Paul Ford: Hi, I’m Paul Ford.

Rich Ziade: And I’m Rich Ziade.

Paul: And you’re listening to Reqless, R-E-Q-L-E-S-S, the podcast about how AI is transforming software. You know what, Rich?

Rich: Yes.

Paul: Today we have a guest who is directly using AI to make software. So we’re going to talk about it.

Rich: Yeah. Okay, great, let’s do it.

[intro music]

Paul: So today in our little mini studio in our temp office, as we grow our AI business, we have Ms. Sara Chipps.

Sara Chipps: Hey! Hey, everyone.

Rich: Welcome.

Sara: Nice to be here.

Paul: So Sarah is a very, very long-time programmer and sort of advocate in the programming community. Has been at places like Stack Overflow, where you and I did a podcast together many years ago. Is currently at LinkedIn. But you’re up to something a little different. So I think maybe, take where I started there. Tell us what we need to know and tell us what you’re up to.

Sara: Yeah, thanks. So I’ve been, I’ve been taking some time off of work, bit of a sabbatical, and I had heard about Cursor…probably about six months, maybe six months ago. Did I make that up? How long have they been around? Like, not long.

Paul: It’s about six months. So tell people, tell people what you did and do for LinkedIn.

Sara: Yeah.

Paul: And tell people what Cursor is.

Sara: Yeah. So for LinkedIn, I lead our frontend web-engineering team. We make infra for LinkedIn, build our products, our dot-com, our LinkedIn Learning, the sales sites, we support those engineers that want to build frontend.

Paul: So components, lots of components.

Sara: Tons of components.

Paul: [laughter] Okay.

Sara: We’re more the JavaScript side. We also have a team that does, focuses on UX. So all the libraries, all the dependencies, all things like that.

Paul: Yeah, there’s a lot of frontend in LinkedIn.

Sara: There’s a lot of front end.

Paul: I’ve noticed.

Rich: [laughing] Boy, is there.

Paul: So anyway, that’s LinkedIn. Rich and I would love to just say things about LinkedIn for the entire podcast, but we’re not going to do that.

Sara: [laughing] Okay, great.

Paul: We’re going to go, so here you are, you download this thing, Cursor.

Sara: Yeah.

Paul: Everybody got really excited about it in Silicon Valley for, like, 40 minutes.

Sara: Yeah.

Rich: It’s still kind of the hot thing.

Paul: It’s still going.

Rich: What is it?

Paul: Yeah.

Sara: So Cursor is a version of Visual Studio Code. It’s a layer on top of Visual Studio Code that you can use to interact with different AI models to build software.

Paul: Okay, so it’s an IDE.

Sara: Yes.

Paul: That has AI baked in.

Sara: Yes.

Paul: Okay.

Sara: What’s the fascinating thing about it, is when I learned about it, it was built by teens. They’re, like, all… [laughter]

Paul: I just see that as almost, like, a little sticker that they put on it. Like, “Built by teens!” [laughter] Yeah, exactly. That’s okay. It’s built by teens.

Sara: Yeah. And basically, like, they didn’t build Visual Studio Code, they’re not building the AI models. But this is a layer on top of the AI model.

Rich: Right.

Sara: So you can use Claude, you can use ChatGPT, you can pick what model you’re using. But the layer on top of it is, the user experience is so good, I think, that it became incredibly valuable very fast.

Paul: I remember once I saw a teen project, and the teen project was, “Hey, you can use Wikipedia as, like, an actual knowledge graph and explore it.” And I was like, “Oh, that’s really interesting. I wonder how they did it.” And they just recursively queried the crap out of the site.

Rich: Yeah.

Paul: Like, they just went and just hammered with thousands of queries.

Sara: Yeah.

Paul: Because teens don’t care. [laughter]

Rich: Yeah.

Paul: They’re not, like, “I wonder what my rate limit is?” Or, “Am I allowed to do this?” They’re like, “Nah, we’re going to go—” Okay, so—

Sara: Yeah.

Paul: —you pick up this teenage code.

Sara: Yeah, yeah. And what the neat, really, the neat thing about Cursor is—I feel like it’s made coding fun again. I feel like, you know, in the early aughts, or the late aughts, rather, I would say, coding was really fun, because we were all doing things that were new. There was a lot of new JavaScript libraries, and things you could do. JQuery, Ruby on Rails, all those things.

Paul: Well you know what, it was a moment when you could really just get kind of a lot of bang for the buck.

Sara: Yeah.

Paul: Like, you’d go in, you’d be like, “Oh, thank God, I can finally do that.” JQuery just sort of, like, took a lot of web sloppiness away and made the web behave.

Sara: Yeah.

Paul: And suddenly you could build things, like, 10x faster. And then frankly, like React showed up and everything started to really slow down again.

Sara: Yeah.

Paul: It’s like, hey, no, actually it all has to be purely functional. It all has that—we’re going to do this with ORMs.

Sara: TypeScript.

Paul: Yeah! Like now you need to learn about types. And everything became, like, orthodox rules for really big systems.

Sara: Yeah. Who did that?

Paul: Money. Money did that. [laughter] Because, no, because you need… Why do people like TypeScript? It’s pretty excruciating to work with, but it does lead to longer term reliability in theory.

