Paul Ford: Hi, I’m Paul Ford.

Rich Ziade: And I’m Rich Ziade.

Paul: And this is Reqless, the podcast about how AI is changing the world of software.

Rich: Play that theme!

Paul: [trumpet noise]

[intro music]

Paul: All right, Rich. Today we have an easy setup for the show, thanks to a listener.

Rich: Okay.

Paul: His name is Rich. We want another Rich. Yes. [laughing]

Rich: It wasn’t me.

Paul: It wasn’t you. And was very nice about the podcast, and said “I am a UX researcher.”

Rich: Mmm hmm.

Paul: So let’s just tell people what that is.

Rich: Do it.

Paul: A UX researcher, this is from Google, is someone who studies users’ needs, behaviors, and preferences to help improve digital products and services.

Rich: Sounds right.

Paul: And they’re focused obviously on visuals.

Rich: Yes.

Paul: Like on what, what makes people, you know, do I, do people click on that button versus that button, is the obvious one.

Rich: A/B testing, and…

Paul: Then more broadly, like, it’s not just clicking on buttons, it’s sort of, like, you know, how do, how are we going to organize this knowledge? What kind of hierarchy are we going to apply? What systems do we want to use, cultural systems, really, like, is this colorful, should it be, should it have little dancing people or should it be more corporate or you know, stuff like that?

Rich: Yeah, but to clarify, they answer those questions by polling users.

Paul: Yes.

Rich: Gathering data and making decisions. That’s their research.

Paul: Yes, that’s right.

Rich: The data for their research is the users.

Paul: Yes. The thing they are researching is human reaction to ideas about design.

Rich: Correct.

Paul: Okay? So that’s what a user researcher does. Now look, a UX researcher is a, there are a lot of roles that are kind of like this. Like, business analyst with a UX focus is a little bit like this.

Rich: Sure.

Paul: A lot of designers—

Rich: Product marketing.

Paul: But the basic role in the world, and we haven’t even gotten to the letter yet, but I think this is important, is there’s a large cluster of humans whose responsibility is to say, “Hey, we want to do this thing and I should go talk to everybody and find out what they make of it. From a visual perspective, from a business perspective, from a money perspective.”

Rich: You’re polling humans.

Paul: That’s right.

Rich: To make sure you make the right decisions, because they are going to be the beneficiaries, ideally, of whatever you create.

Paul: Okay, so there’s our framework.

Rich: That’s the set up.

Paul: Setting us—so, “I’m a user researcher.” This is Rich. “You talked about the importance of conversing with the NGO”—with a not-for-profit organization—”to understand their needs, and you characterize that as something that AI can’t do yet. So that sounds like UX research to me. I don’t take offense that you didn’t label it as such. I am interested, however, to know whether you think there’s an opportunity for a person like me to make a living helping not-for-profits, NGOs, figure out what they need so that they can mostly build it with AI. And if you think that’s possible, I’d appreciate any advice on getting myself into that position.”

So this is someone who is several years into a user-experience research career. They left academia, and they want to do work that aligns with their values.

Rich: Got it.

Paul: So to reframe the question, if you are someone who goes and talks to people about what they need in terms of visuals, in terms of…

Rich: Mmm hmm.

Paul: In terms of business, in terms of whatever. Right now, that’s right. You can’t say, “Hey, AI, go talk to 500 people about the pictures.”

Rich: Right.

Paul: You can, kind of, but it just isn’t, it’s not going to work yet. So we’ll talk about that. So there’s two questions here. One is, I think there’s kind of a foundational question, which is, could the systems we have now, could you conceive of them doing what we think of as user research, user-experience research? On the flip side, what should a person who does this kind of work do? What should they go learn right now so that they’re ready for kind of what’s next in order to help orgs ship software on time?

Rich: This is a great question. It’s a great framing and question, because it actually touches on what’s coming up a lot in our own conversations. What’s happened, I think, in the, in the first couple of years—let’s park consumer use of AI. “Paint me a picture.” “Answer my question,” for a second.

Paul: Yeah, yeah.

Rich: The practitioner use of AI has really taken off, especially in the world of engineering. There are tools for designers, but it’s very practitioner-focused. What that means is it’s sort of like, here’s this wickedly awesome new enhancement to your toolbox, to your workbench, and that’s very cool.

