Engineering Management and AI - Part 2/3 - Jobs

Speaker A: This is the Effective Engineering Manager podcast. In today's episode, Adam and Slava discuss the impact of AI on engineering jobs.

Speaker B: Welcome to the Effective Engineering Manager podcast.

Speaker A: Hello Slava. Today we are continuing the topic of AI and engineering management with the focus on the impact of AI and jobs. Um, why do you think that's an important topic for us to discuss today?

Speaker B: Good question. And uh, so the first, uh, episode we discussed uh, the value that AI brings and now we are talking about the jobs. And I think it's important because uh, first of all it's on everyone's mind. Second, um, we as software developers and hardware developers, um, we've heard a lot of statements that um, software jobs are going to be automated out tomorrow. And um, it's just uh, these two I think are worth covering. Uh, and um, maybe uh, providing our own view and outlook on what is going to be the impact on the jobs, um, globally and um, locally inside our own domain.

Speaker A: Yeah. And as we continue on with this theme and our focus here is to provide the guidance to engineering managers on how they can start to embrace the change and be ready for it. We talked about some of the dynamics in the last uh, episode or podcast and we want to empower engineering managers to be able to lean in, uh, as opposed to waiting for things to change, uh, and then sort of being behind the curve or expecting jobs to get just replaced. Uh, there's a lot more proactive guidance that we're trying to provide here.

Speaker B: That's right.

Speaker A: So let's uh, dive into it a little bit. Um, why don't we start off by why we believe jobs will be impacted and what types of jobs could be impacted by this revolution.

Speaker B: So, and I must say that these are our views and we are talking about the future. And um, we are providing uh, this outlook based on our own understanding and what's happening and uh, our learnings and all the good stuff. But also there are I think, patterns which are very easy to understand. Uh, given that what we call AI or um, more precisely large language models are about the language and uh, the information and how it's processed and what happens to it to create value. And uh, that's why I think personally I believe that the jobs that are going to be impacted first and they're going to be impact on a pretty serious scale. Uh, the knowledge worker jobs that are low level, um, low responsibility and accountability, uh, maybe not accountability but low on responsibility. The jobs which are easy to train on, but at the same time the jobs which are pretty high impact, if you Remove someone whose job who are holding the job today, they're going to be a lot of problems. So that's basically low level, low responsibility, easy to train, hard, high impact jobs. And high impact are important because there has to be a value created by those jobs because otherwise why would you automate uh, them out if you replace them? And there's zero value created. So that's my take on what is going to be impacted. And uh, what are your thoughts on this?

Speaker A: Yeah, I tend to agree. Um, what we hear a lot

Speaker B: uh,

Speaker A: online and in discussion circles and whatnot is that mundane or tedious type tasks are the first to get replaced in any industry. Um, whether it's in healthcare, in legal, uh, like paralegals, in um, secretarial type work, bookkeeping, things that uh, a machine can now do at the same capacity as a human being with the same level of fidelity. Um, and I think some of those are low level jobs but they're obviously still high impact because there's a need for the output. Uh, but it's just easy. It's the first layer of jobs that can be converted over using the new tooling and power of AI. I also think that in the circle of software engineering as the tooling gets better and code can be generated with better accuracy. So the generative component of AI, you're getting more uh, code, uh, not just pojos and boilerplate code but actual meaningful code that requires uh, algorithmic um, analysis and thought, um, and some um, deeper connectivity between modules of code. As that starts to improve you're going to see more and more of traditional software development engineering uh, also be converted over. Now we'll talk in a little bit about where those roles are going to go. We don't, I think you and I are both guiding that those jobs are necessarily not going to go away. Uh, they'll just be converted into other things um, as we continue to feed the technology. But I do tend to agree with that. Um, so as we kind of talk more about this, what jobs do you think will truly benefit uh from this? We talked about some of the high impact. But what jobs do you think will ultimately benefit in the long run?

