When your team leans too hard on AI, the fix is not banning it. Set clear expectations first, then decide intentionally what to do with the time automation frees up. This episode is most useful for founders, owners, and team leads who are embracing AI but feeling uneasy about how their people are actually using it.

This episode is part of the AI for Leaders & Teams series. Browse all episodes →

Mic drop moments

“…as a leader, it is our responsibility to create some kind of semblance of stability on our teams. Maybe we didn’t ask for that, but that is what we signed up for.”

“We’re not replacing the human, we’re replacing some of the stuff that actually was hurting that business from being able to deliver a better result.”

“There is a lot of questions that we don’t have the answers to and we don’t have to make up the answers. We just gotta stay in the conversation.”

“…we utilize AI as a superpower, where we utilize it to supercharge our own skills, where we supplement things that we weren’t very good at or that took forever…”

Episode highlights

Your team is already thinking it

The moment you say “AI” in a meeting, your team members go to one place: “This is how they replace me.” Lia is direct about this because she hears it constantly from both sides. Leaders assume their teams are fine. Team members, when asked privately, say it is all they think about. If you have not addressed AI fears head-on, those fears are already running in the background of every conversation, every performance conversation, and every new tool rollout.

What happens when you skip the context

Saying “Everyone should learn Claude, it’s the future” without any framing is one of the fastest ways to tank motivation. Even tech-savvy, enthusiastic team members carry a background dread that the day is coming when the tool will be good enough and the person won’t be needed. That dread quietly shifts behavior. People start coasting, collecting a paycheck until they lose the choice. The productivity dip is not laziness. It is a rational response to feeling like you are building your own replacement.

The reframe that actually works

Lia’s position has shifted. A year ago the conversation was mostly about uncertainty. Now, with tools like Claude significantly more capable and customizable, there is a real strategy available. The reframe is not “AI won’t affect your job” (no one believes that) and it is not “work faster or get left behind” (that just creates fear-based culture). The honest, productive version is: we are going to use AI to remove the peripheral, time-draining work so you can spend more time on the human stuff that actually requires you.

Real examples: superpowers, not substitutions

Lia walks through two concrete examples that show the difference between replacing a person and replacing wasted time.

For agency clients doing repetitive project kickoff research, she builds custom Claude skills that run searches and consolidate data points that previously required hours of manual connection work. The account rep still has the client conversation. They just arrive better prepared, and they get 10 to 15 hours back per project.

For the snippets tool she uses with her own clients (a bird’s-eye view of what a team is working on each week), she moved it from a Google Doc to a web app built in Claude Code. The app surfaces real-time trends across priorities, bandwidth, and recurring blockers. Lia still brings the human context: knowing that a particular pattern showing up week after week connects to a specific team dynamic only she and the client understand. The AI consolidates the data. The human applies the solution.

Where AI still gets it wrong (and why that is the point)

Meeting note tools like Fathom, Zoom, or Otter are a useful everyday example. In Lia’s experience, the AI-captured main point is off at least half the time, often closer to 75 percent. The model flags what was mentioned most frequently, but it cannot know that the quietest person in the room holds the most influence, or that a popular-sounding topic is actually a lower priority based on a one-on-one conversation you had beforehand. The human has to stay in the conversation. That is the reassurance to give your team: your context, your judgment, and your relationships are what turn AI output into something actually useful. Handing off raw AI notes without reviewing them is a choice, but it is not a good one.

What to actually say to your team

You do not need all the answers. Nobody past the next six months does. What you do need is to stay in the conversation instead of avoiding it. Share your actual vision for how AI fits into the work. Be specific about what tasks you are targeting and why. Name the goal out loud: we are clearing the peripheral stuff so you have more space for the work that requires a human. When team members see that the strategy is about giving them superpowers rather than auditing their necessity, the fear does not disappear, but it stops running the whole show.

Related episodes in this cluster


Want to go deeper?

The Manager Essentials Program. Lia’s flagship program for new and rising managers.

Learn more


Full transcript (click to expand)

Lia (00:00): There is a lot of fear in the workplace right now. Every single day that goes by, there’s a new headline about layoffs and replacing people with AI. And I think there’s no person out there that isn’t at least thinking about what does the next five years look like? What is my job gonna look like? Or my friends or my spouses or my kids future? These are these questions that all of us are facing.