Sara: Yeah, yeah.

Rich: I think it’s cyclical. I think what happens is these cool tools show up, and the barrier to play with them is so low that you could mess around in the—to me, inspect element and the ability of the browser to really be a tool to mess with things, I think, allowed people to play. And then what happens is comp-sci reimposes itself.

Paul: Yes.

Sara: Yeah.

Rich: Inevitably. Inevitably.

Paul: You know what’s fascinating and I wonder if you’ve seen this, too. The thing that Cursor in the AI LLMs enables is that kind of really orthodox, like, disciplined programming is actually in reach for more programmers than ever before. Because you can be, like, “Okay, now make it TypeScript. Okay, now make all the types ROOT.”

Sara: Yeah.

Paul: You can actually make the system behave—

Rich: Yeah.

Paul: —in ways that used to be absolutely boring and excruciating.

Rich: I think that’s right. I just wonder if—you’re a seasoned programmer. I think for young people, you can make the system behave, but you might not understand what the heck’s going on. I think that’s the other challenge.

Paul: Yeah, but then you know what’s going to happen and we, okay, sorry. You know—

Rich: No, I want to ask the question. What made it—what’s making it fun again?

Sara: I think the thing that makes it fun again is the exact thing that you’re talking about. Right? Like, a layer, like, another layer that makes it faster to build things. You could do binary search. But you know what’s better at it? Computers.

Paul: Boy, they are. They’re really good at binary searches. [laughter]

Rich: Yeah.

Paul: It’s like, binary’s in the word, in the term, right?

Sara: It’s like their favorite thing.

Paul: Yeah, yeah. They’re all over it.

Sara: Yeah.

Paul: This is real. It turns out that, like, we were all—there’s a way to look at this. One is, it’s going to take all the jobs and that’s unfortunate.

Sara: Mmm, mmm.

Paul: I don’t think that’s actually real.

Sara: No.

Paul: But that’s one way people look at it. The other way to look at it is we were all wasting an enormous amount of time doing the same thing over and over and over again.

Sara: Yup.

Paul: And it was pretty boring.

Sara: I think a lot of people are afraid. But the thing about it is, it’s not—it’s not 100% there. It’s not, like, you’re, like, “Okay, this is the program that I want, and it’s going to make exactly what you want.”

Paul: You know, specifically, so I use Aider, A-I-D-E-R, which is, like, an open-source—well, I mean, Cursor is kind of open—

Sara: It seems like Cursor, but, like, it makes you feel like you’re doing more.

Paul: Yeah. And it’s more, it’s more in the terminal, it’s actually more, like, Python-oriented.

Rich: Take a minute and explain what Aider does.

Paul: Aider is similar to Cursor. I go into my terminal program, like I was programming in 1981.

Sara: [laughing] Yeah.

Paul: And I’m in a folder and the folder might have some code in it or not. And I go into Aider, and what Aider does is open up a chat in the terminal with the ChatGPT or Claude or DeepSeek or any one of these. And I say, “I want to build a database schema that lets me manage my hat collection.” And it goes, “Yeah, sure, I think it would look kind of like this.” And I’d be, like, “Okay, well let’s, let’s actually run that in SQLite and build the database.” And it’s like, “Yeah, I did it for you. How’s that feel?” And I’m like, “Feels real good. Why don’t we build an API from that?” And it’s like, “Well, I would use Swagger and a little JavaScript.” I’m like, “Why don’t you do it?” Instead of just chatting, what it does is the text it’s spitting out, everybody’s seen this—when you ask the thing to make something and it makes you like a little program, it writes a little program like right in the chat?

Sara: Yeah. And you can watch it.

Paul: It saves that as a file, right?

Sara: Yeah.

Paul: That’s actually all that’s going on with these cool new tools.

Rich: It’s still a handoff.

Sara: Yeah.

Rich: You still have to understand what’s up. You still have to kind of police it. You still do have to understand conceptually what’s going on.

Sara: Yeah.

Rich: It’s not just giving you shrink-wrapped finished code.

Sara: Exactly. It’s like, I feel like it’s like working with a junior developer and you’re pairing.

Paul: Mmm hmm.

Sara: You’re like, “Okay, like, why don’t you write this thing? I’m gonna take a look at it.” And sometimes, like, they nail it and sometimes you have to give them feedback. It’s really, it’s really engineering managers’ time to shine. I feel like I’m engineering managing someone without emotional issues.

Rich: Mmm.

Paul: You know, it is like, it’s real, it’s a good moment for a certain kind of product manager.

Rich: Wait for 5.0.0, which adds emotional issues to the feature set. [laughter] It’s coming.

Sara: “I’m really stressed. I didn’t get a lot of sleep last night.”

Rich: All right, so you are, you are a seasoned engineering manager. This fits for you.

Paul: So actually, give us some metrics here, right? So, like, Aider says that 60% of the tools, of the programming challenges in its benchmark are being met by the best LLM. So 40% of the problems that you might give a junior engineer, it has trouble solving; 60% it does solve.

Sara: Yeah.

Paul: Right? So that’s a, that’s a stat from out in the world. Ballpark a little bit. Like, what’s different from Sara programming two years ago—

Sara: Yup.

