Using AI to replace your leadership presence, whether through AI video clones, copy-pasted team messages, or scripted feedback, erodes team trust and accelerates disengagement. Lia’s framework is simple: use AI to brainstorm and prep, but never let it substitute for the human moments that make you indispensable. This episode is essential for any leader wondering how to stay relevant and keep their team motivated in the age of AI.
This episode is part of the AI for Leaders & Teams series. Browse all episodes →
Mic drop moments
“A manager is a leader. It is someone that enables better results for their teams, that empowers them, that motivates people, that helps kind of look across, well, where is this person best situated and what could that person be doing?”
“…the communication and all of the connective tissue to influence people, to get people aligned and on board, to hold people accountable, to make them feel seen and recognized. That’s the stuff that we really don’t want to replace.”
Episode highlights
Where Sergey Brin gets management wrong
Google co-founder Sergey Brin recently claimed on the All In podcast that managers are the easiest roles to automate. Lia pushes back hard. The claim treats management as a task-routing function: drag tasks into Asana, approve timesheets, click through a Jira queue. Yes, that can be automated. But that was never the job. Google itself tried eliminating middle managers in its early days, and it was a complete disaster. People do not self-organize. That lesson predates AI by decades.
What a great engineering leader actually does (and why AI cannot do it)
Lia points to Rob, an engineering manager she worked with on the HoloLens team at Microsoft, as the clearest example. Rob did not spend his days managing a task list. He set the vision for his team, worked across product, design, and marketing to align on where the product was heading, and removed bottlenecks before they stalled engineers. When a feature came out of the headset and needed iteration, Rob would sit side by side with the design lead, both wearing headsets, talking through what a better solution looked like, how many engineering hours it would cost, and where it fit in the schedule. That is cross-functional judgment in real time. None of it is automatable.
The connective tissue AI cannot replace
A design leader Lia worked with spent her time figuring out where designers were getting stuck, how to get design a bigger seat at the table in an engineering-driven culture, and how to communicate in a way that brought stakeholders on board. Automated meeting notes and ideation tools could have made her faster, but they could not replace the influence, the alignment work, or the relationship-building. That connective tissue, getting people aligned, holding them accountable, making them feel seen, is what Lia calls the real job of a manager.
The skills that keep you irreplaceable
Lia names the leadership competencies that will outlast the automation wave: communication, collaboration, conflict management, recognition, and the ability to read your team. One example she flags is noticing when a team member who used to be energized starts to drift, even if their output looks the same on paper. An AI will never catch that shift. Neither will it know how someone will respond when you deliver hard feedback. These are human reads that require presence and pattern recognition built over time.
How you accidentally put your own job at risk
Here is the trap Lia lays out plainly: if you are so buried in work that the only thing you have bandwidth for is pushing out the task list, you are spending all your time on the things that can be automated. That is a choice that puts your role at risk. Delegating and stepping back to do the higher-order leadership work feels harder in the short term, but it is exactly what makes you future-proof.
What to do right now
Lia’s call to action is direct. Leaders need to be learning about AI and how it affects their industry every day. Not to panic, and not to replace their teams, but to understand it well enough to make good decisions about what to automate and what to protect. At the same time, double down on the skills that AI cannot touch: feedback conversations, motivating teams without budget to throw at the problem, managing your own reactivity when things get uncertain. These are the muscles to build now. A future episode will cover a specific problem-solving checklist for figuring out what to automate and what to leave human.
Related episodes in this cluster
- Ep 232: What to say when your team is scared about AI (and what not to say). When your team hears “AI,” they immediately think their job is on the line, and if you haven’t addressed that fear directly, it runs silently in the background of every conversation. Lead with your vision for how AI enhances your team’s work rather than replacing it, and save 10 to 15 hours per project by targeting the tasks that were never the core of anyone’s job. This episode is most useful for owners and leaders who need to open the AI conversation without triggering a productivity crash or a quiet exit.
- Ep 208: What AI should replace on your team (and what it should not). AI can handle tactical tasks like notetaking and scheduling, but replacing an entire role means losing the invisible human value that holds teams together. Lia walks through a practical way to audit what to automate versus preserve, using the example of an executive assistant whose real job was culture, candor, and translation. This episode is most useful for founders, operators, and leaders who are under cost pressure and tempted to automate their way out of staffing decisions.
