Alane Boyd (00:01)
We added Claude Cowork to our automation mix, not just answering questions, but to do actual work. And it completely broke how we think about hiring because once the AI can draft the emails, update the CRM and build the reports, the thing to hire for is the thing that AI doesn't have. Listen to this episode to find out what that is.
Micah Johnson (00:27)
So there's a clear trend that we're seeing in 2026 in this first quarter, Alane and we're getting close to the end of the quarter here, but the trend is AI is moving from just answering questions and chat to actually doing the work, which is mind boggling. What's crazier is that we see it doing work better and faster and more consistent than the human counterparts.
What does this mean when we think about hiring and scaling a business?
Alane Boyd (01:00)
One of the things that I've really noticed, Micah, is how much humans love to recreate the wheel. Man, we love to revisit a document, rebuild a document. I mean, you are definitely one of those people, but we all are. But what I'm noticing is, and it has to do with how much even context our brains can hold.
Micah Johnson (01:06)
Yeah.
I'm feeling really called out right now.
Alane Boyd (01:27)
and how much memory we can hold, especially as leaders, because we've got so many things transporting through our brain at one time. Like we always feel like, okay, let's just start fresh. Well, with AI being a part of it, it's just taking the context that it has saved in like our vault, like we talked about with Obsidian and being able to reference that and go back to. So we're not recreating the wheel as much. And even when we're having it help draft new documents or new...
IA briefs or things like that, like creative briefs, it's just so much faster because it's just taking the information it has and we don't have to keep rebuilding.
Micah Johnson (02:04)
So Alane, I'm realizing something as you're saying this. I am so guilty of reinventing the wheel, but I also get frustrated when other people reinvent the wheel on my team, because I know how inefficient that is. But I make the excuse that I'm busy, I got to get this done, and I got a race to get it out. And so that's what led me to always go, it's faster if I just get this out, get this repurposed.
do the work however I can in the moment, which is a lot of the reinvention of the wheel. That has changed since we implemented Claude Cowork, because what I realized while you were just saying that is it allows me to work at such a faster clip, a faster speed, and get the results and iterate that I'm not as concerned of how do I find time to do all this? It's not.
three hours of work block that I have to figure out, it's 10 minutes.
Alane Boyd (03:04)
And because of this, we are doing huge company training with our employees so that they're also benefiting from the technology that we're using, the way that we're using it, because we want to empower them as well. But training is happening. We're not just throwing it and going, here you go
Micah Johnson (03:21)
Yes, yes. And what we're seeing, and this sounds crazy, but the tasks are disappearing.
Alane Boyd (03:30)
It is because it's just eliminating some of the details that a human used to have to do and eliminates those because they can just do it all at once. But where we are seeing where a human absolutely needs to be a part of it, is the judgment.
Micah Johnson (03:50)
For sure. For sure. So with the tasks disappearing, I love what you're seeing, Alane, on your side of the house, but what I see and what I hear and what I would say what I feel is the tasks that are disappearing are the ones that suck. They suck the life out of us. They suck the life out of our team members. I know you and I have had this conversation just one-to-one of
wow, this is actually super fun. Like we're able to iterate at the speed that we need to iterate. We can get these things done without hiring a whole separate augmented team to do it. We can prototype stuff in a flash and then we can make decisions. We can make good judgment calls, which is what we want to be doing in the first place. When we've pushed this to our team and started training our team on how to use it, we've been hearing the same thing.
Just this morning, I had a team member tell me, holy crap, I am obsessed with Claude Cowork because this makes my job so much fun.
Alane Boyd (04:57)
I'm imagining which team member said that on the team. But it is. And I think I even said that this week too, from my own perspective and maybe even in our previous podcast episode that it is becoming more fun again, because the companies that are scaling have software in place, right? We can't scale with human, everything being in human's head. So we have software in place to help us scale. But what happens is there's so much
Micah Johnson (05:00)
I'm not going to name names.
Alane Boyd (05:25)
context switching or software jumping, that it does become a little tedious and monotonous, but those things do help scale, but it does drag on us, because the people that are up here in our, I'm using my hands, but like the higher level thinkers, when we're having to do that software jumping, when we're half to digging in into the weeds, that does not make our job fun.
Micah Johnson (05:51)
Yes. Yeah. And the secondary comments that I hear or similar comments that I hear are things like, I don't have to do the shit that I don't like doing anymore. And I echo that too.
Alane Boyd (06:02)
favorite one to me.
Yeah. I mean, as a leader, that is really empowering. And it keeps you motivated because all of us. And that's what I was going to say, like one of the things that is a huge problem with organizations is motivation. How do I keep my employees motivated? And where where we're going in this episode, too, is is really talking about.
