Episode 8: If AI Does The Junior Work, How Do Juniors Become Seniors?

Brian Milner (00:23) Welcome back in everyone. This is the People Over Prompts Podcast. I'm here with you again, Brian Milner. This is the podcast where we talk about the future of teamwork in the age of AI. And appreciate you bearing with us as we missed last week. I was doing some traveling and just was n not possible to get the the show done in time and didn't want to get something out to you that was not of good quality. So we we we delayed last week into this week and there's there's a topic that I've been looking forward to talking to you about for this week. it's it's one that I've heard a lot of conversation about. And I'm sure you've heard conversation about this as well, because one of the things that we've heard quite frequently about AI is that it is doing all of the the work that junior employees used to do, junior developers, you know, that sort of thing. So I I I I wanted to ask this important question. If if AI is doing the junior work, how do our junior employees become senior employees? And I I think that is a maybe developing crisis, if you want to think of it that way. It is an impending crisis. it is something that I think will happen at some point that we we will have this this gap. And I I I've I've fished around and found some quotes from some different scholarly articles about this that I thought were were relevant and and interesting. So we'll we'll guide our way through this. Bef before we dive in though, I just wanna make one more reminder here. I'll make reminders periodically about this until it happens. It's It's not happening until November, but I I'm reminding you of it now because you might want to make travel plans or or something else to attend. But I am going to be delivering a keynote talk at the regional scrum gathering that is taking place in Central America. That's that's going to be in Guatemala at the Porta Hotel Antigua. that's November 4th through 6th. And if you are interested in that sort of thing, if that's somehow possible for you to get there, it's a beautiful hotel. It's a resort kind of hotel and I'd love to see you there. if you mentioned that you you heard this on the podcast, hey, I'll give you my drink tickets, okay? so hope to to see some of you there. but getting back to our topic here, you know, I I was really curious about this idea and about if we're we're automating ourselves out of kind of what we would think of as apprenticeship. You know, we we don't really have a traditional apprenticeship. programs like we used to, but you know, are we are we getting rid of that apprenticeship process of of having someone who who comes in that needs experience but but gains it by doing some of the more routine things. the first quote I found about this, this is from a McKenzie quarterly report from back in July. Brian Hancock and and Charlotte Sealer. It says, these are precisely the activities, and they're talking about some of the junior kind of activities, some of the mundane routine activities. These are precisely the activities, though, which young employees have traditionally built instincts, developed judgment, and earned the right to take on more. And they're talking about the fact that AI is doing these things now. AI is more and more taking on those routine tasks. And those tasks were exactly the kind of task that junior employees were were really making use of in order to learn the systems, learn how to do things properly, and you know, kind of grow in their job. kind of a slight counterpoint to this, this was from Hannah Calhoun. She's the vice president of AI at Indeed. And she said by eliminating junior roles, leaders are effectively cutting off their future talent. pipeline. So she's trying to say here that there there is an an impending crisis there. That the the pipeline of you know if we we've if we fire all our junior developers and all we have is senior developers, what happens when the senior developers retire? Right? Or what happens if a senior developer leaves? We we don't have anyone that has been trained up. We don't have anyone who's been learning underneath them to to take the reins if they step away. So I think there's a good thing that AI removes kind of the more routine work, but that routine work is often how people learned. And, you know, I I I I will insert into that conversation as well. It's not just how they learned. I would say the routine work is also how some of us had mental breaks from the more difficult stuff. We talked about that a little bit the last episode, just this idea that. You know, if if all we're doing are the the heavy lifting things, if all we're doing are the really mentally complex task, it could be draining, right? It could be something that really drains us and we may not be able to get to that point where we we're able to, you know, kind of learn or take the break from the mentally tough stuff with something more routine and then go back to the the the heavier stuff. So The question there is, are are we automating away that path? Are we automating away the path to expertise? that's why I kind of use that word apprenticeship. You know, are we are we automating that out of how we do things within organizations? or put another way, are we killing the talent pipeline in our organizations? another quote from the same McKenzie kind of journal. The the disappearance of entry level roles is gradual and easy to underestimate. And I I would agree with that. It's it's kinda like that thing that they you know, people have often referenced where, you know, you you drop a frog into boiling water and they jump right out, but if you drop it into you know, t tepid water and then you slowly turn the temperature up, they'll b they'll boil. That's a terrible analogy. no frogs were harmed in the the making of this podcast, I promise you. but but the the the analogy works in this scenario because it it is sort of a gradual turning up of the heat here in our organizations and it's it's easy for us to miss that as we