- Ep 197: How to navigate the Gen Z and Gen AI collision on your team. The tension between Gen Z workers and AI tools creates three specific problems for leaders: entry-level jobs disappearing, critical thinking skills going undeveloped, and team members outsourcing work entirely to AI. The fix is not banning the tools but training people to think first and use AI second. This episode is most useful for owners and leaders who are watching their newer team members default to copy-paste AI output instead of doing the real work.
- Ep 189: The hidden cost of using AI as a leadership shortcut. Using AI to replace your leadership presence, whether through AI video clones, copy-pasted team messages, or scripted feedback, erodes team trust and accelerates disengagement. Lia’s framework is simple: use AI to brainstorm and prep, but never let it substitute for the human moments that make you indispensable. This episode is essential for any leader wondering how to stay relevant and keep their team motivated in the age of AI.
Want to go deeper?
The Manager Essentials Program. Lia’s flagship program for new and rising managers.
Full transcript (click to expand)
Are managers the easiest role to automate right now? Well, that’s what Sergey Brin, the co-founder of Google, said in a recent episode of the All In podcast, that managers are the easiest role to automate. And he has been looking for ways to automate managers for years. But I don’t agree with that, right? I think it’s a misunderstanding of the role of managers and the potential of managers when they really step into their role as leaders, that makes them indispensable.
And that’s what we’re going to talk about today. I’m Leah Garvin, and this is The New Manager Playbook, where I break down the ins and outs of being a manager. And now, especially in the context of bringing AI into our workflows, we’re going to be talking about that more and more on the show, right? How to understand where AI fits with being a leader, when to use it appropriately, and when it doesn’t make so much sense, okay?
So I could not be more excited to be talking about this, because it’s bringing in a lot of my background from working at Google, Apple, Microsoft, really, in some of the earlier days of the evolution of the digital assistants and figuring out what appropriate use cases there are for AI. And now, as it’s kind of running full steam ahead, it’s a moment for us all to think about as leaders, well, when is it making sense to adopt some of these tools?
And when is it, really, we still want to have the human interactions and really do that skill building there? So that is something that I want us to really be thinking about. We all should be learning about AI on a day-to-day basis, okay? If we’re not learning about AI and how it impacts our industry, we’re actually completely missing the point as leaders. That’s part of our job now, to understand this new technology. It’s not even that new, right, this technology and how it’s impacting us.
Because if we don’t understand it, we’re going to be left behind. But it doesn’t mean just jumping on the bandwagon and getting rid of all our employees, okay? And you can guess that that’s what I’m going to be talking about and being the drum about on this show all the time. And so when I heard this clip of Sergey Brin saying this, well, first thing was I thought back to early days of Google. This is, I don’t know, how long they’ve been around, 20, 25 years, maybe more, where Google was one of the first companies that actually tried to get rid of middle managers.
They thought they weren’t necessary, they just slow things down, and it didn’t work. It was a complete disaster. Because the reality is, people don’t just self-organize. And this was long before automations, but it’s something that’s been in the back of the mind, kind of in this mindset, I think, from the engineering workflows. Because with engineering type work, yeah, you can assign, you can set a task queue and assign it and have auto approvals. A lot of that actually already is automated.
And so I want us to really discern when we’re thinking about what we want to automate and not. It’s not just like everything, it’s certain kinds of functions, right? So that’s why I hear this and I’m like, dude, you’re looking at management as it’s just assigning a bunch of tasks in a queue, auto approvals and whatever, clicking timesheets. Well, yeah, that can be automated, but that’s not the role of a manager, right? And that’s not what we talked about in the new manager playbook.
A manager is a leader. It is someone that enables better results for their teams, that empowers them, that motivates people, that helps kind of look across, well, where is this person best situated and what could that person be doing? That is what a manager is for. So sure, if you are just literally dragging and dropping tasks into a task list, into a click of an Asana workflow, a Jira, whatever, whatever, then yeah, that can be automated.
So let’s like automate that part and save your energy, save your capacity for what it really means to be a manager. Now I want to share an example from engineering because this is a place that management is being automated very, very quickly. And I think back to when I worked on the HoloLens team at Microsoft, and this is one of the most incredible, fantastic engineering managers I’ve ever worked with. His name is Rob. Rob, if you’re listening, shout out to you and your genius.
Why was Rob incredible? Well, Rob didn’t manage the day-to-day of his engineering tasks, okay, and that was already him because we would do sprint planning and things like that, but he would set the vision for the team. Rob was also a partner working with the product leadership, the design leadership, the marketing to see, hey, how are we all communicating across and understanding where is the vision of the HoloLens product and where are we going with it?