Paul: And Sara programming today?

Sara: The biggest thing I think is I don’t Google anymore. Like, even a day to day, I don’t go on Google.com. I go just to ChatGPT. And so—

Rich: To get an answer, or…

Sara: To get an answer.

Rich: To pass a snag somewhere.

Sara: Exactly. And so even so, it’s like, I’m in the IDE. I’m like, “Hey, I’m thinking about doing this thing. How would you approach it?” Or like, “I want to add…” I’ll think of an example. I want to improve this prompt. I want to improve this prompt that I’m using. Here’s the prompt, here’s what I’m getting back, and here’s what I want.

Paul: Do you find yourself… What I find myself doing is constantly rehearsing and repeating a task to learn how to do it, or are you mostly building things and going on to the next one?

Sara: I’m building things. I’m looking at it. I’m seeing how, like, how good of a job it’s doing. I’m seeing what I would change. Because one thing I’ve noticed is it will add a lot of unneeded complexity if you don’t babysit it.

Rich: Mmm.

Sara: Right? Like, it’ll start, it’ll interpret what you said literally in a little weird way. And then all of a sudden you have an extra loop or you have something going on that you don’t need. And so you can go in there, you can look at the code, you can edit it.

Paul: How do you babysit? What is the process of babysitting?

Sara: I look at the suggestions. So like, I’ll say, “Here’s what I want to do,” you know, like, I want to add a button here or, I want to get an email after I do this thing. It suggests the code. So I just look at the diff. It’s like, it’s like looking at a PR, right? Like you’re looking at the diff. You’re looking line by line. You’re giving feedback on a PR, basically.

Rich: The junior programmer gave it a go.

Sara: Yeah, yeah, yeah.

Rich: And you’re reviewing their…

Sara: Exactly.

Rich: Their code.

Sara: How’d you do?

Rich: Yeah, how’d you do?

Paul: What’s the outcome here? So more joy.

Sara: Yeah.

Paul: Okay. You’re enjoying the actual craft again.

Sara: Yeah. Less tedium.

Paul: Which is, which is real, because it had gotten real tedious. I wasn’t building anything because it was—

Rich: She’s feeling more productive. You’re feeling more productive.

Paul: Yeah.

Sara: Yeah. It can be fast. Yeah, yeah. Exactly.

Paul: How much faster are you, do you think?

Sara: Oh, worlds fast—I’m as fast as I was in, like, 2008.

Paul: Just, just…

Sara: Just moving.

Paul: Chomping.

Sara: Yeah, yeah. Just getting out there.

Paul: Okay.

Sara: I’m like, and it’s super fast. Like, I’m like, “Oh, I want to build this—” For example, on Monday, I spent my day building a machine-learning engine to test a hypothesis I’ve been working on for a while. That would have probably, by myself, taken six, eight weeks.

Rich: And you did it in a day.

Sara: Yeah.

Paul: It’s not just that. You never would have done it.

Sara: I would have been like, “I don’t, I don’t even know where to start.”

Paul: That’s where I’m at. I’m like, “Oh, things that everyone has said matter, like machine learning or—” And you can actually get it to explain what it’s doing as it’s going, or add more docs and so it becomes like the super-senior engineer.

Sara: Yeah, yeah, yeah.

Paul: It’s pretty nice that way.

Sara: It is.

Rich: Let’s go through a few different roles here and let’s give each of them advice. Two-years-out-of-college programmer.

Paul: Because you are, you have managed reasonably large teams of engineers, right?

Sara: Yeah, yeah.

Paul: Okay. Let’s, let’s fix their careers before they all go away. [laughter] Okay, so you said the two-years-out-of-college programmer?

Rich: Two-years-out-of-college programmer.

Sara: I think the thing that is going to suffer the most is frontend developers.

Paul: Because it just writes components all day.

Sara: It writes components. It makes, like, a…it makes design-system…you just, you know, like it’s like design-system plus-plus, like, you don’t need the person. And so I think a lot of new grads were focusing on frontend or, like, web apps. Like I think the thing is, we built all the internet, and now, like, we’ve done all the things. There’s nothing new or innovative that we’re doing with the internet right now.

Rich: Yeah.

Sara: And the AI has consumed all of it and knows how to do it.

Paul: Well, it’s excellent at boilerplate, right?

Sara: Yeah, exactly.

Paul: When you, the kind of frontend you’re talking about isn’t, like, weird processing, hacking to make exciting—

Rich: It’s not a new interface paradox, yeah.

Paul: It’s like, I’ve gone into the bank, I just got my certificate, I’m 26 years old, and they’re telling me, “Make a drop-down for state entry, but it has to work a little differently.” And I’m gonna—that might take me, like, six hours.

Sara: Yeah.

Paul: And it’s, like, 30 seconds now.

Rich: Okay, so I got my comp-sci degree. I’m two years out. Just now they trust me enough to take tickets.

Sara: Yeah.

Rich: What do I do?

Sara: I think the thing that isn’t solved quite yet is, and I think something that will need people for a long time—if you fundamentally understand, so like, we’ll pick JavaScript. That’s where I’m comfortable. If you fundamentally understand how vanilla JavaScript works and how libraries work with each other.