automate away these more entry level jobs. this is a a quote from a group of people that did that had a you know another McKinsey report back in May, and they said, but the cumulative effort will be substantive. fewer pathway pathways into organizations, weaker development at the base, and gaps in leadership capability. And I I think they're hitting on three really big important things there. Right? Fewer pathways in the organization, how do I get my foot in the door? you know, we we've already changed from a culture from from a place where people would work for, you know, twenty years at a place. that's a rare thing now. But how how can I get my foot in the door if I am a recent college grad. Or if I've only had a few places I've had experience at. there's not that entry-level job that I can do some grunt work and and show that I'm a hard worker and and kind of work my way up the the the ladder. I think that weaker development at the base is is an important thing to think about as well, because i it's it's not having the human oversight element as much Since we don't have that junior person there overseeing judgment of each step. so it could make for a a weaker development at the base of what we build. And gaps in leadership capability, absolutely. you know, projected out even the even forward, right? If if the senior people move into management, like a lot of times they do, well, you don't have senior people behind them. Well, those senior people move into management and then move on. You don't have the the complete pipeline. You don't have even people ready to move into management anymore because you didn't have the junior developers to move to senior or senior to management. It it starts to become a very serious gap. and you know the I I think it's important to recognize that junior jobs are actually about more than just you know, grunt work. I I think there there is a a purpose to junior junior work in organizations that goes beyond just carrying out the routine things. I think the purpose there is to to learn. It's part of training. It's kind of extended training. So here's the the thing that I've seen people start to argue about this is should the juniors kind of skip the beginner work? I've seen posts about how companies are are now looking for different kinds of skills in juniors and and they're really kind of more advanced skills. They're not the kind of skills that you would see from an entry-level position. They're in they're skills you'd see from a more mid-tier to upper tier position. so this is a quote from a a piece called Rethinking Early Career Talent in the Agentic organization from McKenzie. And it says entry level wor roles should be defined by developmental task combinations. Not by whatever is left after automation. And I think this is an important quote because it's hitting on some of the things we talk about here in the podcast, right? this idea of the fact that, you know, just giving the rest to people after we've done the automation, right? Just saying we're gonna try to do as much as we can and then whatever's left over, it's sort of the un unintentional design into it. There's not really a design there, right? It's it's just kind of leftovers. this is a quote from Harvard Business Review that I I I think this article is called Help Employees Get Better, not just Faster with AI. And it says, as AI makes generating polished work easier, the scare skill is becoming judgment. Knowing what to trust, question, and refine. And I think that's the important part here as well about junior work is like it says, knowing what to trust, question and refine. That's that's the judgment that's develops from doing some of the more routine work. And if you're not able to do that routine work, then you're not able to develop that skill to to know what to trust and know who to trust and and what to question and what to to not question and and what what needs more refinement, what needs more human oversight. That's just not going to be developed. So maybe AI should be eliminating that pointless beginner work. Maybe that's true. But Sophisticated output is i isn't necessarily sophisticated judgment. And I think that's an important point here, right? we have to think about which struggles are obsolete and and which struggles are developmental. those things are going to be missed if we turn everything over to AI. which which kind of leads us into i is is AI actually hurting younger workers? you know, it's it's definitely hurting. Lots of different segments of the the worker base out there for various reasons, but what about just younger workers in general? Well, this is from a a Stanford digital economy lab kind of a piece that was put out called Canaries in the Coal Mine. And it says early career workers in in the most AI exposed occupations. Have experienced a 16% relative decline in employment. So it's junior early career workers that are are being hit really hard here. we're not hiring them because this this is pretty mundane, routine stuff. you know, we don't have to pay that smaller salary. We can instead spend that on our AI infrastructure and have that take take you know, take its position, take its role. but in that same paper, Stanford, the the researchers warned gave a explicit warning here. They said, We do not believe that AI is always and everywhere the sole determinant of employment. So they're saying that, you know, even though we're trying to now look for junior developers who are more advanced with AI, they're not seeing that that is that sole determinant factor. That there are actually other more human skills that are are being sought out for these more junior positions. so the the question here is, is is AI the cause? The cause? Is it one cause? Or is it part of a much bigger kind of labor market shift that's going on, right? I think that's the thing we have to wrestle with. So let's talk about that first rung on the ladder. Right. where we can all, you know, if you've been in this been in profession for a while, you can think back to that first