360: How do you nurture junior developers in an AI world? (With Barun Singh of Andela)

Speaker A: You're listening to what the Dev, the weekly podcast of ST Times. And now here's Dave Rubenstein, editor in chief of ST Times.

Speaker B: Hello everyone. Today we're going to look at how organizations are, uh, going to nurture future developers if artificial intelligence eliminates the need for entry level jobs. Uh, it's certainly something a lot of people are talking about, the fear of AI taking everyone's jobs. Uh, here to talk about it with me today is Barun Singh. He's the chief Product and Technology Officer at Andela, which is a tech, uh, talent marketplace. Uh, welcome Barun, and thanks for being here.

Speaker A: Yeah, thanks for having me.

Speaker B: Yeah. Let me start off with an obvious question. How do organizations develop engineering talent when the first rungs of the ladder are no longer there?

Speaker A: Yeah, I think that in a lot of ways it's no different than it's always been in that, uh, people learn by doing and it's just that the things that they need to learn are slightly different. Now it used to be the case that, uh, you would have to learn the special almost tactics of writing code. You still have to kind of know the foundational layers of basic data structures and all that, but the AI can usually do that for you. The tricky thing is that it used to be the case that you couldn't progress on to understanding the abstractions and the data modeling very, very well unless you knew how to do the basics. And now you can kind of fake your way through it. And so now it's up to these organizations to really understand how to distill within each of their employees the core knowledge that they need to have that they are not just relying on the AI. So what I mean by that is if you're putting code out there for, um, merging into a code base, do you really understand what's written in there? And the code review process is I think, actually a really, really significant part of that. The planning process, owning a feature end to end, these are all things you kind of always had to do. But now I think people are hyper excited about the, um, first part being so fast that they tend to skip over it sometimes and they really can't afford to.

Speaker B: Yeah, well, you know, if we're getting to a world where people don't need to actually generate code anymore because AI is doing it for us. What these entry level developers would be dealing with are, uh, just the things you're talking about, right, which is code review and making sure pull requests are accurate. Uh, is that level of worker also involved in things like making sure it's secure and the governance is in the right place and access from agents is going in the right place or is that more of a higher level function? Like where, where is the baseline, where are the entry level developers, where are they starting? At what level?

Speaker A: So I would, I would question the premise a little bit in that I actually don't think that it's a base level thing to do a review of a pull request. For example. Um, you may not be needing to review some, you know, the specific lines of code review in particular, code patterns may not be as important. But have you thought about the problem in the right way to model the objects in the right way in this code that you've presented forward? To do that? Well, you actually have to be quite senior. Those are the sort of things that you develop with experience, um, over time. And same for security things like having an eye for what security things are that you need to be looking out for. It turns out that if, I don't know if you've tried out codec security or cloud security, but they are quite good at, um, detecting security issues that previously you had to be quite senior for. So this question of what requires seniority and what can a junior person just be thrown at, I think that's shifting a bit. And there is a senior layer to all of it. There's a senior layer to how you write the code, there's a senior layer to how you, um, assess security vulnerabilities. And I think of the role of the junior developer as almost more of an apprenticeship in all of it. You are doing all of the things, you are doing them alongside people who can mentor you within an organization. And you're learning how to create judgment, how to evaluate where you should prioritize things. Um, and so I wouldn't so much put the junior folks in particular buckets of work. I would kind of throw them into the same work. It's just you get better with it the more you do it.

Speaker B: Sure. Almost like uh, an electrician or a plumber would apprentice next to an expert and learn the trade. And, and then they exactly like, you

Speaker A: don't just say like, oh, I've got my first electrician, let me throw them to stripping the wires. Like that's not how you apprentice to be an electrician.

Speaker B: Right, gotcha. So our colleges and universities now, are they even still relevant? I mean, are they changing their curricula to address this, uh, new era of SOFTW development?