Paul: [comical horrified voice] Vanilla JavaScript? Sarah! [laughter]

Sara: Gross.

Rich: When you say vanilla, you mean not—no framework.

Paul: No—

Sara: No framework.

Rich: Just the raw…

Paul: No types?

Sara: No types, no nothing.

Rich: Understand the fundamentals is what I’m hearing.

Sara: Understand the fundamentals. Because what the AI, what I observed it can’t do is understand the implications of the decisions that you’re making now. It will listen to what you’re asking it to do.

Rich: Yup.

Sara: So it’s almost like we’re asking those new engineers to become a little more intermediate or a little more experienced. So it’s going to take more I think, on the, like, on the education side to learn, but those people are still very valuable.

Paul: Well, I do think, you know, I think about a teaching IDE, right?

Sara: Yeah.

Paul: Because this thing can explain what’s happening all the way up and down the stack.

Rich: Well, I hope you have the con—I mean, what is, all of programming is like a dozen or so concepts, right? There’s certain structures and processes that you just, you should understand that outside of syntax at all, right? And I think one of the byproducts of, of just all these really great libraries that got created is they sort of insulated a lot of learning. You didn’t have to learn a lot.

Sara: Yeah.

Rich: Like, you could just pick things up and just glue them together and you didn’t really understand how they work.

Paul: You could use NPM whatever, and just kind of go get a bunch of code.

Rich: Yeah. And I think what you’re suggesting here is, like, understand it, because you’re gonna need to understand it because it’s not gonna finish the job. It needs you to finish the job.

Paul: I actually have a great—if you want to truly learn vanilla JavaScript, I have a real shortcut for people. There is a game online, it’s amazing. It’s called Universal Paperclips. It’s a clicker game.

Sara: Oh, like a cookie clicker.

Paul: Yes. And it’s about a—it’s actually about AI taking over the world and turning the universe into paperclips.

Sara: Yes.

Paul: You play as the AI. If you look at the source code of it, it’s absolute plain vanilla JavaScript.

Sara: Wow.

Paul: Top to bottom. And so you can—t’s actually one of the better learning tools I’ve ever seen.

Sara: Fascinating.

Rich: Okay. The engineering manager.

Sara: Yeah.

Rich: They’ve got a dozen engineers.

Sara: Yeah.

Rich: And obviously they’re doing their every—each engineer is playing with their own toys. Cursor, Replit, there’s all kinds of stuff out there.

Sara: Yeah.

Rich: Advice for that person?

Sara: Yeah. Like, what should they be thinking about?

Rich: And how should they be guiding their team?

Sara: I think everyone is aware there will be a reduction in the workforce. I don’t—well, I don’t know. Do we all agree there will be a reduction in the engineering workforce?

Paul: I go back and forth, because it’s also, paradoxically now, the backlog is actually decades long.

Sara: Yeah.

Paul: And everyone’s just used to not getting what they need.

Sara: Yeah.

Paul: So everybody can get all the software they ever wanted.

Rich: I think it’s a reduction in roles.

Sara: Reduction in roles.

Rich: And I think there will be new roles.

Paul: Yeah.

Rich: I don’t know what those exactly look.

Paul: Solutions-oriented people who use these tools.

Sara: Yeah.

Paul: As opposed to, yes, I wouldn’t say… Frontend JavaScript development for Fortune 100 corporations as part of a team of 150?

Sara: Yup.

Paul: Like, I don’t know if that job can survive as a gateway job to the rest of engineering with these tools.

Sara: Yeah. Yeah.

Rich: It’s also just the title of, like, engineer, comma the third dropdown. Like, that’s been focused on one corner of the code for years.

Sara: Yes.

Rich: That, I think, has to be something.

Paul: I’ll pitch you a different job. Report specialist. There are so—we need so much reporting in this giant—

Sara: Oh yeah.

Paul: Not-for-profit or bank or whatever.

Rich: To this day, to this day, we talk to business stakeholders and customers all the time. They just want the thing and they can’t get the thing right.

Paul: Only executives have been allowed to have dashboards for the last 20 years.

Sara: Oh, it’s true.

Paul: Right? So now everybody can have their dashboard. They can have their PDF mailed to them. And I can be the report specialist, and I can work with 100 people a year to give them the analytics and reporting they need in order to do a better job.

Rich: I think it’s a migration.

Sara: Yeah.

Rich: I think it’ll net out to more jobs, I think, in the long run.

Sara: Yeah.

Rich: But I think you have to be open enough to sort of shed that expertise that you thought was, like, so defensible.

Sara: Yeah.

Paul: Syntax as a territory you defend. Like, when I was joking about you saying vanilla JavaScript, I—people who don’t this world don’t know that vanilla JavaScript is gross and tacky. [laughter] Right? Like, you’re supposed to use frameworks and you’re supposed to use a certain kind of functional programming.

Sara: I think someone would say gross and tacky. Other people would say incredibly cool. But you know, it depends on who you talk to.

Rich: Yeah.

Paul: Also there are things like—

Paul: Very few might say incredibly cool, Sara. [laughter]

Rich: Let’s go back to that engineering manager.

Sara: Yeah.

