paul: Now you can watch on YouTube. We have some fun illustrations. But we’re going to be describing software on this podcast.

Rich: That’s right. You might be listening in your ears.

Paul: So I’m going to try to make software come alive in your ears.

Rich: Yes.

Paul: So just bear with us. We’re going to do a great job.

Rich: Let’s do it.

Paul: Okay. So we have—okay, when I say we launched Aboard, we’ve launched a product on and off over the last several years—

Rich: Half a dozen times, yeah.

Paul: This I would say is truly the final form. In some ways it’s kind of a return to form for us.

Rich: I don’t think there’s a U-turn, because we’re going to start building software for customers, like, tomorrow.

Paul: I mean, we are now actually.

Rich: Yeah.

Paul: So we—

Rich: So we don’t have… [laughing] This is it.

Paul: So Aboard in some ways is the truly accelerated agency, and at the same time it’s also a product, and I think that represents a—it’s not just us. I think this is the way software is going. So we’re going to talk about what, what we do and what you can do today. And people can actually kind of follow along and build the demo if they want to, because we make that available.

But I think we should also talk about why we landed on the decisions we’re making, because the big thing about these decisions is that there’s a stopping point for where AI can get you. And I think people need to start owning—we believe, and what we’re doing, we’re going to own that stopping point, and then help you get your thing done, whether AI can do it or not.

Rich: So let me say that back to you. AI can do a lot of things.

Paul: Mmm hmm.

Rich: It makes pictures, it can edit music, it’ll make videos, it’ll answer questions, which is how most people use it. But there’s a huge audience out there, there’s a huge population that is using AI to build software. They’re trying to use AI to build software.

Paul: Correct.

Rich: And what you’re saying is it only gets you so far.

Paul: It only gets—it doesn’t finish. We’ve said it on the podcast before, it’s really great at the first mile because it seems human.

Rich: Yeah.

Paul: It is not so good at last mile because it’s not. You need people. So what are we talking about? First of all, let’s just narrow the domain a little bit. We’re talking about business applications, things where data goes in, data goes out, multiple people log in—

Rich: Business logic kicks in.

Paul: I’m managing PTO, I’m keeping track of inventory, I’m doing classic—

Rich: Right, we’re not talking about video games.

Paul: No.

Rich: Right.

Paul: That’s where we’re focused. And actually the easiest way to get people into this and help them understand, thank God, is instead of us just talking about it, we can show now instead of tell. And if you go to aboard.com, there’s a big text box right at the top, you can put in a prompt and it’ll build software. I thought, I’ll do that here. I’ll read a prompt and then we can just sort of talk about what it builds.

Rich: Sounds great.

Paul: And then we’ll talk about what those pieces are.

Rich: Before you start, we encourage you—watch this on YouTube because it’s visual, there’s illustrations, but you can listen to it as well.

Paul: Yeah!

Rich: We’re going to be very descriptive.

Paul: Great. So—

Rich: Let’s do it.

Paul: I’m going to give you a prompt.

Rich: Okay.

Paul: Ready?

Rich: So there’s a box. What do you want to build today? Go.

Paul: I have a large pumpkin patch, apple orchard, and hayride business with $5.4 million ARR.

Rich: Whoa!

Paul: What does that stand for?

Rich: Annual recurring revenue.

Paul: It took me years to remember that.

Rich: Yeah.

Paul: I need a PPP—picking, patch, and planting manager.

Rich: Okay…

Paul: To be fair, I made that up, but I put it in anyway.

Rich: Fine.

Paul: And a system for managing hayride scheduling and apple picking group scheduling, especially for large groups. We’re also doing more weddings. I’d like to be able to manage employees, especially given seasonal variations, and to keep an active pipeline of new seasonal help.

Rich: Okay, so you just read the prompt verbatim. You’re going to put that prompt into a box.

Paul: Mmm hmm.

Rich: On aboard.com.

Paul: Mmm hmm.

Rich: Right? And we’ll see what happens. First off, the way you read it and it started off with, “I have a large pumpkin patch,” it sounded like you were reading out of a children’s book, but that’s okay.

Paul: Fair enough. I don’t think we actually have to go—let’s not go through the whole process. But if you go put this in the website, it asks you a few follow-up questions, you answer.

Rich: Fine.