Micah Johnson (06:09)
Yeah, but these are team members telling us that.
Alane Boyd (06:28)
where you are using this AI technology, whether you're using AI agents or Claude Cowork, is where you start seeing a true separation of your team members. And that you want the ones that embrace the technology, absolutely, because that's gonna help you do things faster, better, and with more efficiency, but that can make good judgment calls.
Micah Johnson (06:40)
Yes.
Alane Boyd (06:53)
Because poor judgment in this is going to kill your organization. Doesn't matter how fast you move.
Micah Johnson (06:58)
Yeah.
Yeah. And I mean, I think we've, we've said variations of this in past episodes, but because of AI it's, you know, a multiplier. It will multiply good judgment as fast as it will multiply bad judgment. I want to share a story here that I think you'll resonate with. It was with our previous business.
Before we did a merger with a competitor, we were really buttoned up in operations. When we did the merger, we took our ops and applied it to the company that we merged with. There were many instances of team members who were absolutely adamant that these systems would be the fall of mankind. They hated every bit of it. The truth was, these were the people that hated accountability.
and the systems shown a spotlight directly on them saying, hey, you're not actually doing that much work over here. And their reaction was exactly what you'd expect for somebody being called out like I was earlier about reinventing the wheel. No. But when they're able to hide in the shadows of not having systems and still collect a paycheck, everything's great.
Alane Boyd (07:57)
Mm-hmm.
You
Micah Johnson (08:21)
when they're asked to be accountable, everything's not great. I share all of that to say we're seeing similar things happen across teams with AI. When we implement AI, it's shining a spotlight on judgment, not just accountability. And those who have good judgment are like, hell yeah, this is awesome. This is what I've been, I'm having so much fun. Let me make more good judgment. And those who have poor judgment are like,
Alane Boyd (08:36)
Mm-hmm.
Micah Johnson (08:51)
Mm, I don't think this can work as good as humans or this isn't going to work or here's all the and you get that adamant pushback.
Alane Boyd (09:02)
And that's where I think at this point, a topic is like hire slow and fire fast is hiring for good judgment can be difficult in the hiring process, but recognizing it fast because a person isn't going to change how well or badly they have judgment. They're not gonna be able to change that. So in any small amount of time. So having good judgment, if you're recognizing those things early in,
Micah Johnson (09:22)
Yes.
Alane Boyd (09:29)
the process is they need to be out of the organization. It is hard, but that is going to severely hurt you at the momentum that things are happening with using AI in organizations.
Micah Johnson (09:43)
Yeah, I mean, I couldn't agree more with that. It's crazy just thinking about how much has changed in the last 90 days, whether it's how does this apply with AI? How does this look at from how you operate and run a business? And now to even how you're hiring.
Alane Boyd (10:05)
And even just looking at ourselves and our team, and we're making changes on our team, we have 41 agents who work for us now. Like the last time I mentioned it, we had 35. Like we are increasingly adding agents to do stuff and using Claude Cowork and empowering our team. And where we're looking at our team members is, do they have ideas? Has this team member had a good idea in the last six months?
Or whatever period of time, because that's who we need. We need the idea people. And then have they had any good judgment calls that we can associate with them, because those are the team members that we want to keep as we scale with agents.
Micah Johnson (10:47)
Yeah, yeah, absolutely. And we see this across companies that we work with. We see this internally that you gotta coach the poor judgment calls. You've gotta reinforce the great judgment calls. And the more that you have people on your team that make great judgment calls, the faster and better everything works more seamlessly. And when you're combining that with the tools like Cowork and n8n agents, for example.
It is off to the races. So we've got three to really get down to brass tacks here, Alane. We've got three core elements that we've identified that work really well for when you're thinking about hiring for judgment. And the first one is being able to frame the problem. This is not just prompting into ChatGPT or Claude. That's one piece of it, but being able to A, understand the problem B,
being able to say what the problem is and explain it in text and C, actually provide the right amount of context to build the entire frame of what the problem is.
Alane Boyd (12:00)
And that's a really, I'm thinking about some situations where they can't frame the situation and it's chaotic. It is just all this extra information and we're not really getting to the root. I mean, part of our hiring process, our hiring process is pretty, I don't wanna say grueling, but that's the only word that's coming to me right now. It's, we have scenarios that we put.
Micah Johnson (12:24)
So if you're looking for a job, yeah, no.
Alane Boyd (12:28)
But we have scenarios that we have the team members go through as a part of the hiring process. And we say like, hey, as a part of your second interview, please run through this situation and tell us how you would handle it. Give us examples, things like that. Cause we're trying to identify those people in the hiring process before we hire them. It's not perfect, but it absolutely does help putting that upfront. So we're not six months later down the road trying to figure that out. So we have...