Rich: You were saying, so look, I think some jobs are going away. I think that is real. I think some roles and responsibilities are going to have to change. So what do they do? Like, how should they approach it?

Sara: Yeah. If I’m an engineering manager and I’m thinking about this, one thing I think we can be sure of is, like, the navel-gazing of the industry is kind of going away. This, like, I’m a special flower. I’m doing this amazing job. You know, like, I think that being a very skilled engineer and without professional skills was something that was okay. And even being a very skilled, a good manager that people liked without professional skills used to be okay.

Paul: So, like, Grumpy Carl in the corner.

Sara: Yeah.

Paul: But he really is, he really understands databases.

Sara: He’s so smart. I know he makes everyone cry, but he’s a 10x. [laughter]

Paul: He keeps, he does that thing where he touches the girls on the shoulder, but he’s really good at databases. [laughter]

Rich: Oh God!

Paul: No, there’s a lot of Carls out there and there’s also, like…

Sara: Yeah.

Paul: The Carl who’s really good at COBOL, and the Carl who’s…

Sara: Yeah.

Paul: Yeah.

Sara: Yeah. And so I think all the Carl—I think, if I’m an engineering manager, I’m focusing on, I’m learning more about the business. I’m understanding, like, what does a business need? What can my team deliver? And, like, my role may evolve, and if you become very—if you’re already in the role, you’re very useful to the business. It doesn’t matter, like, skill set. And I think, like, to your point, our roles may evolve. I don’t think we’re going to get rid of engineering managers. I think there’s still going to be a need for them. But as the team evolves, you’ll need to evolve as well. So just be, like, I would be thoughtful, I would stay current and I would be thoughtful about the contributions I’m making to the business in general.

Paul: So you’re part of a lot of, kind of different conversations around programming through different groups and different, your social network and your friends and so on. Probably more than we are. Like, we’re kind of doing our little startup over in the corner. So when you’re talking to people—doesn’t have to be LinkedIn, obviously. Like, I’m not asking you to betray any confidences or anything. But, like, you’re seeing what we’re seeing, which is, wow, a radical transformation is happening to a multi-trillion dollar industry that is also, like, supposed to be the sort of prince or princess of the entire world economy.

Sara: Yes.

Paul: And it’s really happening fast. And you said, “Hey, I better go actually do this,” because, I think just middle-class roots, like, you just, I better get the screwdriver out. Let’s go. Right?

Sara: Yeah.

Paul: Are other people seeing it? Do you think, like, big orgs are actually aware of this level of change?

Sara: Yes, very, I think very aware of this level of change. I don’t think anyone knows quite yet. Like, no one’s ready to make the big decisions yet. Everyone is trying to build AI into their software.

Paul: Right.

Sara: And I think it’s the biggest challenge is to the larger organizations because their product is established and it’s a lot—it’s one of those worlds where it’s a lot faster to start from zero than it is to build it into something that they want—

Paul: They want to put that chat on and say we did it.

Rich: They want to glue it on. It’s funny, I, I was talking to a, a potential client of Aboard’s and, and like no joke, their business is essentially run on spreadsheets, like, duct taped together.

Sara: Yeah.

Paul: It’s a big, it’s a big business.

Rich: It is a big business. Not atypical. And we’re having lunch, and you know, he’s reading up on AI. He not a technical—he’s the CEO. He’s the founder of the business. And he spitballs, like, 11 cool AI agent ideas of how his business can be better.

Sara: Yep. Yep.

Rich: And then I couldn’t help but ask him, I was like, “That’s really cool. But you still have a dozen duct-taped spreadsheets.”

Sara: Yeah.

Rich: Like, you skipped—

Paul: Well, it’s also like, when—

Rich: —all of it. You’re just gonna—

Paul: When someone sends you an email, Jeff has to open up, like, a Word doc and cut and paste into it.

Rich: Yeah. It’s still the case. I had a friend in high school who had, like, a Honda Civic.

Sara: Yeah.

Rich: And he, and whenever he made any money, he would just get, like, more little things to put on it. [laughing]

Sara: Oh yeah, me too. That was me. Like, stickers.

Rich: Yeah!

Sara: I had flame stickers on my Honda Civic. They were like $300.

Rich: In, like, script, it just said “respect” on his door.

Paul:  [weary sigh]

Rich: Of his Civic. And he got a spoiler. Meanwhile, I don’t know if, like, the aerodynamics matter on your Honda Civic.

Sara: I wanted the spoiler.

Paul: So there’s a reason why marketers love the numbers, 18 to 35 are magical. [laughter] Right? It’s like—

Rich: By the way, I’m clearly placing myself in Brooklyn right now. [laughter]

Paul: But, like, I think there is, like, some little cognitive pop, like, some part of the brain just hasn’t settled. Because you did it. I did it, too. I’m like, “I think I need a, I need a synthesizer.” And you know, the killer for me was like Linux desktop computers, like it just like, “Yeah, yeah, now I’ll be cool.” [laughter]

Sara: Were you putting flames on the side of your Linux?

Paul: Yes. Yes, I was, Sara.

Rich: Back to. AI. It feels like no one wants to do the ugly part—

Sara: No.