And as a leader, you might be freaking out wondering the same things. But I say this because as a leader, it is our responsibility to create some kind of semblance of stability on our teams. Maybe we didn’t ask for that, but that is what we signed up for. And that is what we are being called to do right now. It doesn’t mean we have to have all the answers and it doesn’t mean we gaslight folks and it doesn’t mean we say, nothing’s gonna change and we really have no idea.

It means we stay in the conversation and we figure out a strategy for right now how we support our team members in kind of making sense of what’s out there for them. And that’s what I wanna talk about today is how to talk about AI with your team, how to make space for folks to learn and experiment and utilize tools without feeling like they are creating the thing that will be replacing them because of course they’re gonna be doing that pretty begrudgingly.

How do you just deal with what we’re facing today? I don’t have all the answers. I don’t know what the future holds either, but I do know that when we stay in the conversation and we show our teams, hey, we’re here, we’re all trying to figure this out together, things will go a lot better than if we don’t do that. All right, let’s dive in.

I mentioned on the show last week, I have been diving head first into cloud code. I’ve got terminal open, I’m building web apps. I’m working with clients to figure out how to really supplement their workflows to give their team members superpowers. And as I’ve been doing this, I’ve been feeling really optimistic. Now, if you’re a long time listener of the show, you might’ve heard my tune was a little bit different before. I feel like I still do.

we are having the wrong conversations around AI. So much of it is around replacing people. Should we do layoffs to cut costs and then figure out later? I think we’re not having the right conversation. What I want to do on this show is equip you with the tools to have the right conversation because there is so much out there that we can do to actually save our team members time and give them space back to focus on the things that they really should be doing. We can also…

be utilizing things like Claude code and building out skills and like Claude skills to build our own skills and give ourselves the ability to remove a lot of wasted time, duplication of efforts, all the stuff that I already support you with when we talk about these things on the show. It’s actually, know, automating some of the systems. So there’s a lot there. But when we talk about AI with our team, before we even talk about what’s possible, we’ve got to remember

what the people on the other end are thinking. Where are they going the moment we say AI? And I’m gonna tell you, it’s not a hypothetical question, it’s not rhetorical. They’re thinking, my gosh, you are getting ready to replace my job. Or they’re thinking, gosh, how much longer is it gonna be until I don’t have a job anymore? I’m not guessing. I know that is what people are thinking. That is what people talk about every single day when I meet with business owners.

or VPs or high up leaders, they’re feeling like, I don’t know if people are thinking about that. Your teams, when I talk to your team members, the employees, the folks on the ground, that is what every single one is saying. So just like newsflash, if you feel like, I don’t know if that’s really the case, that’s what we’re all thinking about. With that said, what does it look like if you go to your team and you say,

Hey, everyone, like learn Claude, it’s the wave of the future. You you need to be up to speed on this and you don’t say anything besides that. What do you think goes through their minds? ⁓ here we go, I knew it was happening. This is it, it’s only a matter of time and it starts spinning. And even if your team members are super tech savvy or super into it and adopt these tools right away, there is a good chance they still have this looming feeling in the back of their minds, like, gosh.

this is it for me, there’s gonna be a day when you’re gonna say, okay, well, like the tool is working really well, like what do need you for? This is what’s going on and that is because we haven’t set any context at all. Now, when that starts to happen, you see productivity dip because people think, well, I’m just gonna stick around and get that paycheck till I don’t have a choice anymore. I see a lot of folks in corporate teams feeling like, well, especially if you’re working in tech and you’re building software that’s…

designed to be replacing roles. Meta has a planned set of layoffs coming May 20th. They have publicly said to lay off at least 10,000 people. think that’s 10 % of their workforce, if not more, because of AI. They have said that. So like, this isn’t stuff I’m guessing at and making up. These are things that in big tech is being talked about publicly as like, So of course your teamers are thinking about this. Now, in my opinion,

If we were the thousands of people that I worked with and supported and being better leaders, I’d have to think, well, that’s not gonna be good. That’s not gonna create the right culture that we wanna create inside of our team. That’s not gonna create the right motivation. That’s gonna create a Lord of the Flies situation. But there is a belief in some companies that we should scare people into, know, getting, working faster as if we’re gonna outrun AI. I’m not trying to give you a tools to outrun AI. I don’t think that’s possible.

I think what we can do is to understand where we work with AI, where we utilize AI as a superpower, where we utilize it to supercharge our own skills, where we supplement things that we weren’t very good at or that took forever or that we are reinventing the wheel every single time with our clients. And we use it in that way so that we become even better at the human stuff that we should be doing. That’s what I see as possible.