Speaker A: Um, I'm not sure actually what they are doing. I can tell you what I believe is important. For them to do. We have calculators. So is it important to still learn math? Right, right. We have computers to do all the things. Um, we have had that. But people still need to understand the basics for no other reason than somebody needs to create those machines and the algorithms, the knowledge needs to not get lost in society. But also knowing how a tool works for fundamentally at its core, helps you better use it. Um, if you're only ever operating with a certain superficial understanding of the things you're using, then you're never going to be very, very good. And that is true whether you're a software engineer. It's true if you're a chef. The core chefs will spend forever learning how to do the most basic principles of cooking. Um, but even though you've got all sorts of machines that can do things for you, um, so I think it's the same for any form of labor, actually any form of creative activity that we engage in as human beings. Mhm.

Speaker B: Interesting. So, uh, what about boot camps? Is that still something that is relevant in this age?

Speaker A: Yeah, absolutely. Um, I actually think that the core thing about boot camps has been that people learn well when they are immersed in the work that they need to do. And that thing of learning by doing is as powerful as it has ever been. And what a boot camp gives you is sort of like a structured program to do that learning. And that I think is still very, very useful. How you create those boot camps and what you teach, all of that may be changing quite drastically. Right. Like you can use the AI to help you create curricula. You can maybe make that curricula more dynamic more quickly. Um, the curricula itself needs to engage those AI tools. But that mode of learning I don't think has necessarily changed.

Speaker B: So, uh, as we introduced you earlier as, uh, you know, being the chief product and technology officer at a tech talent marketplace, are organizations looking to recruit new talent? Is it more of a retraining of the talent they already have? What are you seeing out there in the marketplace?

Speaker A: Yeah, there, there's a bit of both. Um, there definitely is a very strong interest in retraining talent. They already have, um, the executives at almost any company, they are in many ways in the same place as everybody else is, or as most people are, that, uh, they are seeing here, they're hearing the news, they're seeing this stuff on social media about how amazing these technologies are and how much they can do, and then they're looking at their own organizations. And by and large they're not seeing a 3x transformation overnight. Right. And they're trying to figure out how to align these two pieces of information of what they see on the ground and what they're hearing from media and, and reading about and seeing largely from companies that are trying to sell them tokens or sell them services. Um, and so they're very interested in understanding, like, how can I get my teams to be more effective in using these tools? And so there's a lot of, um, it's retraining. It's also they're not totally sure all the time what they want to train people to do. So there's a lot of trying to understand what this new space looks like, um, and how you should be building. And so that's actually a large place where we are engaging with organizations now is we help train talent and level them up to use some of these tools. But we're also helping people understand what to expect and what sort of training is most effective. And so, um, on.

Speaker B: That's great. I mean, I'm wondering if organizations are kind of getting ahead of their talent with the rush to adopt AI and try to figure things out. And perhaps that might be part of the reason that they're not seeing the results that they thought that they would see is because their employees aren't up to speed enough to handle these issues that are popping up as they race ahead, uh, implementing AI. What are your thoughts on that?

Speaker A: Uh, yes, absolutely. There's no doubt we are in a hype cycle. And when you are in a hype cycle, things just move very, very fast. Faster than you're able to check the facts and faster than you're able to establish your own under perspective on things. And you saw this, for example, three, four months ago, it became a very popular thing to talk about in certain circles of technology. Um, companies that you should be token maxing, you should be using as many tokens as you want, which is always, never made sense, honestly, as a metric at all. Um, but it was a rush to say, well, how do we incentivize our employees to make use of this new technology so that they can hopefully come of leverage it and we can be on top. Now you're seeing people do a little bit of the opposite and say, oh, this actually costs a lot of money and we aren't seeing benefits. Maybe we should try to understand what it is that we are asking people to do and like have a plan, a strategy of, you know, what is, what is this technology actually most valuable for and how should we leverage it? So I think you're seeing a little bit of backing down from the kind of extreme excitement that was maybe not fueled by, by a clear understanding to a level where people are trying to gain that understanding. I do think that there is a humongous productivity boost that is available if you use these tools properly. But if you don't know how to use them, it's like any tool, it can be very, very powerful. It can make you do the wrong thing much more quickly if um, you don't know how to use it. Or it can help you do the right thing more effectively if you know how to leverage it. But that requires discernment judgment, um, general awareness of the ecosystem. And so that part takes a little bit of time to develop. And the challenging thing is that when you're operating in an ecosystem that is also itself changing every week, then that is not a one time thing. You have to be constantly reassessing that.