Rich: —of, like, getting off the spreadsheets. They want the cool robot.

Sara: They want the robot to show up and do it for them.

Paul: Let me ask you a question I’m asking other people and then we’ll do a quick shift to a very different subject. If you could put it back in the box or make it go slower, would you?

Sara: Oh, I’d make it go slower. 100%.

Rich: Huh.

Paul: I think, yeah, yeah.

Sara: Yeah. Because I—

Paul: [laughing] It’s a lot.

Rich: Talk that through.

Sara: Yeah. Well, I mean, I think one thing we used to do in this industry a lot is say, “Should we?”

Paul: Yeah.

Sara: I think until like, it became super profitable, until the, everyone with an MBA was becoming an engineer, we used to say, “Should we do this? What is the effect it’s going to have on the world? And is that effect more negative than positive?” I don’t think we ask that question anymore. We’re definitely not asking that question with AI. Like, the people you see, like, at OpenAI, different places, the people that are asking, “Should we?” Are getting fired, leaving the board, quitting, all those things. And so I think when we—

Rich: It’s aggressive right now.

Sara: Yeah.

Rich: The posture is very go, go, go.

Sara: Yeah.

Paul: Well, it’s—they’ve over-promised to a point that they can’t deliver unless they actually crack the fundamental problem of human consciousness.

Rich: Yeah.

Sara: Which we’ve been trying to do that for a while

Paul: We’re on year seven seventy of that.

Sara: Right.

Paul: So we’re closer.

Sara: Yeah, yeah, definitely. Yeah. And so I think, I think if we slowed down we could start to ask those questions. I think we’re not going to.

Rich: There’s just a, it’s just a race right now.

Sara: Yeah.

Paul: Look, the positive case actually says the same thing, which is everybody is, like, look, even if there’s no more progress, it’s going to take us 10 years to unlock what these things are and what they can do.

Sara: Yeah.

Paul: The growth that comes from them. So part of me is like—I’m with you, which I’m just like, whoa. But you can’t put anything back in the box. So I get it.

Sara: Yeah.

Paul: But it is a level of change. It’s one of the things we’re really struggling with because we’re trying to communicate how fast you can build now. And the product that we’re going to go out with will show unbelievable speed.

Sara: Yes.

Paul: And I’m actually finding it hard to communicate that.

Sara: Yes.

Paul: Because people can’t tell the difference between the demo and the product. They can’t, like, it’s just very—everyone has just been hit and—

Rich: It leaves them actually, also, a little anxious. “Wait a second. That’s neat that you did that. But I can’t put my whole business on that. That’s too scary.”

Sara: Yeah. I think it’s like, with all new technologies that we’ve seen related to this, it’s like, the “Hello, world” phase where you go in with hello, you can make a sick “hello world,” and everyone’s, like, “Oh my God, this is going to revolutionize everything.” When you try to do it from soup to nuts, it’s like, “Oh, here’s what’s missing.”

Paul: It’s still a big ugly set of engineering problems to build software in the world. That’s real. Right? We, we talk a lot about the “$15 Volvo” because nobody wants a $15 Volvo. You think that sounds good, but imagine if you saw that on Craigslist. You’re like, “I’m not putting myself in that.” [laughter]

Rich: “My children.”

Paul: Yeah, I’m not gonna, like, that thing is gonna explode. Let’s do a quick shift because there is a forbidden topic that we should—

Rich: Don’t be dramatic, Paul. [laughter]

Paul: So Sara very graciously came to the office recently and shared some of what she’s working on. We’re trying to get people together to just talk about how AI is changing software and try to, because New York City is a little behind. Right? Like, we’re, it’s always trying to figure out, like, how can it still be New York but also use technology.

Sara: Yeah.

Paul: And so you are building crypto-related things.

Sara: Yeah.

Paul: Crypto!

Rich: But you’re using AI to help you build them.

Sara: Yeah, yeah, exactly. That’s been really neat.

Rich: So tell us what you’re doing.

Paul: And talk a little bit about what’s going on in Crypto World with AI and with this stuff.

Sara: Yeah. Well, I think that like, Crypto World is being run exclusively by 12 year olds right now.

Paul: That’s great. [laughter]

Sara: Yeah, it’s, it’s all—

Rich: As a father of a 12 year old, that’s a very concerning statement.

Sara: Don’t let them hear. Don’t give them too much cash.

Rich: Yeah.

Paul: Your son has $2 million. [laugther] You don’t know about it.

Rich: I don’t even know it yet.

Paul: He’s just going to come back one day with, like, a lot of Skittles. [laughter] All right, keep going.

Sara: Yeah, it’s all, and so, what’s really interesting, like, keeping, like, it’s finance, it’s impossible to keep ahead, but, like, looking for trends, figuring out where people are investing their money, figuring out, like, what’s exciting, what people are doing that—

Paul: So more of the cultural part of it. Like, where is the money going? Where is, like, what are people thinking is valuable and exciting?

Sara: Yeah. So I think the thing that’s really, it’s an interesting time because—

Paul: Tell us about the coins.

Sara: Yeah, it’s all, it’s all meme coins. Last week was a big week for farts. There’s Fartboy, Fartgirl. There’s probably about, like, I mean there’s millions of coins that are meme coins that are created all the time, but there’s trends. So last week it was all, like I said, all farts.

