Down the Rabbit Hole: Why AI Is Stealing Solo Recruiters' Time (And How to Take It Back) Episode 4, Season 2

The rabbit hole is real — and if you're a solo recruiter, you've probably already fallen down one.

In this episode, Kat and Nick get honest about one of the less-talked-about side effects of AI adoption: the hours, days, and sometimes weeks that quietly disappear when curiosity meets a shiny new tool. They dig into why solos are disproportionately at risk, and why that entrepreneurial brain that makes you great at your job is also the one that just cost you three days of BD.

Nick shares his system for staying focused — building well-contextualised AI projects that filter the noise before he goes near a demo. Kat adds her take: the rabbit hole isn't always the enemy, but you'd better go in with guardrails.

They also take an honest swipe at the LinkedIn hype cycle and the age-old truth that a great sales team and a great product are not the same thing.

If you've ever started exploring a tool on a Monday and surfaced on Thursday wondering where your week went — your wake-up call just dropped. 🐇

Like our page and connect with us!

LinkedIn: linkedin.com/company/re-boot-ai-and-recruitment/

Nick: https://www.linkedin.com/in/ai-agents-for-recruitment/

Kat: https://www.linkedin.com/in/katkingshott/

Transcript

SPEAKER_01 0:00

It has been a time through technical difficulties uh and other difficulties we've we've managed to make this conversation happen again.

SPEAKER_00 0:08

Yes. Yep. So what we're gonna talk about today, we should talk about rabbit holes.

SPEAKER_01 0:16

Mm-hmm.

SPEAKER_00 0:17

Not literally, but figuratively, and how AI is creating huge rabbit holes for solos in particular, and what the impacts of that might be on them and their recruiting capacity. Um, is it a hindrance or is it a benefit?

SPEAKER_01 0:42

It's a good question. It's a good good way to think about it. I have had quite a lot of experience with rabbit holes recently and also hearing about other adventures. This is a terrible analogy, but we're just gonna keep going with it for myself. I am guilty of spending way too much time talking mainly to Claude. I I am trying to see other people, but Claude and I like we're going pretty deep. Like it's pretty serious. But I think what comes with that is you can't really often until you try something, you don't know how good it is.

SPEAKER_00 1:25

Yeah. Are you using like cowork called co-work?

SPEAKER_01 1:32

And um once once I could actually whatever the thing it asked you to download, the GitHub thingo, that took me like a week to get. So yes, now I am, but I nearly threw my laptop out of the window. But yes, and it and then then the first time I did something when it was like uh I think uh I asked it to yeah create a couple of docs and come back, I was literally like the smiley. I wish you could kind of bottle the feeling of like when you get your AI to do something for you and it does it even better than you thought it could, and you're just like yeah, and then you go, I wonder what else it could do.

SPEAKER_00 2:13

And then that's where the rabbit hole starts, isn't it?

SPEAKER_01 2:16

Yeah, I think I think you need to have, yeah, I think it's also a distinction, and I think it's it's very reasonable that it's just the solos, particularly, that are feeling this is because by character and by nature, typically um, you know, you're more creative, you're more just entrepreneurial, you although you have like a recipe in a way that you like to work, you are looking for new things. So I think because our brains are that way inclined, you know. I talk to some people and they then they're not trying this stuff out, and they probably don't own their own recruitment business that started in the last like two to five years.

SPEAKER_00 2:59

But also employees, we've spoken about this in another another pod, but employees usually get what they are um effectively given, right? Um, so if the owner of the business doesn't invest in you know fraud and says, Hey Kat, you can use co-work, right? And it's because it's not going to be that productive for them, uh, they won't. Um whereas solos they can invest in that productivity, or at least explore that productivity. Uh let me reframe, or at least explore the technology to see if it's more productive, right? And that and the problem is, I think, with again, we're talking about Claude co-work, but even if you use like open claw, that's a massive thing at the moment, right? Or one of the other millions of products that are out there, um, they've actually existed for quite a long time. They've just been repackaged. So when I was going down rabbit holes, I was using MCPs with Claude, and MCP is what Claude now calls a connector, really. So you know when like you go into your Claude and you've got that little plus button, yeah, and then it's got connections, and then you can connect to like you know, super base or whatever.