I think a five part interview process with the actual interview calls plus the scenarios that they go through. And we have them record a Loom to hear what their thinking is. So we're getting that feedback within the scenario too. And it's not just written. Like they can't just go use AI to try to answer it. Like we really need to hear their thoughts on
Micah Johnson (13:16)
Yeah, and this made sense even before AI came into play. It's even more important with AI and tools like Claude Cowork being in play, because if you can't frame the problem, you have no place to start. You're already dead in the water. And this used to be like, when people record those Looms, like you're talking about, Alane, we could almost predict like,
without even watching the Looms. And it's by the length of the Loom. If it was an hour long Loom, and no joke, we have gotten hour long Looms, then probably not a great fit. If it takes you an hour to frame a problem, we're gonna, yeah, we're gonna pass on that. If it's a five minute Loom and we watch it and it's super succinct, it's amazing.
Alane Boyd (13:46)
Right.
Yeah, perfect.
Yeah, when I was thinking about that, Micah, I also wanted to just for context for our audience, like the entire scenario with the Loom video should take that person 30 minutes to complete. Like they're not meant to be these long four day solutions to get back to us and long explanation. Like the whole thing is meant to be figure this out, give us the answer, give us the Loom. The whole thing should take about 30 minutes of time.
Micah Johnson (14:31)
Yeah. Can you perform
quickly, efficiently? Can you explain, you know, we do all these podcasts in like 15, 20 minutes. Like imagine if it took us four days to put together this podcast. We would never be able to execute on this.
All right, so the first one was framing the problem. The second thing you're looking for is being able to evaluate the output. And again, this was important before AI. You needed to look at what somebody was producing or what you were producing yourself and look at it and go, cool, yeah, this works, this doesn't work, I should make some changes. The people that produce something look at it and go, I have no idea.
If this is actually good or not, I got to go ask my manager. That slows everything down. It slows everything down before AI. It dramatically worsens if you apply a person like that and give them the power of Claude Cowork. And Claude is producing documents and PDFs and slide decks and spreadsheets and any type of output to do the work. And the person can look at it and go,
I have no idea if this is good.
Alane Boyd (15:48)
Or worse, they give it to the team without looking at it and then it's even more.
Micah Johnson (15:52)
And you know what? Honestly,
this happens on a team member level and it happens on a leadership level.
Alane Boyd (15:58)
Absolutely,
Micah Johnson (15:59)
Alright, so the important thing here is when hiring, you have to have people that can look at an output, whether it's human or AI, and again, make good judgment on it. If they have to refer to somebody else or like you were saying, Alane, if they just believe whatever is coming out of AI or whatever another person produces, then you're in for a world of hurt. So the third thing and final thing,
is being able to have orchestration thinking. And this is, can they look at a problem that's been framed correctly, even get some output and go, this is actually what we need to do next. And these are all judgment calls. Can we make good problem framing? Can we evaluate the output correctly? Can I make decisions on what needs to happen next? Because with AI,
I could actually orchestrate that across multiple agents, across multiple co-works. I could instruct Claude Cowork to create sub-agents to get this work done. And if you can't think in that way, then really, what are we doing here?
Alane Boyd (17:12)
I mean, this is a hard part for humans because how often do we hear people say, well, that's the process instead of why are we doing things this way? Let's improve it. Getting people out of that habit is hard. So you want the thinkers that are, you know, I want to improve it because that is where all of this is going. We even, you know, just looking at AI basics, that's how we're looking at things. But that human part of this is realizing
what needs to go through AI, where's the process improvement and what can be removed from the process because AI can do it now and be comfortable with that. Not AI has taken my job, not, I don't want anybody to know what I'm doing because then that takes away my value. No, you want the people that are seeing it and embracing it.
Micah Johnson (18:01)
Yeah. I mean, define another way. This is inherently systems thinking. I'm looking at this problem. I can make good judgment calls on the output from it as I'm working through this, but I can also think about this in a process, like you're saying, in a system. I can orchestrate that. That's who you need to hire for.
Alane Boyd (18:20)
I mean, I know from my team, and I think I've mentioned this, you can have personality types that can do the job, but have bad judgment. So I'm saying a little bit contrary to that is like, my team is built up of an architect personality. Most of my team that works directly under me, and I'm saying that because I am a little bit up in the sky. I don't like being in the weeds at all. So I need to be around builders.
but they have to have good judgment and be able to do this orchestrated thinking or else we're constantly just doing the same things over and over again without changing and being adaptable. So my team members that are architects also have great judgment. So they're constantly making things better and improving what we're doing as we learn.