And how is every sprint that we’re planning? How is every milestone leading into that? So he was working as a cross-functional communicator, understander, really removing the bottlenecks that were happening in engineering. So that was something that we really can’t replace with AI because he was having conversations. He was also influencing stakeholders and saying, hey, we have this deadline, we don’t have enough resources here, where can we shift things? Where can we prioritize the different features that we want to launch and where does it make sense?
He was also sitting down with design leadership when we would have something come out in the product. So we would create a piece of the product, we’d look at it in the headset, and then we would iterate and make changes and resolve issues. He would sit down hand in hand with the headset on, the design leader in him, and they’d look at it and say, hey, what’s a better way to do this? How many engineering hours is that going to take?
Where does that fit within the schedule? How can we adjust things? This is what he was doing, okay? So none of those things can be automated and the things that Sergey Brin says can be automated, he already had that automated and this was like 10 years ago, right? So I just want to clarify that there’s sure there’s a lot of stuff we can automate, but the leadership skills you saw, I was going to call it soft skills, the leadership, the hard skills that we bring as managers, as leaders to make sure our team members understand what they are doing, that they are best positioned for their roles, for their tasks, the moving of the chess pieces, if you will, when we see, oh, there’s an opportunity to go over here and we pivot and we look at the bigger picture and we go bird’s eye view and then we go narrow and then we go wide and narrow, we figured out we’re always looking across these dimensions.
That is the power of a manager. Now I say this because I want us all as leaders to use our own discernment when we’re thinking about what kinds of things we should automate or not. And I’m going to do another episode around kind of the problem solving checklist that I run through with teams when I’m consulting with teams or working with folks like, well, how do we know what to automate? Because it’s not everything, but it’s not nothing either.
There’s so much power to these tools, so we want to use them the right way, but it’s about understanding them. And so today we’re talking about what are these examples of leadership skills that really can’t be automated. And then again, I’ll talk next time about really where we can utilize them and some of the great tools that are out there. Now let me share another example of a design leader, right? So working with a lot of my career in tech was in design program management and operations and design leader that would sit down with the team and observe, okay, well, where are they getting stuck when it comes to tools?
What are some of the issues with communication across design and engineering and product? Where can we really make sure design has a bigger seat at the table? In the tech industry, it’s engineering driven as no surprise. And so how do we ensure that we are communicating in a way that gets folks on board, excited, and ensures that everyone is really positioned in the right way? That is another leadership skill that you really can’t replace with automation.
Things that would have made that easier would be having automated meeting notes and cascading of information, right? Like those are things that could have really streamlined that workflow because connecting the dots around communication can be a challenge. Maybe there’s some sort of use of ideation tools to say like, hey, here’s a couple options we have that now blow out a bunch of ideas for this, right? But it wasn’t removing the communication pieces. And so that’s really my point here is the communication and all of the connective tissue to influence people, to get people aligned and on board, to hold people accountable, to make them feel seen and recognized.
That’s the stuff that we really don’t want to replace. No one is asking you to replace that, right? So it goes back to what problem are we solving because that’s not the issue. And I want you to think about what leaders that you have worked with throughout your career that really, really embodied the qualities that you want to emulate as a leader. How did they communicate? How did they show up? A big one is how did they manage their own reactivity and responses and dealing with change and uncertainty?
How did they show up when they didn’t have all the answers? These are the things that you really want to lean into because these are the qualities that keep you relevant, that keep your job necessary. The more we’re leaning into that, the more we are removing this, is this necessary? Should we automate it, right? It’s when you pull away from those things and you are drowning in work because you’re not delegating, you’re not managing your time and all you have space for is, well, just I got to get the task list out there, at least that’s happening.
If you’re leaning only in skills and in tasks that can be automated, you are actually putting your job at risk by choice. And I know that it is not easy to delegate, I know it’s not easy to take a step back and look at your calendar and think like, okay, here’s some things team members can take on and but I got to get them up to speed and delegate it and figure out what the handoffs are on decisions.
That is something that can take a little bit of time. It’s not necessarily in one 15 minute conversation, that is future proofing yourself. So it’s really important that we are always building these skills around, I said it already, communication, collaboration, conflict management, recognition, how are we understanding all of the different people on our team and what their strengths are and how to best position them and put them in front of the right people and in front of the right opportunities.