Speaker B: Yeah, something that we've seen every time there's been some sort of a major advance uh, in the technology, whether it was going to Agile or DevOps or the Cloud or any of these things that got ultra hyped. The first movers are always the companies that are the large companies, the ones that can afford to uh, max out their tokens to get people incentivized to use. What are you seeing in smaller companies where they may not, they really have to be careful with their spending because if they make a wrong move or two that could really hurt their bottom line and perhaps even threaten their organization, its existence.

Speaker A: It's interesting because it's in some ways you, it's not exactly the size of the company, it's kind of how tech forward they are or how senior their people are. So there are some very small companies that are actually the most significant users of the technology. Like teams of 10 people who are leveraging AI for 95% of the code that they're writing and moving at uh, a much faster, much higher velocity than they were six months ago. We've got an internal team that's operating that way. Um, and a lot of tech startups are working that way. And sometimes it's. Then you have the very large companies where they are, the tech companies, uh, the metas, the um, Googles, Amazons and there it operates one way. But the very large enterprise companies that are not tech companies at their core, it's a very different thing in those places. Right. Um, and there you have governance issues, you have a lot of tech companies like they don't even have a procurement process that allows Them to access the tools, the same tools that a smaller company might have access to. And so I wouldn't say it's size necessarily. I would say it's the ability of the organization to adapt and to experiment and to kind of focus on the outcome of what they're trying to build rather than, um, focusing on creating um, requirements for the people in the organization to meet certain, um, benchmarks of usage.

Speaker B: Right, yeah. So how would you recommend a future junior developer to break into this world? What do they need to know now that perhaps they didn't need to know 10 or 15 years ago?

Speaker A: Look, the technologies, the tools, how you build things, they're all going to change really dramatically. The core foundations are not different and they are very important to learn. You have to learn how to fundamentally software engineering. A core part is understanding how to model the world and how to think in abstractions and ultimately any form of knowledge work. There's really only one meta skill and that is a skill of how to learn. And so the process of learning, if you just focus on constantly learning and improving, then you will be placed, you will be situated in a way that you'll be able to adapt to whatever the new technology is. Right. You'll be able to figure out how to best use it. The more you are trying to follow along with a specific fixed playbook of here is how you get ahead in X, probably the less likely you are to get ahead in X. The future belongs to the people who can develop their own core understanding of things. And much of that will come just by playing around and experimenting and just really, if you enjoy the thing, if you're having fun with it and you're putting the time into it and you're trying to deeply understand it, you will learn the things you need to learn.

Speaker B: Last question. I'm curious to know. Uh, organizations for a long time have, uh, outsourced looking for the appropriate talent that they need for specific jobs. Uh, is that still a thing with AI? I know that they were talking about regions of the world that were hot. It was Ireland for a while, uh, Southeast Asia, uh, all these places around the world. Our organization still like mining those areas, looking for that kind of talent for AI.

Speaker A: Um, I think that people are looking for the right talent wherever they can find it. We have a core belief at Endela that we think talent is globally distributed even if opportunities are not. And I think that that will come back around. With these new technologies, it becomes even easier for somebody to build anywhere in the world. I would argue. Um, and I think right now, honestly, a lot of organizations are still trying to figure out what they should do. There's a lot of uncertainty and upheaval. And so I don't know that there is one fixed trend in that regard. I think we've got to give it a little bit of time to see how it plays out. Um, but I don't know that there's a massive shift in how people are seeing remote work versus non remote work or global talent versus domestic talent, um, as a result of this quite yet.

Speaker B: Interesting. All right. Barun Singh, chief product and technology officer at Andela, thank you so much for your time today. Excellent conversation.

Speaker A: Thank you.

Speaker B: Okay. And to our listeners, thanks, as always, for tuning in. Until next time, I'm, um, Dave Rubenstein, editor in chief. So long. For now.

Speaker A: Hm.