Paul: Takes about four minutes, all-in. It asks you about migrations, about all kinds of stuff. Let’s skip to the result and talk about what it made.

Rich: Okay.

Paul: So I put the prompt in, I answered three or four follow-up questions. Just gave it my email.

Rich: It wanted some clarifications.

Paul: It turned its wheels for a few minutes and it built me a thing called Harvest Manager.

Rich: Okay.

Paul: Okay? I’m going to try out Harvest Manager. Harvest Manager starts with a dashboard.

Rich: Uh huh.

Paul: At the top of the dashboard it says, “I’ve got events, applicants, staff.” So it gives me numbers. I have 69 staff. It’s pretty big pumpkin patch.

Rich: Right?

Paul: Okay, stop for a second and just tell me what I’m looking at.

Rich: You’re looking at a first cut of full-stack software.

Paul: Why does it have all this data? How does it know about 69 staff?

Rich: So once it built the software—and we should talk about that for a minute, I think it’ll be helpful—it pours in sample data so it can look somewhat recognizable to a user.

Paul: This is our experience as salespeople.

Rich: Yeah, yeah.

Paul: Unless you put real data inside of an application that feels real and looks real?

Rich: Yeah.

Paul: They don’t understand what you’re talking about. It’s just boxes until they see their world.

Rich: People throw documents and decks about what software is going to be. And then when they see real—when you see real software with, with pretend information, and by the way, it’s not, like, lorem ipsum. It’s not, like, filler text. It’s actually…looks relevant to whatever you tried to build.

Paul: So I’m going to move fast because we’re describing business software and people only have so much time in their life. But if there’s a left side, it’s a built me an app. This looks like a business app. It is a business app.

Rich: This is worth pausing and putting an asterisk on this point. This is not a prototyping tool. It’s actually full-stack software.

Paul: It didn’t draw me pictures, it built me software instead.

Rich: Like down to a Postgres database.

Paul: That’s right.

Rich: Like, it’s a full-blown first cut of software.

Paul: For the nerds, this is actually just a web app.

Rich: I’m going to go as-seen-on-TV for a split second. I don’t think anyone’s doing this.

Paul: Not like this. We’ve looked around a lot.

Rich: We’ve looked around a lot. Usually it rushes to code and it kind of does a bunch of stuff and then shows sample data and, like, React components. This is actually a full-stack instance of software.

Paul: You could in fact kind of hit eject, take that application, and just build old-school.

Rich: Code on top of it, yeah.

Paul: It is just a web app, except it’s a web app that is managed and controlled by a system that also has some AI tooling.

Rich: Right. So describe the UI.

Paul: Okay, so, on—you know, if you’ve seen Slack or really any business app, you’ve got your left nav and you got your right area which shows your data. And so what it did from that prompt with very little other context, it created an activities and events section, a crops and field management section, and a staff assignment section, and also has room for applicants and hiring. And each one of those is kind of what you’d expect. If I go into activities and events, it’s a Kanban board and it says that, it’s like a sales CRM or a scheduling tool.

Rich: Mmm hmm. Mmm hmm.

Paul: It has things set to, scheduled, confirmed, in progress, completed. I have a “Lincoln elementary third grade field trip for apple picking” here set for September 23rd, and it has information—and that’s a card. I can go in, I can edit that card.

Rich: Yeah.

Paul: I can delete the card, I can change the date. It has a phone number in case I need to call the person, there’s a rain date. So it has made things real.

Rich: Yeah.

Paul: And we don’t have to belabor this. If you’ve ever used business software and you go and see—

Rich: Looks very familiar.

Paul: And you can follow the link and look at the Harvest Manager. That’s public. We share it out in the world. If you build something like this with our tool, you can share it with people too to show them.

Rich: Yeah.

Paul: So I don’t want to sit here and distract software and people can go make it in five minutes.

Rich: Yeah.

Paul: But it built what is a, frankly, a credible first-pass business app.

Rich: Yeah.

Paul: When I was building software with you at our agency, we spent months getting to this point every time.

Rich: Yes. Yeah. And it’s worth saying out loud when you put in a paragraph or two, even if you’re very verbose, it still has to make a ton of assumptions about what you actually wanted.

Paul: Yes.