How are we motivating our team members even when there’s all this stuff going on, everyone’s frustrated, right? And we can’t pay them more. What are we doing? This is all the stuff I talk about in the new manager playbook book, because those are the skills that will withstand the test of time. And that’s the stuff that I hope we are re-energized around. We feel more empowered and we get support if it’s not our strength. So if you feel like, well, I get really uncomfortable when there’s a conflict or when someone gets defensive when I’m giving feedback on who doesn’t get uncomfortable when someone’s defensive, That is an opportunity to develop a skill there, to flex those muscles, to strengthen those muscles, to get better at it.
It’s really hard to give difficult feedback. Someone, Hey, this is like, I was looking at my episodes from the past couple of years. I’m like, I had like 25 episodes on feedback because it is the hardest thing. Feedback is something that sure we can, you know, utilize AI to understand like, Hey, what are some different ways I can say that, but it is not going to know what the response is on the other end. The, you know, any AI is never going to be able to predict what your team will say.
So that is a competency you want to develop. How do you recognize when a team member was really, really dialed in and motivated? And then when stuff started to shift and their whole energy’s off, even if they have the same output, that’s what you as a human is understanding and diagnosing. Okay. So these are the skills, again, we want to be building, cultivating, developing, really strengthening in ourselves and across our teams. And if you’re overseeing managers and managers of managers and things like that, this is where we really, really want to focus development.
This is where we really want to make sure folks are, are, are highly competent, are feeling confident about it, are able to, because this stuff is very, very scary. A lot of people in teams, in companies hear AI and they start wondering like, gosh, how long do I have? So it’s on you to be developing these skills so you can bridge this transition. That’s really, really part of the job. So again, that’s why I’m going to start to talk about this on the podcast all the time so that you feel equipped.
You have the new tools you need. You have literally the new manager playbook for dealing with all of this technology. All right. So the first thing we want to do is understand what are the skills and to be having a clear path for how we’re going to be developing those. Okay. So I cannot wait to support you on this. We can all get there. We can all do this together. A lot is changing. Okay. So let’s all say a lot’s changing.
I’m going to do my role, which is to become versed in it, and I’m also going to build these leadership skills because those are going to keep me relevant. All right. See you next time. Transcribed by https://otter.ai
Questions this episode answers
Are managers really easy to automate with AI?
The claim gets it wrong because it confuses administrative tasks with actual leadership. Yes, dragging tasks into a project management tool, auto-approving timesheets, or assigning a ticket queue can be automated. But the real work of a leader, setting team vision, removing bottlenecks, influencing stakeholders, recognizing when a team member's motivation has shifted, and navigating conflict, cannot be handed off to an AI. The moment you reduce a leader's role to only the tasks a machine can do, you have already misunderstood what the role is supposed to be.
What leadership skills can't be replaced by AI?
The skills that hold up are the ones built around human connection and judgment. That includes cross-functional communication, conflict management, giving and receiving difficult feedback, recognizing when a team member is disengaged even when their output looks the same, advocating for your team with stakeholders, and knowing how to motivate people when raises aren't on the table. These require reading the room, responding in real time, and building trust over many conversations. No AI can predict how your specific team member will respond in a tense one-on-one.
How do I future-proof my leadership role against automation?
Stop spending most of your energy on tasks that can already be automated, and start investing that capacity in the skills that require a human. That means delegating more, building your team's ability to operate independently, and leaning hard into communication, collaboration, and recognition. If you are so buried in task management that you have no room for actual leadership, you are choosing, by default, to make yourself replaceable. The leaders who stay relevant are the ones who keep developing their ability to connect with, align, and advocate for the people around them.
What parts of a manager's job should actually be automated?
Anything that is purely administrative and repeatable is a fair candidate. Task assignment in tools like Asana or Jira, meeting notes, cascading routine information, and generating a first draft of ideas are all areas where automation can save time and free up capacity. The point is not to automate nothing and not to automate everything. It is to be intentional: use these tools to clear the low-value work off your plate so you can spend more time on the leadership work that actually requires you to be in the room.
Why did Google's experiment with eliminating managers fail?
Early on, Google tried removing middle managers entirely, believing they slowed things down and were unnecessary. It did not work. People do not simply self-organize without leadership. Teams still needed someone to set direction, resolve conflicts, connect the work to a broader vision, and make sure individuals felt supported and positioned well. That lesson came before AI was even part of the conversation, and it still applies today. Coordination and human judgment are not inefficiencies to eliminate; they are what makes a team function.