Speaker B: Well, I think um, well first of all number one is we are talking about automating those jobs out with AI. And the answer is who is going to be automating those jobs out? Right. And uh, these are ah, the software developers. And I think maybe next five years plus uh, hardware is going to join because it's going to start impacting the or other developing the robotics um, uh, LLM driven robots. Right. So software developers I think are going to benefit uh, immensely because to automate those jobs to create more value uh, will require a lot of work. And um, I can attest myself that I've been working on something for like last uh, 12 plus months. And I can tell you this, converting the value that AI may provide to something that actually works, it takes huge effort. Uh, so I would say that software development is going to be as heavy and as valuable or maybe even more valuable, maybe even um, many access valuable next 5 to 10 to 20 years. So we are going to be very busy. And I think um, the second part that is gonna um, uh, the second kind of jobs which are gonna continue to benefit more, um, or rather benefit more from the rise of AI is high um, qualification jobs that require interaction with the physical world. And it means that anything that requires work outside, outside of the computer, uh, there going to be a greater demand because that productivity, informational productivity bump will have to eventually be translated into the physical world. And in that physical world anything which requires interacting with that world is going to be extremely valuable because robots will come at some point maybe. But um, I would say next five to 10 years being able to operate the outside world using the benefits of AI is going to be huge. And it can be, I'm just, I mean there are maybe 5,000 titles that describe jobs which deal with the real world, uh, or physical world. But I'll just give just simple examples. Right. Electrical engineers. Right. Um, if you want to. Or just think of building a data center. Right. Imagine how much of a physical effort is going to go into building the data center. Someone must come in, put those racks in, wire them up, look at the diagnostics, rewire, pull cables, turn the ac, climb on the roof. Uh, that's a lot of work. Just a simple example, plumbers, um, I don't know, maybe at some point they're going to be robots who do plumbing. But until then someone must come in and uh, fix uh, the sink. Right. And just air conditioning, construction, mechanical engineering. Basically any jobs which require building stuff with your own hands or by managing machines, that's going to continue to increase in demand. That's my uh, that's my take sort of. It's maybe just a segue into. Because we are going to talk a bit more about it. But that's my initial take on this.

Speaker A: Yeah, I agree with you 100% on the uh, the specialty jobs, especially manual labor, particularly, particularly in areas of higher demand as we're building out more Data centers and building out um, more space and capacity to um, host this compute. So that certainly will be there, um, and all the derivative engineering exercises, mechanical, structural, etc. Uh, for sure in software engineering I 100% agree with you in that there is going to be this shift, uh, away from uh, writing large volumes of code to leveraging generative AI to replace some of that, but also to allow engineering to shift to uh, more deeper thought tasks, engaging more with the underlying tooling and data sets and building out capabilities so that uh, you can leverage more of the uh, benefits of AI. And that I don't think is going away, certainly not in the next five years. It may be maybe longer and we'll see in the next five to ten years how the industry continues to pivot. So that brings us to the next question. Um, what types of jobs do we think will actually get created and how can we guide towards that?

Speaker B: Right. Uh, so my expectation, I think uh, this expectation is driven by just looking at the history. If you look at every time there was a revolution, uh, in uh, productivity, like for example going from uh, manual labor to steam engines, right. Uh, before steam engines, if you're making shoes, you would sit and work for hours at a stretch to make a shoe or a couple of pair of shoes. With the appearance of steam engines and the appearance of machines, that life cycle went from one person working for 20, you know, 20, 40, you know, 60 hours to make a pair of shoes, it went to a thousand people on a conveyor being able to crank out thousand shoes per day, right? Not uh, one pair of shoes, you know, in a week or whatever it takes, right. I'm not an expert in shoes but let's say two days if it's a good pair of shoes. So I think, and if you just look at further how those revolutions that drove the large scale increases in human productivity, some jobs disappeared sometimes within years, but it continuously led to creation more and more jobs which allowed humans to operate faster, with a better quality by using their heads more and more. And that's a very solid pattern. If you look at the Internet, if you look at the Internet gave rise to E commerce. Uh, before if you wanted to shop a car, you had to spend you know, a week going through dealerships. Now you can pick a car you want within, I don't know, depending on what you are using an hour, right? So essentially, but that drove creation of large, large scale creation of jobs, uh, in um, if you just think of E commerce, right. Someone had to write those E commerce websites, right? Someone had to write those workflows because now ecommerce goes from two weeks shopping for a car to one hour. Right. There's a great increased demand on receiving those um, um goods, those cars for example. So that automatically pushed creating more and more effective and efficient supply chain. Right. That requires supply chain specialists that require more people building more ships. And it's just sort of just one single thing that creates this pyramid of jobs. Right? Just if you take Internet, it's just micro example of E commerce. How many other amazing things Internet created is just, you know, it's mind boggling. So it's just a simple example and uh, it's just going to continue. And uh, so it's just the initial, initial thought that uh, every time there's a revolution of productivity we humans come up with better, faster ways to do things and we just can create more good stuff at a lesser cost. And essentially whole humanity, I'm very hopeful the whole humanity is going to get richer. Right. Um, the level of um, um, level uh, of wealth globally is going to increase and I hope that's going to also impact the disadvantage. There's 4 billion people are living without running water. Right. Maybe with AI we'll figure out how to enable them to be creative and productive. That's just my initial take. What do you think?