Rich: It’s not gonna be, like, “Oh, you got it all wrong.” Of course it got it all wrong. Because whatever’s in your head wasn’t expressed that explicitly to Aboard. x

Paul: But—it’s plenty of hand-waving, but I can take that and I can go back and I can edit that prompt, right?

Rich: It’s a hell of a starting point.

Paul: Yeah. I can iterate this way three or four times.

Rich: That’s right.

Paul: And now I have something I can go show my boss and I can say, “Hey, boss, do you think this is what we want?”

Rich: And it’s worth highlighting what you skipped, which is tons of conversations, note taking, requirements gathering, passing around documentation of what the thing should be.

Paul: Let’s be clear, too, that is very valuable, that time. We’re not actually trying to throw away the idea that people need to talk and think about the software they’re building. But so much of that stuff is pro forma and sort of territorial.

Rich: Exactly.

Paul: When you build software, organizations reject it out of hand, and then you have to spend all this time convincing them to just to let you put the box on the screen.

Rich: Yeah.

Paul: That’s gone.

Rich: Yeah.

Paul: Like, I can say that unequivocally, even though it’s our company, even if we went out of business tomorrow, someone else is going to do something—like, that era of they won’t let—

Rich: The ceremony and the steps….

Paul: [overlapping] They won’t let you build it until you’ve begged long enough?

Rich: Yeah, yeah, yeah.

Paul: That’s done forever and it’s never coming back.

Rich: Yeah, I think you’re right.

Paul: So God bless us all.

Rich: God bless.

Paul: So—

Rich: All right. What do we got here?

Paul: Well, we can build—let’s just talk about what actually happens, right?

Rich: Yeah.

Paul: So you talk to the AI.

Rich: Okay.

Paul: Okay? And then in the background, how is it doing the magic? What’s actually happening here? Because let’s be really clear, and I think that the emphasis for the rest of this podcast is like, people are learning about things like vibe coding.

Rich: Yeah.

Paul: And AI kind of will do all this stuff and agents and so on.

Rich: Yeah.

Paul: We have a very, very specific philosophy. And I’ll skip ahead, which is the AI doesn’t write the code.

Rich: Okay.

Paul: Okay? It helps organize things, and then the software is produced in a different way. And we’ll talk about that in a second. But what happens after you put the prompt in, okay? A lot of agents spin up. Tell me about these agents. What’s an agent?

Rich: So these agents are representations of specialized skills. A datum architect agent, a user experience agent, a roles and rights agent.

Paul: This is like when you use ChatGPT and you’re like, “Hey, you are an experienced database administrator. Can you please tell me how to define this database?”

Rich: That’s right.

Paul: But it’s ten of those, or five of those.

Rich: That’s right. And I think what’s different about our approach is it’s nonlinear, meaning a group of skilled agents show up, talk to each other, and talk to you about what you want, and then they fan out and do their specialized work.

Paul: Well, what you and I did, we’re 20-year agency veterans and we’re working with agency veterans on our engineering team..

Rich: Yeah.

Paul: We simulated the agency process.

Rich: We simulated—and why did that process come to be? It came to be because you don’t run to the construction guy and say, “I need a house.”

Paul: Right.

Rich: Like, there’s a lot of steps. There’s permit approvals, there’s architects, there’s picking materials. There’s so many steps before you can just start construction. And so we actually codified that process and simulated it through here. So that’s why it takes three, four minutes. It’s worth also, like, clicking in on that for a second.

Paul: Yeah.

Rich: Everybody’s used to things just showing up instantly with AI nowadays. Our process is pretty complicated and there’s a lot of work happening in the background. So it takes a few minutes. People will be like, “Well this is kind of slow,” but it used to take six months. [laughing]

Paul: I actually, once they see the output, they don’t say that anymore.

Rich: Yeah, that’s true.

Paul: I’ve noticed that.

Rich: That’s true.

Paul: Okay, let’s be clear. This is not a product for people who need—if you want a fully robust enterprise software app spun up from a one-sentence prompt in less than five minutes? You might be looking for something other than what we offer.

Rich: Yeah, yeah.

Paul: That’s okay.

Rich: Yeah.

Paul: So those agents talk—now they make a thing. I’ll describe it real quick and then you riff on it. So what the agents do is they say, like, there’s a database agent and it says, “Here’s the kind of data we’re going to need.” And then there’s a UX agent that says, “I think it should look like this.”