Rich: What’s the market cap of, like, Fartboy?

Sara: Fartboy’s market cap I think was like $37 million.

Rich: Okay.

Sara: So it’s not like—

Paul: Go ahead, let’s open that window so I can jump out. [laughter]

Rich: And Fartgirl?

Sara: I didn’t check the market cap of Fartgirl, but, like, Fartgirl came after—Fartboy was a big one. It was big because they made, like, a, you could tell there was thought put into it because there was a, there was an illustration of Fartboy. He had, like, a superhero outfit on. People had put a whole, like, website together. I mean, all these things take about an hour. But someone did it.

Rich: They marketed well.

Sara: They had some marketing. Yeah, yeah,  yeah.

Rich: This isn’t NFTs. This is an actual coin.

Sara: No, this is an actual coin.

Rich: Okay.

Sara: Yeah.

Rich: And so this thing goes out and there’s interest because they’ve created some buzz around it.

Sara: Yeah, yeah. And people buy them. And people sell them. And things get rugged all the time. Like, so rugged is like. So I, I keep telling people this story because I think it’s really indicative to the industry. In December, there was this kid. He was just, like, a random 12-year-old kid on the Internet. His family, like, they’re not in finance, or not—they’re, like, in the middle of U.S.—

Rich: Goofing around on his computer.

Sara: And he’s, like, “I’m gonna make a coin.” He had some kind of social following, like, a social community. So he made a coin. It was called, like, Gen Z something—something cool. [laughter] And he put $350 of his own money in the coin. Marketed it to his community. His community starts buying. He runs downstairs. He’s like, “Dad, I just made $30,000.” His dad’s like, “Okay…” You know, like not believing.

Rich: Sure. [laughter]

Sara: “Are they like Fortnite Bucks, or what’s going on?”

Paul: Yeah.

Sara: And so what the kid does is he takes his $350 investment out, which is $30,000 of the coin. Tanks the coin. Everyone else’s money goes away. And so the kid’s made $30,000. He’s, he’s amped. Everyone else is pissed off. Everyone’s pissed off.

Rich: Like, how much money was lost in that?

Sara: I don’t know. If you were, if you were to guess, like, I mean, probably a few hundred thousand dollars.

Rich: Okay.

Sara: If you’re, if you’re taking from $350.

Paul: This is a 12 year old’s afternoon.

Sara: It’s his afternoon?

Rich: Yeah. Okay. So they’re angry at him.

Sara: So he’s like feeling bad. They’re angry at him. And so he’s like, the next day he goes back to his community, and he’s like, “You guys, I’m so sorry. I didn’t know this would happen. I wanted to give you an opportunity to make your money back. Like, I didn’t want to take everyone’s savings that like, that was really hard. I’m gonna make a new coin. It’s called I’m Sorry. You can buy I’m Sorry coin. I’m not gonna do this again.”

Paul: [laughing] He’s gotta rug pull again in a minute. Yeah. Come on, come on.

Sara: That’s exactly what happened. [laughter] Everyone bought I’m Sorry. He did it again.

Rich: No!

Sara: It’s amazing. Everyone was furious. It’s not illegal, it turns out.

Paul: No.

Sara: Because it’s not an actual currency.

Paul: Craziest thing about this? He is our new treasury secretary. [laughter] So…

Rich: What’s your take on all this?

Sara: Yeah.

Paul: And bring it back to, like, AI and what you’re like, trying to figure out.

Sara: So it’s like the Wild West. I always like the communities that are like the Wild West, where everyone’s just doing something crazy.

Rich: Yeah.

Sara: You don’t know what’s going on. You know how it’s going to—so what I’m just doing with AI is I’m making tiny investments. I’m seeing how they’re performing. I’m seeing, like, I’m doing things like what market caps are promising? Like, right now, like, everyone’s coming out with the Dave Portnoy, Portnoy—

Paul: The Barstool Sports guy.

Sara: Yeah, he just came out with a coin this week.

Rich: I’m sure that’s big.

Sara: Everyone’s coming out with—so you have to think about the Dave Portnoy coin is dilut—is not taking money from the, from the Fartboys of the world. The people selling their Fartboy, putting it into Dave Portnoy. So you have to think about like, what, like, what does a bear market look like for meme coins? What does a bull market look like? What size of coin should be you be investing in right now? So that’s what I’m doing with AI is, like, making tiny investments, monitoring them, seeing what works, seeing what doesn’t. It’s super easy.

Paul: Cool.

Rich: Yeah.

Paul: I’m going to start huffing paint. [laughter]

Rich: Yeah. Yeah.

Paul: Okay. So for you, it’s like, so you’re, you’re building, you’re using Cursor to build a little observatory that then lets you take action based on stuff that’s happening in these worlds.

Rich: She likes the chaos of it.

Sara: Yeah.

Rich: Is what I’m hearing.

Sara: Yeah. It’s fun.

Paul: No, I get, I find this, too, which is because you can build in such an accelerated way and so many of the boring parts get taken away.

Sara: Yeah.

Paul: That you start to do things you would never like, that would have been an unbelievable amount of work.