Speaker A: I agree with that. I think a shift to solving more problems that are currently out there, like you mentioned, um, is going to be at least in theory one of the things that is going to benefit from AI taking over things that were previously um, only done by humans. So there will be a shift. I do agree that in general more jobs are going to be created. The biggest challenge or I think daunting thing with this innovation wave compared to others in the past is that this is really the first time there's a direct impact to replacing humans and potentially being even better and more efficient. Right. And you're already seeing that with companies trying to reduce the number of humans they have and replace it with automation. You know when, if you go back in the industrial wave, I mean when um, mobile phones and um, the Internet age and cloud computing and everything before that there was always a clear path forward, say oh okay, my current role and my current skills make it obsolete but I can look at, follow the industry and take a new role um, if I retrain myself or if I stay with it. Right. As engineers we're always looking at the latest and greatest technology for our interest but also to see how we can leverage that for the ability of our um, uh to apply something new and be current. Right? Always about being current and uh, sometimes bleeding edge, but not always the case. So I think there is this element of fear, like, oh, my job is getting completely replaced and there's no new path for me, um, similar uh, to what you've seen with manufacturing jobs and other things. But I think within engineering, the spirit is always around innovation and to innovate. Right now innovation is purely a human activity. Um, until AI can get to a point where it can generate new ideas that are meaningful and that can, uh, replace humans in that regard, um, it's still going to be very much a human element. So that means there will be opportunities to learn new things or solve new problems that we were previously not able to or didn't have the time to do because we were consumed with more mundane stuff. And I think that's. That to me is one of the biggest takeaways. Um, what do you think, Slava?

Speaker B: Yes, I agree. And um, I think we should maybe double click a bit more on the practical, um, aspects of uh, like I said, um, you know, what is the fear? I think the fear is, uh, from. I mean the fear itself is real. And I also believe that it's not justified. Right? Because we as humans, we just scared shitless of any change. We do not like change. We like stability. Um, and now we are very much entering at a very high speed in the uh, year of instability. And uh, we are fearful naturally. Um, and if you just think, I mean, how about we just, you know, step back for a second and maybe a little circle into uh, the benefits if think of this, right? What are the low level, low responsibilities to train high impact knowledge work jobs? M. Who are those knowledge workers? Um, and I'll just give maybe just a few practical examples. Office, um, managers, uh, legal assistants, uh, accountants, human, um, resources, uh, radiologists. And I'll go and say closer to the home, uh, project and program managers and even up top, uh, such as chief financial officers, quality assurance. These are just a few examples where these jobs are super important. If someone stop doing those jobs that work, we are going to be in very serious trouble. But at the same time they're highly trainable. Um, and then the responsibility are uh, fairly low, the accountability is high, responsibility is low, and they're characteristically common that you have to know just to be a, um, um, uh, legal assistant. Right? It's an important job because you have to prepare the briefs for the big guys, right? You have to know a lot. You have to know. I'm pretty sure Everything that legal firm ever done, uh, you have to know, you have to know the processes, the workflows, just a lot of stuff, right? What of it? Which part of it cannot be taught and automated? I would say all of it can be taught and automated. So and the question, natural question becomes what happens, uh, to uh, the legal assistants? And I would say that even we as software engineers, we coded up. It's amazing who is going to take the responsibility for those, uh, artificial legal, uh, assistance. Someone has to be accountable and someone has to be responsible. And it means that just like an internal, they had to be, uh, they should be given a bit of a responsibility. Those artificial A.I. assistants, right? Uh, A.I. legal assistance, they have to, you know, try things in a safe environment, you know, learn from it. Who's going to be providing the feedback? Who's going to be teaching them? Who's going to be um, watching them? Who is going to be, well, answering questions, right? At some point they're going to get in the situation, hey, per my understanding of all the things, I have no idea what to do. What do you think? Right? And essentially what I'm is just first line. I think right now the people who do those jobs, they will have an immense opportunity to be teachers with significant practical experience to help those systems to come online. Right? And I think it's going to take years and years and maybe even tens of years to create this level of, you know, level of um, AI workers who can be trusted. Right? Because right now someone comes to me and says, hey, I have this great legal assistant. You know, he's going to be managing all your uh, you know, your legal uh, case. Oh, thankfully I don't have any legal cases. But it just, you know, just in case, I'll say, wait a second, how do I know if this thing even works? Right? So. And no one will be able to tell me, I expect. Right, because just to train a knowledge worker from zero, for them to come online and to be operational, that takes five, you know, four to five years, right? Maybe with AI and the help of, you know, this sort of a narrowing loop of feedback and improvement, you know, the sort of driving towards singularity, right? Still, humans will have to be in the loop for very long time. They'll have to be able to become good teachers. They'll have to become good, um, um, they have, they have to develop the ability to understand on a higher level what is the things that we have built those level. Level, um, mundane, um, low level mundane knowledge work jobs. And I think this alone is going to start creating a new layer of high grade knowledge workers. Right? Because we just, and I think in the end of the day there are going to be more jobs. We're just going to be much, so much better, doing mundane things faster with absolute quality and cheaper while leaving us to do the creative work regardless of the level. Right. So, uh, that's my, uh, take on, uh, you know, I'm super optimistic, right. And uh, being fearful, I think, just unwarranted because history shows us that every technological revolution, we came out, uh, richer, uh, with more time on our hands, with more opportunities to be creative and more opportunities to do good stuff. So that's my sort of like a ripe take.