Rich: Mmm hmm.

Paul: And they put all of that into an old-school—it’s not quite a computer program. It’s a description, it’s a syntax.

Rich: Yeah.

Paul: And it says what—this syntax describes an application that looks a lot like that prompt.

Rich: Yeah.

Paul: Okay. And that’s, and that is—there’s a lot of detail in there.

Rich: A lot of detail.

Paul: I mean it is long. There’s a specification we’re building.

Rich: Yeah. It’s not code.

Paul: It’s not—it’s a description, it’s a description, but it’s a description that—and the interesting thing about this description is, like, AI can generate the description, but a human can edit and understand the description as well.

Rich: Yes. And why did we do that? The best analogy I can think of for Blueprint is if you view the source of a webpage—I’m talking about a plain webpage that just shows pictures and words and whatnot. It’s a description of what the page should be.

Paul: Mmm hmm.

Rich: And then the browser takes that description and turns it into something pretty that you can look at and read and whatnot.

Paul: That’s right. HTML tags become…

Rich: Exactly. It’s not that dissimilar, what we’re doing. Why did we do this? I think it’s worth talking for a minute about why we did it. Because we’re forcing AI to fill in a form, effectively—

Paul: Mmm hmm.

Rich: —and it has to stay inside the empty boxes of that form, it minimizes the sort of randomness and it reduces the likelihood of AI straying off. And look, AI still might fill the form out not great. But now we have given it strict form, other definition of form, to stay within. And when you do that, you end up with a much more predictable platform.

Paul: An enormous amount of our effort over the last year has been about getting AI to do less.

Rich: Yes.

Paul: In more structured ways, right?

Rich: Yes.

Paul: And so all those little bots, they talk together, they come up with a good plan, they turn into this syntax.

Rich: Yeah.

Paul: And now it’s called the Blueprint. Okay, now the Blueprint gets transformed.

Rich: Yes.

Paul: Okay. And we call this generation. It’s a little bit like compiling code. You go from this Blueprint thing, which don’t worry about what’s inside of it. It’s data objects inside of it, if you’re a nerd.

Rich: We’ve been struggling with this verb. I, I—

Paul: Let’s go to compile for this, for right now—

Rich: Go to compile. You could do render, like a browser renders Markup. It is a lot of words.

Paul: So what it does—now, what’s inside of the Blueprint is a description of the data model.

Rich: Yeah.

Paul: A set of instructions for the user experience, a set of user roles like administrator, user, subscriber, whatever.

Rich: The common building blocks of software.

Paul: And then what’s evolving now, also a set of business rules.

Rich: Yeah.

Paul: What to do—somebody updates the inventory. Go ahead and check the other external API, because maybe—

Rich: Fire off some event, yeah.

Paul: We might be out of shoelaces.

Rich: Correct.

Paul: That kind of thing. So it’s software. It describes the software and it compiles it into this application, much like Harvest Manager, which I just showed you.

Rich: Yeah.

Paul: Harvest Manager is a demo on the web. The sort of more formal version of Aboard is that plus actual login through your Google account and all the real sort of businessy stuff.

Rich: Yeah.

Paul: So let’s talk about the pieces. Right? So first we have—and this is why the five-minute part is really wild, and why it’s a little more than a prototype. First, we have a database.

Rich: Yes.

Paul: Okay? The database has, like, all the things a database does. Like, you can’t put, you know, you might not be able to put two things with the same ID in.

Rich: Yeah.

Paul: It might—

Rich: It holds information. It’s a bucket.

Paul: It’s a bucket. A bucket with rules.

Rich: With rules, and also information is related to other information. So there’s nuance to it, but it really is a place to hold information.

Paul: And we stand that up. And it is, like, it is a cloud-based Postgres database.

Rich: Classic stuff. Modern….

Paul: It runs inside AWS.

Rich: Yeah.

Paul: One of the interesting conversations is people are like, “Ah, it’s really hard to get stuff like this into my organization.” We’re like, “If you can deploy AWS apps…” And they go, “Oh, well, I can do that.”

Rich: Which, by the way, we’re not married to.

Paul: No.

Rich: If you told us, “Look, this needs to be on Azure,” fine.

Paul: Yeah, that’s right.

Rich: It’s pretty agnostic.

Paul: Because it’s just a web app. Right?