rung that you had on the ladder and what it looked like and you know, how you kind of paid your dues there and that. this is from a Mackenzie quarterly that was put out. It the quotes sa is is from a piece called Building Expertise in the Age of AI. And the quote says, The resulting incentive hire seniors, automate juniors Quietly dismantles the bottom talent pyramid on which every senior role depends. Just saying that as a senior, you need those juniors to get increasingly better over time. And you can trust them with more and more things. you like those capable junior people who are are growing and getting better because you can get more things off your plate and you can focus on the the more intricate. higher level things. That's what's being missed. That's what's being lost here. from from kind of that same paper it says those expecting AI to increase entry level hiring in t twenty twenty six outnumber those expecting it to decrease hiring by nearly three to one. Which is sort of a counterpoint there. That doesn't seem logical. Why would they expect it to entry level hiring you know th those who are expecting AI to to increase entry level hiring, it it's not really in increasing entry level hiring though that we saw in that the that last statistic a little bit earlier on. It's actually decreasing it. It's actually hitting it very, very hard. So as more junior jobs are kind of disappearing and going away from our companies, i it it's kind of interesting to to think about what what are those old junior jobs? What are the actual tasks that we're we're offloading and and and giving over to AI. it's it's an i you know it's a it's it's kind of one of those analysis things that have to be done. so let's talk a little about codified knowledge versus tacit knowledge here when we're talking about the knowledge of junior developers this is from another McKinsey quarterly called Building Expertise in the age of AI I've quoted this one before early in this this podcast But this quote says: Tasks such as research, documentation, data cleanup, basic coding, and preliminary analysis are being streamlined or absorbed into AI systems. I think we'd all probably agree with that. That's probably the kinds of things that we imagine that AI would do is feed those things into AI. AI, if it can do things with better quality, if it can do it faster, right, if it can offload some of that. more routine things, automate those things. That's the kind of stuff that we want AI to do. going on from the same same paper, they there's another quote that says: what remains for humans, including junior employees, is higher value. And I thought that was a kind of an important thing to highlight. What remains is higher value, interpreting outputs, making trade-offs, and connecting insights across context. So The point there is that the human work is actually shifting towards judgment. Where that trust boundary is, which I think is is kind of interesting, right? and one of the the interesting things I think about this is what this kind of leads to as far as our our leadership is concerned. you know, if if We're we're kind of looking at our process for onboarding lower level employees, people who are new into our profession. you know, I think we have to ask the question if we were gonna gonna design it today from scratch, knowing what we know about AI, would we design it in the same way that we have it today? and and honestly what I what I wanted to get to here in the podcast was kind of a a a leadership crisis that I think is developing as far as these junior developer positions are concerned. In the past, managers, lower level managers, mid-level managers could rely on that more routine work as being the the root the regular building block of skill and knowledge. Because if if I can have those junior entry-level employees Do the things that are more routine, they'll learn the system, they'll get better. I I'll have a pool that I can promote and and everyone can move up the ladder, right? So leadership in that realm, I think, kind of mostly relied on that paying of dues to be the dojo, right? To to be the the learning environment, the school, and the trainer really, right? Because they they're they're they're allowing the routine nature of the stuff that's being done to reinforce again and again and build up the capability within those employees. Well, if AI is taking those things and taking that away. There's still a need, right? There's still a need for those more junior to gain skill, to get better, to improve, and move up the ladder so they can fill the higher level senior roles. I think what this means, and this is kind of my idea, I'm proposing that here to you in the in the podcast. I think that what that means is that those mid-level and lower level managers are gonna have to be much more deliberate about. Mentoring. I think mentoring is going to become a skill that you know I I think in the past we've kind of looked at mentoring and given sort of a a sideways nod to it. it it's not been a a primary focus. You know, you'll ask someone in interview, give me an example of someone you've mentored and they'll tell you a story about somebody who has, you know, gotten better over time and the Problem is they don't really describe a system of mentoring. And I think that's the era that we're going to enter into with those lower and mid-level management positions, is really establishing a mentorship system. having to be deliberate about meeting with the entry-level developers is going to take more time from those managers, right? It's going to take more time to kind of assist to guide those more junior developers and and help them get to a higher level. It's not going to happen by accident. And it's not going to happen by routine because we're taking the routine out of it. Right? We're stripping out that routine of of doing the mundane work. We're offloading that to AI now. So we still need people to go through the pipeline. We still need more junior developers who can eventually fill our senior level