SPEAKER_01 4:16

Yeah, it was like vibe, like a year ago, it was we we were vibe coders. Well, I I wasn't, but I I knew some vibe coders.

SPEAKER_00 4:24

I mean, I think vibe coding now as a separate subject together. Like, I suppose technically I'm a vibe coder, technically, although I don't use lovable, I don't use all that stuff I I I build within my local environment using claw code. But um for things like for the for the average recruiter, well, not even the average recruiter, for like your solo, if you get told that you can use clawed co-work and you can, you know, now it can access your documents and create documents and save documents, and it can access your emails. I mean, it has actually been able to do that for a long time anyway, but now it's more accessible, you go down these mad ass rabbit holes, and then the issue is because they don't understand sometimes the fundamentals of how uh like AI and LLMs work, they're not prompting them correctly, for example, right?

SPEAKER_01 5:18

Or they're yeah, or they're accepting information that they shouldn't. Yeah.

SPEAKER_00 5:24

Yeah. And then once that happens, and like you with your example there of you know it does jobs better, right? Then the next step there is I wonder what else it can do. And then before you know it, those you know, that BD stuff that you had to do, because that's boring as shit, right? You know, where you've where you're looking at intent data and all that stuff, you start thinking, can I can I get claud code to do this, or can I get you know, open claw to this? And then you're investing days and days and days trying to get that to work, but you can't get it to work because you don't know necessarily how to, and then you've wasted days and days. You know what I mean? Hence the rabbit hole, right?

SPEAKER_01 6:00

Um Yeah, I want I want to be able to phone a friend, like uh not a literal friend. I guess I want to phone, can you phone an AI friend in that moment? Because it don't you just then just aren't you then just meant to go to like Claude or another LLM and be like, I'm stuck, what do I do next?

SPEAKER_00 6:17

Yeah, for sure, but again, you need to but then you need to provide your project within Claude, you need to provide it with knowledge. So like one of one of the things that I I did because again with MC, I was doing MCPs like two years ago, it was new. I had no idea. Really had no idea. There's hardly any information out there, really. So um I just did loads of deep research. I I didn't, um, but GBT did, perplexity did, and Claude did. I asked the same thing for all three LLMs, do deep research on MCPs, how to use it, blah blah blah blah blah blah. It created a whole document and gave that to Claude and said, turn that into a universal document, upload that into the project, and then I can generally ask Claude's information and it'll have a really good reference point to everything. So you can you can use the same principles with cowork, right? You you can just uh train up a Claude project who only focuses on co-work and how it works or what you can do. Um and then you can use that as your you know, first line, second line support. Uh, how the fuck does this work? What do I do here? What do I click here? And then it will tell you step by step on how to do things. So that will still get you down the rabbit hole, but it will get you out of it quicker. Does that make sense?

SPEAKER_01 7:37

So the the lesson um could be before you go down the rabbit hole, have like a like a pre-check or uh evaluation of like uh what you might need with you. So for instance, even just setting up that project um specifically for co-work is good before you then start.

SPEAKER_00 8:00

Sort of, sort of. So basically, I I let me just tell you what my process is, right? Um and then that might provide easier context. So I've got loads of projects, but there's actually only really three that I use consistently. One's called the PM, one's a marketing director, another one's a note director. The PM knows everything about me. It knows exactly what I'm building, it knows the tech stack, it knows everything, right? So when about uh six months ago, when I heard about OpenClaw, um, which if you don't know what that is, it's basically you get loads of LLMs to do loads of shit for you. Uh and there's a lot of risks to it, but people are just all jumping on this like crazy. My perception was um I think it's gonna cost a lot with LLM calls, it's gonna be high risk security-wise, and it's gonna completely neglect things like GDPR and stuff.