Rich: It’s a web app.

Paul: So we’re trying—and let me just get meta for a sec. We’re trying to take this new, exciting world that everybody says is completely different than everything that ever came before?

Rich: Yeah.

Paul: And bring it back to the good old reliable world of solid software that people use all day.

Rich: Yeah.

Paul: As opposed to a weird chatty robot that tells you things.

Rich: Great.

Paul: So now the other thing, because of what we did, you get something that works pretty good in mobile.

Rich: Mmm!

Paul: You get all those niceties of the web app.

Rich: Yeah.

Paul: Regardless, it could have RAG and AI and all these wonderful features of chat, but we’re not shipping chat. This is just an app.

Rich: And let’s take a minute and acknowledge, you know, software moves fast, but two billion people use Excel, and many of them are not doing math in Excel, they’re just putting information in it.

Paul: The world thinks in tables and then—but that next step where you’re organizing the tables? Really hard.

Rich: Yeah.

Paul: And so then you saw with Harvest Manager, we’re going to give you, when you do this on the website, get your own little place. It’s aboard.com, there’s a secret URL.

Rich: Uh huh.

Paul: You can share it out and you can talk about it—

Rich: With your colleagues or friends or whatever.

Paul: And look, this is a demo product. We live in the future. It’s very exciting. Many wonderful things are happening. But I wanted to just pause for a second and talk about what actually happened, which is that you typed—you’re a smart software person, even if you’re not a coder—you typed some words in a box, about a paragraph of words. They’re pretty well-informed words, you know about your business.

Rich: Yep.

Paul: And a set of processes spins out and the first draft of a robust, real software application exists in five minutes. And we get some credit for this because the Blueprint is novel, we’re doing some good stuff, and so on and so forth. But this is the real world now.

Rich: Explain.

Paul: I don’t think that people have metabolized quite what I just described.

Rich: It’s coming.

Paul: It’s not just coming, it’s here now.

Rich: Yeah.

Paul: You can now say to us, “Hey, that was pretty close,” or, “I need these two things,” or, “I need to migrate some legacy data.” And our lift, you know, the way Aboard works is you’d come to us, you’d say, “I just got to finish this.” But we’re thinking you’re going to finish it sometimes in weeks.

Rich: Yeah.

Paul: Like, what used to be—

Rich: That’s unheard of. It used to be nine months, right? Or six months, or whatever it is.

Paul: And it’s been very confusing, right? Because you get this promise, like, “Come to this website, vibe code a little bit, and you’ll have your app.”

Rich: Yeah.

Paul: And for really smart people who kind of are connected to computers and understand how a lot of stuff works, they can get really far. Like, product managers who have a good technical understanding?

Rich: Can play around, yeah.

Paul: But that’s still, like, less than 1% of computer people, like, you really…

Rich: Yeah.

Paul: Like, it doesn’t quite get you over the hump that much.

Rich: Yeah.

Paul: And this, I’m not promising that we do that in a minute or that you can do that by just sort of sitting there and typing in your underwear.

Rich: Yeah.

Paul: What I’m saying is, like, this is a trillion-dollar category of the world.

Rich: Yeah.

Paul: We’re trying to carve off our little piece and hustle and make—

Rich: Just a few hundred billion.

Paul: Just a few. But, you know, we have our office and we have a team and we’re going to—

Rich: Yeah.

Paul: We’re actually, we already have clients and we’re—

Rich: Yeah.

Paul: —going to keep working with people on this. We want to make lots of friends. We want to work with partners. We want them to use it. But the reality is five minutes.

Rich: Yeah.

Paul: Okay?

Rich: Yeah, you’re right.

Paul: And it’s, so we’ve been living inside of it for six months.

Rich: Yeah.

Paul: And it’s five minutes to something real that is now—but, both real and substantive and used to be. I’ll give you an example. When I used to go build stuff like this with our company and before, that first meeting was terrifying, because you’ve now put four months in. They’ve rescheduled three times because it’s not the most important thing to the CEO. [laughter] You’re going to come in and you’re going to show them this thing. The entire company that you’re working for has changed management four times.

Rich: [laughing] Yeah.