roles, right? How is that going to happen? And I I think that's where I'm kind of interested to see what happens in the next few years, as far as that's concerned. It's one of those gaps, right? And I I think the important thing to recognize here is there's sort of two approaches to this. There's two ideas, two camps. I think there's one that's much smaller that I'm advocating for here that says, no, take a more deliberate approach to it. Right? Map out what your your mentoring process should look like. And then follow that process no matter what the who the employee is. Right? you know, adjust for the person, obviously, but you know, have a system. Have a system approach to how you mentor someone. I think the secondary approach. And I think this is what the majority are doing, unfortunately. Is I think that the most senior leaders are making a bet. And it's a big bet. I think they're betting that this gap that's going to develop, that AI is going to be able to cover that gap. Right? I I think what they're saying is. Yes, we recognize that we're not training the junior developers now, and we're not building that pipeline into our senior developers. We we recognize that we're not going to have people to backfill the senior developer roles eventually. But AI will get better. AI will prove improve over time. So, yeah, we're just gonna bet that it won't just take the junior developers' jobs, but it'll take the senior developer jobs. And if it takes the senior developer jobs, then I don't have a problem with not having a pipeline. Well, I think that's short-sighted. And I I think you're just kicking the can down the road because those senior developers turn into managers, directors, VPs. Right? They they turn into the higher-level positions. And If you don't have the contextual knowledge, if you haven't spent your career working your way up, then by the time you reach that senior level, you don't have the base of knowledge to make the judgment calls. Right? So, how far do you want to kick that can and how far do you want to make that bet? That's the question I'd have for senior leaders. Is if you're betting that Yeah, AI is gonna catch up in time to cover the the senior developers. Is it gonna catch up in time to cover the managers? Is it gonna cut catch up in time to cover your job? I I don't think that's the case. I I I do think AI will get better, I do think it'll improve over time, and I do think it'll increase in its capability. But I don't think it's going to reach the level of making human judgment calls Anytime in the i in the near future. I think that's that's maybe more or more years away. I I don't say never. I think that maybe it can eventually do that, but it's it doesn't have the context, it doesn't have the capability. It can get better at the reasoning for what it does now, but you still need a human to determine whether it's hallucinating. We haven't fully eliminated that. And if we want to have quality controls and make sure that we don't have an embarrassing situation, then we're gonna have to have human oversight. Well, if you're not building that pipeline, you're not gonna have anyone qualified to to have that human oversight. So This is a gap that I'm going to be watching over the next few years. I'm curious what your thoughts are on it. How would you cover this gap? How do you think this gap is going to covered? I I'd love to hear your thoughts. You can send me an email about this. I'd love to hear from from listeners here of our show. So send me an email, Brian at agilityvolved.com. Brian is with an I, not a Y. So Brian with an I at agilityvolved.com. Drop me a note and let me know what you think about this topic. Let me know what you think about maybe other topics that you want to hear me cover here on the show in the future. are there guests you want me to have on? I'm kind of interspersing guests with you know, bringing you some some of the latest industry trends here as well. And I I plan to kind of continue that mode as we move forward, but I I want to have interesting guests on that you want to hear. I don't just want to fill it with, you know, guests. I want to make sure that the guests are people you want. So drop me a note and let me know who you want me to have on. the other thing I'd ask for you is, you know, tell a friend about this. Maybe just just tell someone that you know that you you like this podcast and it's interesting. Maybe it's something they would be interested in as well. I think this area is going to grow, and I'm really curious to see what the science and data tells us over the next few years as we learn more about what happens when AI becomes a team member. Right. And of course, like and subscribe to this podcast and whatever your podcasting platform of choice is. That always helps us get found. it really does push us up the charts and helps other listeners find this this podcast as well. So I'd really appreciate that if you could do that. if you're just trying this out for the first time, you know take a shot. See what you know, see what you think about it. But if you've been listening to a couple of epis episodes of this, then you maybe that maybe it's time for you to hit that subscribe button and make sure that this doesn't Go away that you have that in your your inbox. just a reminder here as well, we don't have any commercial sponsors of our podcast. My company, Agility Evolved, is really the only sponsor, if you will, of this podcast. And we do training that's through the Scrum Alliance. We do advanced certified product owner classes, advanced certified Scrum Master classes. I'm trying to stick to just the advanced classes. at the moment, so I can focus on you know, getting to that deeper level with people. So if that's where you are in your career, I'd love to see you in a class and I'd love to hear that you heard about it here on the on the show. So we'll wrap things up for this week and I appreciate you joining us. We'll talk to you the next time on another episode of People Over Prompts Podcast. Have a good week, everyone.