SPEAKER_01 9:01

Yeah, that's right, yeah. I I I asked from last week. Yeah.

SPEAKER_00 9:05

Yeah. I asked my PM this. I said, I want you to do some deep research on OpenClaw. Um conclude. You know what I do? Conclude. I didn't tell it what I thought. It did some research for like 25 minutes, scraping git counts.

SPEAKER_01 9:20

What does the word conclude mean? I mean, I know what it means, but in that context.

SPEAKER_00 9:26

It will get loads of information, it will look at what I'm doing, and then conclude whether I should invest my time in doing it.

SPEAKER_01 9:32

Oh, okay. Yeah, all right.

SPEAKER_00 9:34

And then and then it concluded and said, I'm not sure, these are the risks. Funnily enough, it's GDPR, security, LM costs, what you're building is database centric, don't think you need to do it. So I've gone to it and said, This is what I think. Well, I hadn't told her what I think, but I've asked her to go and do research. So whenever a new court tool comes in that I think to myself, this could be interesting, I'll have my idea of it, but then I'll always want to speak to Claude about it, who can then do research and then come back. And if he said if they say, um, I think that actually this could be really useful, then I'll invest some of my time and actually look at how I can implement this, or what I could implement. So that because I used to go down rabbit holes all the time. We joke about how I got ADHD. Um, I I get very fucking hyper-focused on stuff. And if uh if a new if a new shiny thing comes out, a while ago, I used to be all over that shiny thing, and then I realize I wasted a month because it's shit. Now I don't focus on any new tech whatsoever. I really don't. Uh anything new that comes in that might be of interest, if it passes my initial preconception of what it might be, uh I speak to Claude about it. Claude gives me the answer, then I decide I'll set up a new project which is just about co-working. So with co-work, I don't bother with co-work because I was doing co-work a while ago with the MCP stuff.

SPEAKER_01 11:04

Uh and it's got no is really good, but it's got no impact in my actual but you're you're to be fair, you're probably a bit more of an advanced user of it.

SPEAKER_00 11:15

Because yeah, but it's the same principles still though. Right? It's the same principles. So if I'm a recruiter, if I'm a solo, and I've got a project, and and and this project, this called project knows what I do, and I've even told it, listen, I like shiny things. I need you to to to help me strip this away so I don't focus on shit, right? This is what I want to get out of you know using AI. Then when Open Core comes out or Court Co-work comes out, I can just ask that same project, the same thing, but the context is different. For me, the context is I'm building a platform or platform, blah de blah de blah, right? But for the recruiter, it's I'm a special specialist in X, Y, and Z. I really hate doing this, and I hate doing that, but I like doing that. Yeah, will this tool help me? Right? It's again it's like the context.

SPEAKER_01 12:09

Yeah, I think that that's I'm gonna add that into mine for sure. It's accountability uh practice, really.

SPEAKER_00 12:19

Yeah, I used to spend way too much time, man. Way too much time, and my productivity suffered as a result, which I think is what a lot of solos are going through. There's a guy who is um he was a member. I'm probably gonna kick him off now. Um the the hub because he doesn't he doesn't need it. Um he he hasn't used it for a while. Not because it's not good, it's because he's getting overwhelmed still with everything, with AI. Yeah, he's going down rabbit holes and he ain't picking up the phone and fucking calling people. Yeah, he cannot be the only person.

SPEAKER_01 12:52

Well, I would you know, as much as I love uh as much as I love chatting to Claude, I actually enjoy talking to people humans more. So I yeah, I get it too.