Paul: And you’re going to put this thing up on the screen and they’re going to say, “Well, we don’t really do that deal—we didn’t really do that deal with Meredith Magazines that we thought we were going to. So…I don’t know if we still—”

Rich: Stuff moves, right? The world keeps turning, and…

Paul: “I don’t know if we still need this. You guys are great.”

Rich: Yeah. Well, the cycle of building software is so long.

Paul: And it’s been four months. And two of those months were you building, and two of those months were corporate politics and slowdowns and so on. And so you tell me socially, culturally, that it’s five minutes? And I’m going to tell you in response, well, that for better or for worse, probably some ways will be worse, is a radically different software industry than the one we were in three years ago.

Rich: Absolutely.

Paul: Okay. So that—

Rich: Yeah.

Paul: And so when we talk about AI, and we’ve been kind of, we’ve had to be elliptical because you gotta get your product done so you can show it to people—

Rich: Yeah.

Paul: —before you can talk about your product.

Rich: Yeah.

Paul: But when we’ve been kind of riffing on this stuff, this is the reality that we’ve been—this is why we did the podcast. Because it’s genuinely—we’re not young dudes. We’ve been in this industry for a while. I made a career out of explaining—

Rich: Speak for yourself.

Paul: [laughing] You’re very young. I made a career out of explaining this industry before we did.

Rich: Yeah, yeah. Sure.

Paul: Before we started the agency. This is hard to explain.

Rich: It is hard to explain…

Paul: Which is why it’s so exciting to have something to show.

Rich: Yeah. And here’s the other thing I think that’s worth saying. I think for a lot of people, software in companies that don’t care about software, that need software, it sort of lives in the category of, like, the internet bill or the…

Paul: Rent.

Rich: The rent, right? It’s sort of this necessary thing. Otherwise it’s not, and so what’s happened is—

Paul: This is important, but let me pause you. Many people who are in the software industry, because it’s such a big industry and they have so many people who are also in this, and so many people are tech-adjacent, don’t have the exposure to realize that 99% of people don’t care at all.

Rich: They don’t care at all. But they also know—but they’ve been so beaten down, they also know that if software can be incredibly powerful for your business and incredibly additive and incredibly competitive. If you can get a good team in place, it can change the game. But it’s so rare, it’s so rare such that business stakeholders just, they’ve sort of resolved themselves to just make sure it keeps running. Just make sure it just runs.

Paul: One of the worst parts about running an agency is no one came to us because they were happy.

Rich: Yeah. That’s true.

Paul: They came to us because—

Rich: We were last resort a lot of the time.

Paul: It was expensive, things had failed.

Rich: Yeah.

Paul: They just couldn’t get their thing.

Rich: That’s right. That’s right.

Paul: And so they’re like, “Well, I guess I’ll do this.”

Rich: Yeah. That’s right.

Paul: And then we would deliver it.

Rich: Yeah.

Paul: But we were very expensive.

Rich: We were expensive, and you really needed the will to even deal with it. Which is the other point I want to make. If you can spin up credible-looking stuff that you can walk around in a few minutes, it doesn’t have to be perfect because we’re not, it was a few minutes, for Christ’s sake. It is not just a—

Paul: I love how defensive we all are about that.

Rich: Yeah. It’s not just a sort of starting point for a project. It’s also a social tool. It’s a political tool. There is no more powerful political tool. You can have the best, most lucid, persuasive bullets in a deck, but when you show someone a thing, it is very, very powerful.

Paul: You know what’s brutal here too is that, and I do believe this, and again, whether it’s us or somebody else, this is much, much cheaper.

Rich: I mean the idea of charging for peoples’ time for across many months times many people are sort of behind us.

Paul: This doesn’t mean—

Rich: It’s not there yet, by the way.

Paul: It also doesn’t mean that like no one has jobs in the future. I actually don’t buy that. I think there’s much more software.

Rich: Yes.

Paul: There’s, like, the need has been so great for so long, but there is a realignment because if somebody… If these slick huckster fellows with their AI toolkit can show up at your company and build you a prototype in five minutes and then credibly deliver a real thing in a couple of weeks?

Rich: Yeah.

Paul: And it costs a fraction of what the old thing used to cost?

Rich: It’s very disruptive.

Paul: Which, let’s be clear, that’s our proposition.

Rich: Yeah.

Paul: That is not great if someone else has been telling you that it’s 36 months.