Paul: It’s also, you know, it’s interesting. It doesn’t get as much attention because it’s much less controversial. When writers, when college students are cheating on their essays, that’s a good news story. When AI companies are soaking up everybody’s content and generating images that look like art that’s made by people who are trying to make a living off their—

Rich: Also news.

Paul: Also a news story. But people just in general, there have been some class-action lawsuits and people are like, “Hey, you didn’t honor all the licenses of the open-source when you spidered it,” and so on.

Rich: [laughing] Yeah.

Paul: That’s all going on. But for the most part, the larger world of culture is like, “Hey, nerds? Whatever.”

Rich: Have fun.

Paul: Have fun. Go use your toys, right?

Rich: Correct.

Paul: So quietly—and we keep, we see this, and I kind of want to iterate it over and over. Like, there are multiple conversations going on with this big new technology. There’s the kind of the culture and cultural-artifact conversation, and then there is the code conversation, and they’re very different.

Rich: Very different.

Paul: The code conversation is like, “Hey, everybody, this is accelerating really, really quickly and you better pay attention because it’s like a tidal wave.”

Rich: Yeah.

Paul: Right?

Rich: Yeah.

Paul: So that’s where we are.

Rich: Yeah.

Paul: And you’ve got programmers writing modular code, like, my React, like, the components that I click on.

Rich: Yeah.

Paul: I can write that in a minute now.

Rich: Yeah. And it’s amazing. And, and—

Paul: And it works. It’s good.

Rich: It’s good. It works. I think the question that sort of he’s alluding to, I think, if I can reframe his question, and I don’t know Rich, and I might get this wrong, but I want to reframe it because I think, I think fundamentally what he’s asking is, “Can I be a liaison between what an organization needs—” in his case, an NGO.

Paul: Yep.

Rich: “—and how to best use these tools that are, like, sort of piling up in the world.” No one’s having that conversation right now. The idea of someone that is taking the actual real-world context and real-world needs, whether it be an NGO or a business or whatever, and then bringing it over to these incredible tools that accelerate not just development of software, but development of anything, is really, really interesting.

Paul: Well, let’s go to a, I’ll throw out a hypothetical so we can talk about something real. Right? So here’s this person, and their job, nominally, they’ve been brought in on the team to figure out design. Okay? And so let’s come up with the project. The project I’ll throw out is: It’s an organization. It’s headquartered in New York City.

Rich: Mmm hmm. Mmm hmm.

Paul: We know it well. Got some money from Bloomberg. And the whole goal is to increase awareness of childhood vaccination. Like, let’s get more kids vaccinated.

Rich: Okay, okay.

Paul: Okay? And your kids get vaccinated at school. We wanna send out pamphlets, and so on. So we do all this stuff. So let’s say that what I wanna do here is I wanna create a system that works with doctors so that they can report back on vaccination rates.

Rich: Mmm hmm. Mmm hmm.

Paul: And I wanna track that information.

Rich: Yeah.

Paul: And sort of track how communities are either saying, “We love vaccines,” or, “We don’t want vaccines anymore,” because that’s going on.

Rich: Yeah. Okay.

Paul: Okay, so I need to build a system for doctors and maybe nurses and schools, and I got all this stuff going on, Rich. You’re a UX researcher, and now here we are.

Rich: Yeah.

Paul: Okay, we’re in the room. We’re both shirtless, oddly. [laughing]

Rich: So, I mean, look, here’s the best way to sort of highlight the gap. You can’t take that ask and run to the engineers that are messing around with AI, and skip everything else. You can’t run into the engineering room and just say, “Hey, guys, listen: Vaccinations. Really important. I want to raise awareness, and I want to have an idea of how to—” And they will just look at you, and like, “Well, first off, how’d you get in here?”

Paul: I did a thing on purpose, which is I specified the problem about as well as they often walk in the door.

Rich: Yeah!

Paul: Which is, I literally don’t know what I just asked for.

Rich: Exactly.

Paul: Yeah.

Rich: And so what happens is, when organizations that don’t have any interest in the nuances and the complexities of how tools get built, have needs, there are a host of, let’s call them translator/liaisons that take what the need is and then turn it into finer and finer-detail artifacts that then can be used to build something.