And that’s the conversation I encourage you to have with your teams is to figure out like, what is your strategy? What is your belief about this? And then how are you helping your team members figure out this yes and. And this is what a lot of folks have been bringing me in to do is to look at, okay, let’s look at where we have these inefficiencies. Where’s their duplication of efforts, especially agencies that have to do some kind of like a lot of upfront research on kicking projects off.

or on research and strategy and be kind of putting out the same reports over and over. They’re bringing me in to evaluate, okay, where can we utilize and set up our own custom cloud skills, which I then build for them so that they’re totally dialed in so that they can save their team members 10, 15, 20 hours per project, per team member on things that they were getting really stuck on. So this is one of the ways that we’re using these tools to give our team members superpowers.

supercharge the skills instead of just saying yeah like it’s a matter of time until you’re out of here. And we I think there’s a way to reframe a lot of this fear by talking about our own vision as leaders. Our own vision as of incorporating it into giving ourselves more space to do the human stuff. Now for example

When I’ve recently built out a tool for the snippets tool that I talk about on the show, this is coming soon. Right now I’m beta testing it with a few of my clients to make sure I work all of the bugs. It was a Google Doc. I built it into a web app in cloud code so that we can identify proactively trends across what’s getting done, what priorities are coming back, where people are getting stuck, where bandwidth is at. So this snippets tool that gives you a bird’s eye view into what people are working on instead of having it as a Google Doc.

I’m using, I used AI to build this into a web app that gives you these real time trends. This is pretty freaking awesome. Now I didn’t replace myself in having that as a tool. I use this so that I can have something at a glance, show me what’s happening on this team. And then I can go back through it with my clients and say, hey, when this kind of thing is showing up week after week, here is the…

here’s actually what’s going on behind the scenes. I can then say, because I know that context with that client, I can say, ⁓ you know, we’ve struggled with Joe in the past, feeling, you know, wanting that more recognition. Now I’m seeing this show up again. Let’s try this. And we have that human conversation. So we’re utilizing a tool to consolidate the data, help us extract the themes so that we can then apply a human solution to it. I’m not asking AI to just spit something out for us.

No, it’s pulling together the data that would take a lot of manual work to find so that we have an answer there. And this is why I’m so excited, right? Is because when we are finding what are things that we actually couldn’t do before or would have taken a lot of different connection points behind the scenes, how can we use that to bring it together? This is where the superpowers are. Same thing as I mentioned with my agency clients. By creating custom cloud skills that

that run a lot of these searches and different connection points that they were doing to be able to prep for client kickoffs, this is not replacing that account rep on the job. This is giving that person the ability to actually have a better and more informed conversation that can then support their client in achieving their goals faster. We’re not replacing the human, we’re replacing some of the stuff that actually was hurting that business from being able to deliver a better result.

Those are the ways that we talk about the power of AI without it feeling like it’s you or the computer and one of you isn’t gonna win this battle. So there’s a lot out here. And I think, again, if you’re a long time listener, can see my tune has shifted against, hey, I’m nervous. There’s a lot that we don’t know to, hey, as these tools and capabilities have gotten so much stronger and so much more fine tune.

I think now we’re at this moment, we’re using it very, very strategically and in a very customized way with our teams. We finally have a better way to talk about the strategy of utilizing AI. A year ago when it was like, we’re typing into chat GBT and we’re just copying and pasting and half of it was garbage, that’s where we weren’t.

quite in a maturity, I think, to be able to have a really clear strategy around it for how to engage with it in our teams. You know, it’s always been a really powerful tool for saving time, reducing steps, you know, the note taking is obviously like stuff like that. But I think right now with the ability to design something that is taking the things that are really the biggest sticking points on your team, it’s bringing together all the things that I always talk about to life.

It’s not you’re placing you in feedback conversations. It’s not letting you skip the hard part of delegating, which is kind of figuring out the context and what success looks like. It’s giving these other, I would say, kind of periphery activities that take away from the core of what you’re doing. That’s what, that’s where we can really save time. Give you another example. ⁓ Clinicians or therapists or some coaches, you know, they capture notes.

of their sessions and everything’s like that so that they can ⁓ just kind of have a longitudinal plan. Utilizing AI for analyzing trends and patterns and figuring out, like what’s happening across the board? That can be really useful. Now we don’t take that at face value. Now we layer in the fact that we were in that session, we were in that conversation. We say, here’s a perspective based on a transcript.