Rich: Not just that. There’s people who—to build a new thing. Then there’s armies of people who just maintain the thing because they don’t want to change the thing because it’s been too expensive to change. So you just have people who just sort of, they wait for the thing to, like, hit those rare conditions and they just sort of monitor and they man the systems and they restart them once a day because there’s memory leaks. There are hundreds of thousands of people whose job it is to just make sure it just keeps running.

Paul: I believe that, too. I think we’re in this incredibly, it’s a really complicated moment in the world, but I think we’re in this incredibly wild moment for the software industry where something radical and transformative is happening—nowhere near as radical and transformative as AGI, but rather all of the systems that we’re dependent on that are slow and that we’ve been waiting to kind of improve, and “I really need that report” and all that stuff? I just think that’s over—and I don’t think it’s over in a day. I think it might be five years.

Rich: We’re in the beginning of that transition, I would say. But you could see it.

Paul: There is a point in the future—today I know a mid-level person at a charity and she really needs a report and it takes her a week to cut and paste it from Salesforce to get the data to put it in Excel to do the report so she can get her funding.

Rich: Yeah.

Paul: And she can’t get the Salesforce people to do it because it’s not on the roadmap and blah, blah, blah.

Rich: [laughing] Yeah.

Paul: Okay? I do not believe that a few years from now, that can continue to be the case and the Salesforce vendor can continue to thrive.

Rich: Yeah. Yeah.

Paul: I think that you have to be—a few years from now that she is going to turn and she’s going to say, “Hey, can I get that report?” And either her or, like, a nice nerdy person is going to type a report description into a box—

Rich: And they’re going to get it.

Paul: And it’s there forever.

Rich: I agree with that.

Paul: And that is really coming. That is around the corner. And that’s so much of software. People don’t talk about it because they talk about it, like, Adobe.

Rich: Yeah.

Paul: Or Apple.

Rich: No.

Paul: But that’s so much software.

Rich: 100%.

Paul: So that’s the way we’re going to ride. That is what Aboard is. We wanted to tell you about it. Go to the website. Check it out.

Rich: So wait, that’s it? I mean, I just type in a box and I get my software and then you pay a bill?

Paul: No, we’re going to talk about this one in the next podcast, but after that you can talk to us if you want to get this thing, like, deployed and say—we are also now building some of the characteristics of an agency. It’s very product-focused, but we are building and partnering with firms that can actually get these things across the line. Partly because that’s where our baseline is, but also that’s the limit of AI. AI cannot ship you an enterprise application out of the box no matter what it is.

Rich: It’s a relay race. AI is a really great first leg runner.

Paul: Mmm hmm.

Rich: But it’s got to get handed off.

Paul: Well, I’m going to give you a challenge. If you don’t believe us, go get the documentation that this thing produces for your app. Okay? Look at what we built and then go upload it into Claude and say, “Build me this app.” We give you all the instructions.

Rich: Yeah. Aboard generates a spec.

Paul: Yeah.

Rich: In fact, if you use it at aboard.com, you’ll get a PDF of the description and you can take it wherever you want.

Paul: That’s yours. Go see if you can—literally, go see if you can do better. I’d love to know.

Rich: Yeah, yeah.

Paul: For real. And reach out.

Rich: How do people reach out?

Paul: They send an email to hello@aboard.com. They can go and try out our tool and put a prompt in. If you see something surprising or confusing to you, we’d love to hear about it. We’re still very new. But we’d love to hear from you. Get in touch. And then next week we’re going to talk about how we’re building this organization, and how we’re going to provide services.

Rich: Awesome.

Paul: And then we’re going—these are the two where we just sort of talk about where we’re at, and then we’re going to bring back the rest of the industry into the conversation.

Rich: Sounds great.

Paul: More guests.

Rich: Thanks for listening.

Paul: Thanks.

Rich: This sounded like an ad. I guess it was kind of an ad.

Paul: It’s an ad. It’s okay.

Rich: But guess what? It’s a killer tool that can save you a ton of money. Oh shit. I did another ad.

Paul: It’s okay. I had a really good time building this thing with you. There were some ups and downs. It’s been a long couple of years to get here, but, man, I like this thing.

Rich: It’s cool.

Paul: Because everyone can get their thing. So go check it out. We love you, and we’ll talk to you soon.

Rich: Have a great day.

Paul: Bye.

[outro music]