Paul: All right, so let me refine my ask, and then you start to do that.

Rich: Okay.

Paul: So my ask is, I actually send out 20 million pamphlets a year.

Rich: Wow!

Paul: It’s a lot of pamphlets.

Rich: Yeah.

Paul: Yeah. It’s expensive, et cetera, et cetera. And I actually don’t know how those pamphlets—we poll a little bit, but I have never done, like, a broad analysis of how those pamphlets are performing.

Rich: Yeah.

Paul: And I need to validate, I need to go back and get more funding. And they’re usually pretty generous—

Rich: Yeah.

Paul: But they said to me last year, like, “We need to know what these pamphlets are doing.” You can’t—and they’re in 12 languages, and so on and so forth.

Rich: Yeah, yeah.

Paul: So I need to build a system that pulls all the people who, like, I need to figure out—

Rich: I have a CRM that has all the addresses in it.

Paul: I have all the addresses. I know how to reach these people. I have a lot of email addresses. I know some of them are doctors, some of them are school nurses, so on and so forth.

Rich: Yeah. Yep.

Paul: And I need to pull them, gather that data, analyze it, and then report it back.

Rich: Yes. Yes.

Paul: So this is a fundamental problem, or otherwise they won’t give me $5 million next year.

Rich: Right.

Paul: Okay?

Rich: So here’s what’s interesting about your problem that you’re stating here. Everyone is talking about what AI is going to do to your job.

Paul: Yeah.

Rich: You want to excel in this environment? Go and train yourself to be able to answer that problem, and be able to have a dialogue with the stakeholder that’s asking you to help with that problem. Because AI can’t do it, because it’s human, it’s extremely context-sensitive.

Paul: Yes.

Rich: And look, will it do it one day? Will there be a business-requirements version of Siri? I don’t know. I don’t know anything.

Paul: You can get—you could present this and say, “Outline how the project should go.”

Rich: Yes.

Paul: And ChatGPT will do, like, an okay job. But here’s what you have to educate yourself on. Okay? When we talk about a lot of this stuff because of the cost of engineering being so high, that is like, it’s like Jupiter in the solar system of process. Right? Like, whatever happens, we need to orient it around the fact that engineering is really expensive and takes a long time.

Rich: And we don’t want to get it wrong.

Paul: We don’t want to get it wrong. So a process emerges. We’re going to start with, we’re going to start with basic requirements analysis. We’re going to do all the meetings, then we’re going to do UX research, we’re going to do more analysis, then we’re going to present. Then we’ll do wireframes.

Rich: Yeah.

Paul: Then we’ll present the wireframes to the users, and so on and so forth.

Rich: Here’s the other bit around this. There’s a key political component around it. You create these stages of artifacts.

Paul: Yeah.

Rich: And then you go back to the stakeholders so they can sign off.

Paul: Right.

Rich: Because what you don’t want to do is have a quick five-minute conversation in a hallway and then disappear and spend $3 million and end up 2,000 miles west when you should have been 6,000 miles south.

Paul: Now here is the superpower that AI gives you, and I’m going to tell you.

Rich: Yeah.

Paul: Generative AI, large language models, are good at generating artifacts. They translate one kind of statement to another kind of statement. A use case could become a module, or you know, things like that. So what you are actually beholden to do, if you want to help these organizations, what I think one of the most useful things you could be learning is how far you can go in creating artifacts that represent the entire process, from business analysis to research to code generation. You may not be doing all of it, but you should educate yourself as to what these tools can do to accelerate that.

Rich: Yeah.

Paul: Because I actually don’t think those artifacts are going away. I don’t think anyone is ever going to sit down and go, “Make me this.”

Rich: Right. It’s too far of a leap.

Paul: Because, because it’s like, what are you going to do? You’re going to say, “Make me a thing that sends out polls to a bunch of nurses.” And so—it’s like, you’re just so far down the rabbit hole that you’re gonna have to start over. Like it’s just—

Rich: Yeah.

Paul: It’s not—

Rich: Yeah, exactly.

Paul: It’s not a real question. So instead, you know it’s gonna continue to be a 50-step process to get that thing out the door. Right?

Rich: Yeah.