SPEAKER_00 13:03

Yeah. So just just I think that all you gotta do is again, we're talking about Claude, you can do this with GPT or you know, Grok or Gemini or whatever. It's the same principle. Make sure that you have a project and they just know what you do. Not talking about your personal life, just talking about like they know what you do, they know what your objectives are, and they know what you like and what you don't like. Then when you get a new tool, because this is the thing about LinkedIn, isn't it? Right? It's like this tool is gonna destroy Figma, this tool is gonna do this, and then it makes people go like, whoa, oh my god, I better I better invest some of my time into this and have a look at it, and then they realize it's dog shit. Um or it's actually not gonna benefit them in their specific U case. Right?

SPEAKER_01 13:46

Yeah, so it can be good, yeah, good, good to use, yeah, maybe as a reflection point after, you know, if you do find yourself on a demo or yeah, I mean before you even go on the demo, don't even go on a demo.

SPEAKER_00 13:58

Just just just get clawed to do some research on it. Right? Yeah, and then you're saying there, just you know, I'm just about I I read this amazing post um uh this morning, and I loved it. It said um we invested loads of money in juice box and we hate it, it's shit. It's a glorified Boolean string.

SPEAKER_01 14:21

I saw that one.

SPEAKER_00 14:23

Did you see that? I loved it. I loved it. I was like, finally, finally, somebody understands.

SPEAKER_01 14:32

But you know what? Different strokes for different folks, yeah.

SPEAKER_00 14:39

Maybe, but it is right.

SPEAKER_01 14:47

I'm trying to think of a good analogy here, but not everyone wants is picking up what what what you put down.

SPEAKER_00 14:56

100%, yeah, 100%. But I think the reason why I liked it is not because you absolutely slated a company that I've been suggesting people shouldn't be using in the first place, is because um it's another example of this revolutionary tool that's gonna change the way you work, but all it's doing is redistributing your your time.

SPEAKER_01 15:20

But it's an example, and and maybe not to to pick on a particular business or brand that's just raised a boatload of money. Um, but to say, yeah, but it's also that's it, like the popular, the popular, shiny, the popular shiny thing, at least on the exterior wins, I think so many of the tools, so many, so many of the tools that people are using in recruitment are because recruiters like to be sold to by good salespeople, not bad salespeople. So if you've got a shitty tool but a good sales team, we know we know who we're thinking about.

SPEAKER_00 15:57

So on that note, um but I think overall, to conclude from my perspective, I think that um solos need to be a little bit careful, they need to utilize AI absolutely 100%, but they need to be mindful of these rabbit holes, they are real, they really are, and you you spend so much time on this stuff that it isn't relevant. Um, I've been there and I've I've experienced that a lot. Um, but you can mitigate that risk uh by setting up projects right on Claude, GBT, whatever, um, that can evaluate and help you see it as an employee um sort of thing. Uh and and I think that that's gonna continuously happen, it's gonna get worse and worse, and then loads of solos are gonna be like, oh shit, now what? Because I haven't got jobs in or whatever else. Right. Um, and then it's gonna shake things up unless they get a hold of it. That's what I personally think. Um but I might be completely wrong as well.

SPEAKER_01 17:07

Yep, my takeaway would be it's usually not the silver bullet you think it is, or that it's it's marketed as, but it's I think it's I think it's really important as a solo to take time away from the day-to-day grind of whatever you're doing to be playful and creative and think bigger picture or think about a new way of doing something. And I think that that's when the rabbit hole can then produce something really hopefully really exciting if it's set up correctly, particularly if you have those guardrails where you're telling it, hey, look, don't let it, don't, don't let me go off on these complete tangents.

SPEAKER_00 17:47

Yeah.

SPEAKER_01 17:48

Yeah.

SPEAKER_00 17:49

Yeah. Cool. Nice. Good. All right. Well, until next week or the week after or next year, whenever we do the next one.

SPEAKER_01 17:56

Yeah, whenever I'll probably see you, yeah, at some point. Yeah. Nick.

SPEAKER_00 18:00

All right, ciao, ciao everyone.

SPEAKER_01 18:02

Bye.