Speaker A: I think there's, um, three competing, uh, well, not competing. I think there's three things going on that together could work against this wave of, you know, just wiping out all jobs. The first is, you know, companies are still in business, right? They still have a business to attend to. Uh, we can call that, you know, legacy stuff now or we can call that whatever it is to run the business. But those things, you don't just flip a switch and automate that out. So there is, there's already a time, you know, there's already a, uh, sort of a ramp period there because you have to maintain your business. You can't just walk into office tomorrow and say, okay, we're autumn, you know, we're gonna automate everything. Everyone go home.

Speaker B: Right?

Speaker A: It's not quite there. Otherwise, you know, your business goes nowhere. The second thing is the market. The market. I mean, our stock markets, our industries, the things that people consume that come out of technology, whether it's services or consumables or products, those are not going away. Right. They're still going to have demand and m. Publicly traded companies are still going to have responsibilities to their boards. And boards are always looking to make money. And making money means you got to keep often turning out the same thing you're doing. Um, sure. They're always going to say, yeah, be more efficient, cut this, cut that. But that's not something again, that happens overnight. And it is an evolution. And I think each company is going to go through a different way of getting there. In our previous episode, we talked about how companies are reacting very quickly to, oh, I got to have an answer for generative AI. And so they're racing to do stuff better. Companies are taking the time to do it right. But, uh, they're not necessarily deviating from their core business and how they do things. Yes, it's a revolutionary industry but it doesn't happen overnight. And you are bound by your financial obligations. The markets, the industries, the economies, the consumers, all these things are still going to be pulling at you for things that you can't just immediately replace. It's great to say you're doing that, but until you, you demonstrate that and build products that um, are just as valuable, that's not going to go away, right? So you still need people to staff your firms and businesses to be able to continue to keep the business going. The next thing is, like you mentioned, there's this spirit around innovation, right? And especially as scientists, uh, as engineers, that doesn't just get replaced. And that's a very, very difficult thing to just change. So people are going to continue to want to create and innovate and inspire and do new things and try new things. Uh, and the human mind works at a much faster rate creatively than anything out there today that is meaningful, um, that can stitch things together, that can create communication between people and systems and whatnot. That is valuable. Those things are not going away and those things will continue to press on everyone. So that will itself continue to create, either keep jobs or create new jobs. And I agree 100% with your point that we will pivot away from things and move to more deeper thinking roles that enable us to leverage the tooling and inspire new opportunities as much as possible. And I think that's basically part of our guidance here. As an engineering manager or leader, you have an opportunity, right? Don't wait for, don't treat this as something coming down that you have to um, uh, adhere to, embrace it, be ahead of it and lead the curve on innovation so that you can uh, deliver more meaningful stuff with the team that you currently have, retrain, retool, regroup, right? Um, so that people still have a meaningful place in our professional society.