Paul: What you can do that you could never do before is get good at understanding those steps and actually do some of the work related to those steps.

Rich: This is the new profession that is coming.

Paul: Like, to sort of show our own hand here, what we’re learning as we’re playing with this stuff, Rich and I used to run an agency, and we’re learning that the real power in this framework isn’t that it can make code fast, it’s that it can simulate the whole process. You can say, “Hey, write me a product requirements document.”

Rich: Yeah.

Paul: And it’s not a great one. It’s not—it doesn’t have, like, intelligence that it needs.

Rich: No, no, it’s pretty generic, often.

Paul: But boy, can you—if you are experienced at creating PRDs, you can see the gap really quickly, and you can be like, “All right, let’s close that up and go on to the next phase.”

Rich: You’d ask AI to keep filling in the gaps.

Paul: You can, you can say—and you can then uh, translate that.

Rich: “Beef up the security requirements.”

Paul: Right. And now, let’s translate it to a specification. Well, now let’s turn to each module—

Rich: Are you going back to AI when you say, “Let’s translate.”

Paul: Yes, yes.

Rich: Okay, so you’re highlighting something here implicitly, though, I think it’s worth saying out loud, which is if you jump to the end of the movie, AI doesn’t do a good job. But if you’re willing to step through the artifacts and feed them back into the brain, it does a much better job.

Paul: Okay, so if you’re a UX researcher, I think you have, what you want to do is know where you are in the process. Okay? UX research is one step in from basic, like, we’ve started the project, we need to figure some stuff out—

Rich: I’ve got some wireframes to show or whatever. Yeah.

Paul: So now as you can go up one level and you can write the overall definition document or the vision statement.

Rich: Yeah.

Paul: And you could do that yourself, or you could do it with the help of AI saying like, “Hey, what’s this project going to be?”

Rich: Yeah.

Paul: So that could, that’s one way to go into your career. Or you could continue all the way until you have it writing some components for you.

Rich: Yeah.

Paul: And see if you can get a prototype built.

Rich: Yeah.

Paul: You know what, we’re being very abstract—

Rich: No—

Paul: —but you just stated really, really clearly, which is, simulate the whole process. Like, simulate each artifact that gets generated.

Rich: Yeah.

Paul: Don’t just think you’re going to get to the end state.

Rich: To punctuate that, and I think this is a kind of hopeful observation about AI, because it sounds like everything’s getting automated away and we never have to look at each other ever again, it’s actually a very human-driven role.

Paul: Yes.

Rich: It’s much more about, “Listen, these tools are crazy. I’m going to step you through how to tease value out of them, and it’s going to take a lot less time. But we do need to talk, so I can manage this.”

Paul: So this is the value that you can drive—this is a person who wants to work with not-for-profits.

Rich: Right.

Paul: Okay? So the value that he can—if you get that meeting, like, let’s say a couple years from now or a year from now, things have kind of evolved a little bit.

Rich: Yeah.

Paul: You sit down and they say, “Hey, I really want to do this.”

Rich: Yeah.

Paul: You can go, “Let me show you exactly how it will work.” And how it will work is incredibly high-definition.

Rich: Yeah.

Paul: Like, it’s, it’s almost like we’ve upgraded wireframes to the whole level of prototyping.

Rich: Yeah.

Paul: Right? So where you used to go to a whiteboard and sketch, you can show them the whole solution.

Rich: Yeah.

Paul: The flip side of all this, the thing that’s key, and we’re seeing this in our product. Like, we’re going to be able to deliver services to people. We’re going to be able to get them a product in their hand and deliver a software experience for something like 10% of what we used to charge.

Rich: Well…

Paul: All right, maybe, maybe a—

Rich: 20, Paul.

Paul: We need margin. [laughter] Like, for real, we do, because we have to keep developing the product. But, like, as—

Rich: I’m kidding. 10% sounds great.

Paul: No, but we’re talking to people.

Rich: No, no, it’s…yeah.

Paul: And they’re leaned in and they’re like, “Let’s go.” But the numbers are low compared to where they used to be.

Rich: Look, this is a very murky podcast episode?

Paul: But this is huge for NGOs.