Sara: Yeah.

Paul: This is this, actually, I’m going to bring this back in a kind of weird way. When we talk about all the jobs going away? You never would have had the time to do this well. You might have hacked around, written some scripts and so on and so forth, and you might have dabbled, but actually what you’re doing is probably equivalent to like two or three people in a startup.

Sara: Yeah.

Paul: Right? But it’s just you with the time you have.

Rich: And you get to see these outcomes.

Sara: More than two or three people. I say it like, it’s like a team of 10.

Paul: Yeah, that’s right.

Rich: Which is insane.

Paul: I mean, I am able to take just about any SQL schema, and using a set of prompts that I’ve rehearsed over and over again, I can get it to a really robust Rest API.

Rich: Yeah, I think allows for—

Paul: I can do that in like an hour.

Rich: This isn’t just about productivity. It’s about exploration. I think, I think I have young kids and I have friends who are asking me what do I, I want my kids to be technical and learn computers. But what do I tell them now?

Paul: Just say, just look them in the eye.

Rich: What I tell them now—

Paul: No, no, you just say one word. You say: Fartboy.

Sara: Fartboy.

Rich: Yeah, I say Fartboy a lot.

Paul: You just walk away.

Rich: Yeah, but no, but I think the message is like, hey, they have these incredible tools to explore with, right? Like, we had them, we’re old. So I take apart my computer and I would like, I would code. I learned BASIC was the first language I learned and that was like, it was kind of this open-ended space to play. And I see that a little bit here. The thing is, and I want to go back to this point, which is you have to understand those fundamentals. And I worry about, like, I think Scratch, which is this like kids programming tool, they got to reboot it now so that it is in the context of understanding these principles, so that you can play with all these amazing new things. I think that—you could see things dying or you could see new opportunities. I think that’s, that’s the difference.

Paul: I want to throw a question to close this out. You have been on and off for many years a hardware hacker.

Sara: Yeah, yeah.

Rich: Oh!

Paul: Tell the people about what you, what you built.

Sara: Yeah, I built a company called Jewelbots. We made smart friendship bracelets for young girls that could teach them how to code.

Rich: Very neat.

Sara: That was very fun. Yeah, yeah.

Paul: So now take this world—maybe not the crypto world, but like, the… How do we… What’s going to happen to hardware hacking?

Sara: Hardware hacking. Oh, that’s so interesting. I was just talking to someone about that the other day. What’s going to happen to hardware hacking? I have no idea. [laughter] No, I really have no idea.

Rich: Thanks, Sara!

Paul: No, but everybody, everybody, we’re talking to a lot of big brains and everybody gets to that point. Because something’s going to give.

Sara: Yeah.

Paul: It used to be really hard to do hardware specifications and yeah, now you could—

Sara: Just do it. You could just ask the AI to do it.

Paul: That’s right. And like, at what point I keep thinking about China with all those CNC machines. Instead of having to give it, like, various kinds of shape files and stuff, you’re going to be able to get it to keep writing code to iteratively improve the things it’s manufacturing.

Sara: Are we going to get replicators, finally?

Paul: Well. Or something replicator-ish. Because it can write code to make itself better.

Sara: Yeah.

Paul: But then turn that into things in the physical world.

Rich: You can’t end the podcast like that.

Paul: Absolutely. [laughter]

Sara: “Earl Grey, hot.”

Paul: So if I wanted, so if I wanted to buy some Fartboys—no, if I wanted to get in touch with you or see what you were up to, what would I do?

Sara: Oh, I’m off all the socials. You could email at me at Sarajchipps@gmail.com, that’s what you could do. Or find me on LinkedIn. I’m always on LinkedIn.

Paul: You better be. [laughter]

Rich: There we go. This was great.

Paul: Yeah, it’s, I, we learned. We learned. We laughed. It was really good.

Rich: There’s a lot of to be continued in all these conversations. As we talk to really smart people, the smartest ones say, “I don’t really know.”

Sara: Yeah.

Rich: “I’m not really sure where this goes.” [laughing]

Paul: How far in the future do you think you are right now?

Sara: What do you mean? Like, how far can I see into the future?

Paul: No, just like you, I feel like that we’re all a little bit ahead of where everybody else is, and that’s sometimes a liability.

Sara: Yeah.

Paul: Do you think everyone will be where you are in 6 months? 2 years?

Sara: 6 months to a year. I think it’s actually, like, the problem with the future right now is it’s really a lot closer than it used to be.

Rich: Yeah. That’s real. That’s real.

Sara: Yeah.

Paul: All right, there we go. So, Rich, if people wanted to learn more about us, they should go to aboard.com. They should send us an email at hello@aboard.com. We are building an AI-accelerated software platform. You can try a little tiny taste, a little nibble on the top right, it says, “Try it.”

Rich: Watch the trailer.

Paul: Click that button. But there’s a lot more coming. Excited to keep talking about it, but in the meantime, we’re going to talk to really smart people like Sara, who help us understand what—

Rich: Thank you, Sara, for doing this.

Sara: Thanks for having me.

Paul: —the hell is happening. Okay. Thank you. Yes.

Rich: Have a lovely week.

Paul: Bye.

[outro music]