Micah Johnson (18:48)
Mm-hmm.
I mean, so what you're saying, Alane, I'm going to paraphrase back to you is you've been hiring for these three things before AI was even.
Alane Boyd (19:18)
I mean, I'm trying. I don't always hit a home run, you know, it's what I've tried to do with my team because these things are so important. But with AI, it is ripping that bandaid off an organization.
Micah Johnson (19:31)
Yeah,
you have to. If you're employing AI, you have to hire for these things. And you have to stop hiring for the opposite, the task doers. If you're only hiring people to do the time consuming tasks, to produce reports, to produce files, to move data between systems, that's not a position that is going to be a viable position now. Like it's not in the future. It is now.
The technology is here, the technology is working. And within the last 90 days, we've had a huge step forward. Within the next 90 days, every single major platform is pursuing Claude Cowork. OpenAI just announced it. They want to do a powerhouse desktop app. You know what that is? A competitor to Claude Cowork. Gemini is being integrated into everything even more. Why? So it can do the tasks.
Claude and Microsoft just partnered and announced, they didn't just partner, they just announced their partnership on Microsoft Copilot Cowork, which is the Claude Cowork for the Microsoft platform. So that covers every single area and base and platform and anything you could think of where, and I mean, hell we had this conversation yesterday in one of the mastermind groups.
which is essentially, it's not even about being competitive anymore. It is about setting the baseline and the standard for work as a whole. And clients and customers are gonna come to expect that baseline, you know, that baseline is rising as more and more people are using AI, the ability to produce stuff faster, to respond quicker.
Alane Boyd (21:18)
Mm-hmm.
Micah Johnson (21:21)
to be more accurate, to have more diagrams, to have clear documentation, to have clear agendas, whatever it might be on your operations, that baseline is rising. And you're either above that baseline and impressing your customers, or you're below that baseline and you're frustrating your customers and your clients, and that's when you're gonna lose them.
Alane Boyd (21:31)
Yeah.
Nobody's writing a letter with a contract and mailing it and expecting that to be the baseline. Like, no, it's email docuSign, Panda doc. Like technology has already been doing these things and is disruptive because I haven't sent a fax in 10 years. I haven't mailed a contract in I don't know how long, ever probably.
Micah Johnson (22:06)
Yeah, and so to that point, that's exactly where this change is happening. It's just so much faster. The next 90 days, we're going to see these releases of these other platforms. They're going to get stronger. They're even all in beta right now, essentially. So that means these aren't even the full releases.
Alane Boyd (22:13)
Yeah.
Micah Johnson (22:27)
There's so much more that can happen. There's so much more that can come out that if you are hiring right now, and we do have this conversation with a lot of people, especially in the mastermind groups that we run Alane, which is if you're hiring right now, you have to understand this technology today. You have to understand what it can do and what it means because if you hire somebody today to do any of these things that do not fall under that good judgment outline that we did and is more task centric, you're shooting yourself in the foot.
You're going to be below the baseline and you're paying somebody to do this work that could be done in seconds or in minutes. maybe they're not even if they're the task doer and they don't have good judgment, you've shot yourself in both feet at that point. And ultimately, to stay at the baseline and just tread water, you're probably going to have to repurpose that, retrain them for good judgment or ultimately let them go.
Alane Boyd (23:25)
And if you're listening to this podcast, like you're already thinking ahead because we're talking about this, we're having a conversation about this, but your team has to be trained. Your team has to know what your place is, what your vision is with AI, and then train them on how to use Claude Cowork or Copilot Cowork, it's a tongue twister. So they have to be trained and it's ongoing training. We weren't training for this.
Micah Johnson (23:46)
Wow, that one's so hard to say.
Alane Boyd (23:54)
in Q4 of last year. So we're constantly training our team on AI Agents, Claude Cowork, and your organization needs to look the same way.
Micah Johnson (24:03)
All right, so what this all boils down to, to kind of wrap up this episode, Alane, is we're saying stop hiring people to do tasks. And we're saying start hiring people that can make decisions that AI can't.
Alane Boyd (24:19)
And this reminded me, Micah, that we have the free job toolkit on our website as a resource, and we can link to it in the show notes. Our job toolkit creates a job description, but also for that role, a multi-interview process and questions to really dig into how this person thinks and if they can do the job. And then also creates sample scenario ideas so that if you like that idea of...
getting them through the interview process and seeing how they think it produces that. So we'll link to that. And then we also have a Claude Cowork workshop coming up on April 16th. And we can drop that into the show notes as well where we're teaching how to build skills, how to use Cowork, how to use connectors and plugins so that you're really giving your team or yourself the tools to know how to automate this work.