Speaker B: Yes, I agree. And um, I think your point is that introduction of AI does not necessarily mean loss of jobs at all. It may also mean that the speed, uh, uh, of delivering good stuff is going to accelerate. Maybe you have the team, same team, same size team that delivers twice more in the same amount of time, right? And maybe even more than twice without killing up. Um, you know, if you, if you want, if you pick an average technology team these days, if you ask them to deliver twice more, uh, they're going to have to work to twice more, right? They're going to have to have a, you know, 60 to 70 to 80 hour week to deliver twice more now. But if you ask yourself where is the time going Right. Uh, getting the information together, understanding where things are, understanding why things are working, understanding where things are not working. Figuring, uh, out how to make things work. Right. Imagine if this is, uh, instantaneous. Right? And so that building this idea of how the real world looks like through continuous meetings and conversations and emails and all the good stuff, and then you just push a button and say, hey, what is and how it is? Right? And you have a completely truthful, clear picture of what is and how it is and what's next and what needs to be done. All you have to do, just sit down and go and do it. Right? Uh, and that's all. Right. And that's where the creativity kicks in. Right. Because you have a gigantic universe in your head, which is. I don't think LLMs will ever be able to do that, to build a universe inside. Um, and, uh, um. How many human brains do we have on the planet? Eight billion. That's extremely powerful. Try to replace it with two dozen data centers. Yeah, let's see that.

Speaker A: Absolutely. I mean, you know, that's a. I think that's a great way for us to kind of summarize here because, um, you know, we should be inspired, right? We as engineers should be inspired that we have a role to play here. We have it. We can influence. But part of the influence comes from how we, uh, adopt and appreciate this revolution, this new phase of this, uh, you know, our industry growing, and we should be a significant part of it. Right? And as leaders, the more that we are able to pivot and be versatile, the more successful we and our teams are ultimately going to be. So, Slava, to summarize here, can you summarize and provide a checklist that our listeners can use, uh, starting today?

Speaker B: Uh, thanks for asking. And usually we try to provide. Not try. We almost always, I think, um, out of, uh, more than 30 episodes, we always do a checklist so that engineering managers can pick it up and start implementing. I think this time we are going to loosen it up a bit. And um, um, uh, I have only one thing on my mind. And uh, uh, I'm wondering what you have. In the end of the day, it's going to come down to this. There are going to be two kinds of jobs. The ones which are getting automated out and the ones who are doing the automation. And the question becomes, which job would you like to have? And if you want to be in the job of, um, um, automating things out using AI, I would say start automating things around you today. That's like a One single checklist point for, for today. Look around you, look at unsolved problems inside your own domain and start asking yourself question and start implementing. The question is, why does it exist? Why is it painful? And how can I apply this new amazing technology for that pain to go away? How can I automate things out, Problematic thing, Problematic things out of my life? And, uh, how much of time you as an engineering manager are going to spend on this depend? I mean, it's your choice. You can spend an hour, you can spend two hours, you can get half of the team to start working on it. You can get two people, um, and rotate them because everyone will want to participate. But, but my take here is that start automating things around you.

Speaker A: 100% agree. And the only other thing I would add to that is be, uh, creative. The industry is ripe right now for change. Right? It's ripe for whatever. And that means it empowers especially managers and leaders to set a path forward. So on you and your teams, yes, 100%, automate what you can and then create new opportunities that is going to empower your business going forward, leveraging these tools. So you're going to set your own path in your company, in your space. It doesn't necessarily mean everyone's going to do it the same way, but if you're creative and you can be a part of this innovation curve and um, right, period, you absolutely can sort of control your own fate, destiny, uh, to a large degree. So that would be the only other thing I would add to that checklist.

Speaker B: Yep, 100% agree. Good, good take. And, um, yeah, Adam, thanks. Uh, um, that was good stuff, as always. And, um, um, more good stuff is coming. And, um, I encourage our listeners to share this episode if they like it. And as always, we are looking forward to feedback and suggestions. And, um, I'd like to mention that, um, we do offer customized training, uh, on effective engineering management to help companies accomplish more with less. And you can reach us@contact effectiveam.com and www.effectiveam.com to learn more.