Now I was actually in that conversation or your team member or whoever, what is the actual truth in the middle? What is the actual reality? And I’ll give you an example. ⁓ Take any meeting where you have your meeting notes transcribed, right? So you’re using Fathom or maybe Zoom is doing it or Otter or whatever these tools are. And it’ll give you the top points, you know, action steps. If you read through those,

I’m gonna say at least half the time, maybe 75 % of time, in my opinion, when I read through that, the main, main point wasn’t fully accurate of what the AI captured. The AI thought, you know, this thing was mentioned a few times. So then like by virtue of like understanding language learning models or whatever they’ve captured because it was mentioned a few times, it must be more important, right?

What it doesn’t have is your own context. It doesn’t know that, okay, this thing was emphasized more, or this person has this title. It doesn’t sound like the most important title, but they’re the most influential person in the room. Or you have the context that you had the one-on-one with someone before, and you know that, hey, this is a popular thing to say, but it’s actually not the priority. This is why the human still has to be in the conversation. And this is what I want you to reassure your team members with, is…

Don’t just spit the notes out and send them to the client. Go through them, read it, figure out what is the piece in the middle. Add that layer of personalization. Your team members, when they decide to opt out of being a human, I was gonna say, of doing that human piece, and they just wanna hand off that AI output without looking at it, that’s their choice too. So what we need to do as leaders is show them how to…

apply this to take it to massage it, to use it as a data point, but not the all or nothing, right? That’s part of the conversation. And again, that’s what I’m helping businesses with. It’s so exciting. There is so much out there. ⁓ It’s a scary time. People are worried. Am I obsolete? Nobody knows what’s happening, you know, past the next six months. Seriously, it’s weird, but we’re here. We’ve got to show up. We’ve got to equip people with the tools. And if you need support, I’m so excited to help guide you on this journey.

Reach out at hello at liagarvin.com and we will build what this looks like to just figure this out and have these conversations with your teams. There is a lot of questions that we don’t have the answers to and we don’t have to make up the answers. We just gotta stay in the conversation. All right, see you next time.

Questions this episode answers

What should I say to my team when they're scared about AI replacing their jobs?

Don't pretend nothing is changing, and don't stay silent either. Your team members are already thinking about job loss every day, so the silence makes it worse. Stay in the conversation. Share your actual vision for how AI will be used on your team, and frame it around giving people superpowers, not replacing them. When you say 'learn this tool' without any context, people hear 'you're on the way out.' Give them the context. Tell them what problem you're solving together and why their human judgment is still the thing that makes it work.

How do I get my team to actually use AI tools without them feeling like they're digging their own grave?

Start by being clear on what problem the tool is solving, and show people how it removes the stuff that was slowing them down, not the work that makes them valuable. If an account rep used to spend 15 hours on research before a client kickoff, and now a custom tool does that in minutes, they can show up to that kickoff sharper and more useful to the client. That's the frame. You're not replacing the person, you're replacing the bottleneck. When your team sees AI saving them from the tedious parts, they stop feeling like they're competing with it.

How do I talk about AI strategy with my team without creating panic?

Be honest that you don't have all the answers, but make clear you have a direction. The panic comes when leaders go quiet or drop a tool on people with no explanation. Instead, share your belief: AI handles the peripheral, time-consuming tasks so your team can do the human work better. Walk through a real example from your own business. Show them where a tool is pulling together data or trends that used to take hours of manual work, and then point out that the insight and the decision still comes from them. That's what keeps people in the conversation instead of quietly checking out.

What's a practical example of using AI to help my team without replacing them?

One example is using AI to analyze meeting transcripts or session notes to surface patterns over time. A clinician, coach, or account manager can use that output as a starting point, but they have to layer in their own context because the AI doesn't know who the most influential person in the room was, or what was said in the one-on-one before the meeting, or which priorities are actually real versus just frequently mentioned. The human reads the AI output, checks it against what they actually know, and then applies judgment. The AI saves time on the grunt work. The person still owns the insight.

What's the biggest mistake leaders make when introducing AI to their team?

Introducing AI tools with no context and expecting enthusiasm. When you tell your team to start using a new AI tool without explaining why or what it means for their role, the default assumption is that you're preparing to cut headcount. That fear causes people to quietly disengage, stick around for the paycheck, and stop investing in the work. The missing piece is always the conversation before the rollout. Explain what you believe, what problem you're solving, and how the tool fits into a bigger strategy that still needs people at the center of it to work.