Rich: It’s huge for NGOs. It’s huge for anybody that’s been stuck in the same lousy tools they’ve been using, because they can’t afford or don’t have the time to try anything new. Here’s the thing. It’s going to be a corny title, but a new role and title is coming. It might be three years away. I don’t know when it’ll show up. But there is going to be a new sort of recycling of existing skills to fit them in this new way of working.

Paul: Product specialist…?

Rich: I don’t know, I don’t know what it looks like. What I do know is a lot of the sort of ground rules have been tossed aside because of this new stuff. And no one knows it yet. Like, no one is really playing with it yet, because it seems like it’s, the engineers are having all the fun right now, but it’s coming.

Paul: Coders are taking advantage of it and getting more done.

Rich: Yeah.

Paul: Product managers, consultants are doing it on the down low and not really telling anybody.

Rich: Product managers, to this day—and I’ve been a product manager for almost 25 years—to this day, have to explain what they do in a long paragraph at any cocktail party.

Paul: Every Thanksgiving.

Rich: Every time.

Paul: Yeah.

Rich: To this day. And I think now, I think the word “AI” has to be in the title somehow. I don’t know what’s—

Paul: Well, you know what it is—

Rich: I don’t know if it’s AI specialist, or AI—

Paul: No, it’s going to be horrible to be AI Process Reengineering.

Rich: Oh, boy, oh, boy, oh, boy, oh boy.

Paul: Yeah. That’s where we’re headed.

Rich: Let’s see where we’re at. We should revisit this episode.

Paul: Okay.

Rich: It’s November 2024. When the world changes, I’m curious what that title looks like. But—

Paul: So let’s give this nice person concrete advice, which is go in and be, pretend to be an entire software organization. Take on each role and talk to the bot.

Rich: Yes.

Paul: Okay?

Rich: Yes. Talk to the bot. Produce stuff. And…

Paul: Solve a problem for an NGO with AI.

Rich: You’re a non-engineer who doesn’t fully realize just yet how much power you actually have now in this new world, because no one knows it yet.

Paul: You’re also still going to get trapped by bugs. Like, that’s the—you still actually need to be an engineer to deliver with this stuff.

Rich: We still need engineers.

Paul: Yeah.

Rich: We’ll do other podcasts to talk about how engineering needs to rethink how they work, because we still need engineers. One final point I want to make here, which is all of these steps and all of this process and all of this ceremony before code gets written? You alluded to it. It’s because getting it wrong is so expensive. You could spend a million dollars and turns out it was the wrong tool.

Paul: NGOs in particular do that a lot.

Rich: They do. But there’s a double hit here, which is everything we just described, but also the fact that if you decide to kind of barrel through and get it wrong, it’s not expensive anymore.

Paul: Yeah.

Rich: So all of the reasons you had for creating all those speed bumps are kind of gone, in a way, because you can kind of whip it together and see. And if it works, we keep going. If it doesn’t? Eh, we tried.

Paul: There’s a local way that this person could adapt their job to learn more. And one would be to partner with an engineer and then prototype experiences, as opposed to going with designs and design ideas and doing the research. You could actually put software in front of people, get them to play with it, and it would take about as long as the current UX research process does.

Rich: The process of creating things is being exploded.

Paul: Yes, that’s right.

Rich: That’s bad English, I think.

Paul: No, it’s okay.

Rich: Okay, thank you.

Paul: It’s exploding. You could just—

Rich: Being exploded.

Paul: Yeah, being exploded.

Rich: Yeah. So it’s an exciting time. I say, what you’re essentially saying also, by the way, is get out of your lane and go look around, because no one knows what’s going on. [laughing]

Paul: Yeah, don’t worry, you’re not behind.

Rich: You’re not behind.

Paul: At all.

Rich: Exactly. Exactly.

Paul: All right, friends.

Rich: Check us out at Aboard.com. We are a—

Paul: Go up to the top right of that site and click the “Try it” button.

Rich: Try it. You can actually run AI software, our AI tool, to get a glimpse of the software we can create for you in days, not months. And it’s really cool. And we’d love to talk. Check us out: aboard.com.

Paul: Hello@aboard.com. Thanks for that question, Rich.

Rich: Thank you, Rich.

Paul: Other Rich.

Rich: Other Rich.

Paul: Bye.

[outro music]