Subscribe
Share
Share
Embed
The built hapily podcast is about building apps, companies, and relationships in the HubSpot ecosystem. As HubSpot grows, so does the opportunity - and this podcast puts you in the room with the people making it all happen.
Hosted by Dax and Max, built hapily goes behind the scenes with HubSpot developers, solutions partners, startup founders and community leaders. Each episode delivers tactical insights into launching and scaling businesses around the HubSpot platform.
However, this podcast is about more than just building software. It's about building authentic connections, fulfilling careers, and lives you can be proud of. Guests share their personal journeys, hard-won lessons and philosophies for not just achieving success, but finding purpose and happiness along the way.
After all, this is about more than making apps. It's about building hapily - and you're invited along for the ride. Join Dax, Max and their guests to construct the life you've been dreaming of, one conversation at a time.
bhp S2E03 - Nico Lafakis, HHH - 16x9 - Full
===
Nico Lafakis, HHH: [00:00:00] In this world's economy, most people need two jobs to facilitate. So why not? Now you've got two employees for the price of one when it comes to a business. So
Max Cohen, hapily: The question is, did it take that job away from someone else?
Dax Miller, hapily: Yeah, I'm going to take it regardless. I'm going to take two if I didn't even have an agent because I'm crazy and I did take it away from somebody.
Nico Lafakis, HHH: Okay, but if that's true, why are there so many? How come I can go on Fiverr and find so many HubSpot jobs?
Dax Miller, hapily: On today's episode of the built hapily podcast, we have Nico LaFakis, the creator of Hub Helper Harry.
Niko shares the story with us on how he built Hub Helper Harry, why you should be using it as a HubSpot admin and gives us some pretty deep insight on what he sees coming to the HubSpot universe as it pertains to AI. [00:01:00] All of that, and more on this episode of the built hapily podcast.
We got Nico in the building, Lafakis
Nico Lafakis, HHH: Lafakis
Dax Miller, hapily: Galifianakis, not, no
Nico Lafakis, HHH: It's the same. Actually, you can go either way.
Dax Miller, hapily: but I attribute you Nico to the forefront of AI in the HubSpot ecosystem. I don't know anybody that was there before. I don't know anybody that is there after, and you left your flag in the dirt with Hub Helper Harry. So, Nico, welcome to the built hapily podcast. Man, we're super hyped to talk to you today. About all the things. That two letter word that has been floating around the universe for quite some time now.
Max Cohen, hapily: Hold on. Can I just Nico, can I just give you your flowers for one second, dude? Like I, I thought I knew a lot about HubSpot, right? And then I realized how much I didn't know about things like CMS modules and all this stuff until one day, I was just like, "Man, I really need to build some modules."
You know and ChatGPT by itself doesn't seem to really understand a lot of the [00:02:00] nuances of this stuff and then I found out through you somehow, right? Hub Helper Harry and that was actually like my very first experience with a custom GPT By the way. Your tool that you built, dude, and the process of using it... And it's for anyone who's unfamiliar with this story I've talked about it a lot, but I built a bunch of CMS modules on HubSpot just by using Hub Helper Harry to explain it to me and write the code and give me everything I needed and I just want to make this like clear to everybody.
I am not a coder. I am not a coder. I do not write code. I don't know how to write code, right? I understand the basics of HTML and CSS and like what it means. I can't write it for shit, right? I was able to build custom CMS modules in HubSpot just by talking to this bot that Nico created. Okay. And I think what you built is probably one of the most like monumental things.
It's going to be like one [00:03:00] of the most impactful things to HubSpot admins, especially in this age of AI, that if there are things you think you couldn't do in HubSpot, now you can, right? The amount that I learned about things like HubL and like how CMS modules work and like all these other concepts, I wasn't familiar with it just because it was able to explain it to me because my favorite thing to do with Hub Helper Harry is be like, tell me how to do this, but explain each step like I'm a five year old, right?
And it easily walks me through everything. Dude, the stuff that I was able to build from that has is literally, reinvigorated, a lot of my like love for some of the scarier parts of HubSpot that I've stayed away from. Cause I'm not a coder. And so I just want to say, thank you, dude.
It is like one of the coolest tools ever, but maybe boy, we can do is you can explain it in your own words for anyone who doesn't know what Hub Helper Harry is and give us the story there, but obviously tell us about yourself first, who you are.
Nico Lafakis, HHH: Well,
man. I, well, I guess to start off. I'm [00:04:00] just tremendously beyond tremendously humbled, because I just do stuff is the way that I describe it. But, yeah, oddly enough, my background it's super varied. So, I was doing graphic design for the longest time. My degree is actually in 3D arts and animation.
So, even the career job didn't end up being what the degree was, so for everybody that's out there is wondering about that. They don't always have to align. I consider marketing to be this sort of like car production process. So I look at graphic design as being the team that sort of builds the outside of the car and makes it look good and puts all the, paperwork and all advertising and everything together for it.
And then was doing RevOps for a couple of years and, still doing that as well. And, that to me is like working on the engine, right? That's the heart of the CRM and how everything functions. And, now definitely doing way more with technical solutions and with AI. And to me, that's definitely the electrical systems. It's like tying [00:05:00] all of the parts together and making everything work together seamlessly. So, that trip has been insane because there's been a lot of jobs in between. There's been a lot of stuff in between. Life happens. but you end up here and yeah, to Dax's point, I've just wanted since the beginning and knowing what this technology can do, all I've really been waiting to do is just disrupt as hardcore as possible and to use this technology in the way that it was meant to be used in order to do that. So, if you follow this stuff closely enough, one of my favorites, I would say as Mustafa Suleyman, who works and headed up the team for Pi, and now works with Microsoft and he gave a talk probably eight months ago and was talking about how, what this technology does and the profound aspect that it brings to humanity is that it makes knowledge free for everybody. High Super high level knowledge, [00:06:00] free to everyone. Right? So, to your point in terms of you can go on and learn about things without necessarily the need for a guide, without necessarily the need for a course, without necessarily the need for a tutor or teacher, you have that, right?
And being able to do that is really what pushed me to create this in order to satisfy the, what I kept hearing from people, which when you get into this space and not just this ecosystem. From what I can see, it's like just the SAAS ecosystem in general, I think that's really where the imposter syndrome started, because to me, it's an entire field that is degreeless.
Max Cohen, hapily: Yeah
Nico Lafakis, HHH: You don't go to college to learn how to work in SAAS. You just use the platforms. You just work with the people. You just work at the places. So it really sucks because a lot of these places move at a lightning pace and they have zero onboarding. Or even if they do have onboarding, you have zero time to do it.
[00:07:00] Right? So you're just left out in the wind.
Max Cohen, hapily: I think the other thing that's interesting too about it is like I mentioned it when I said explain it like i'm five. It's not only is this information there, and you can learn about anything you want, you can learn about it in the way that works for you, in a way that you're going to be able to consume it.
Obviously with all like the political stuff going on and people talking about budgeting and this and that and the other thing, I didn't know what deficit spending was. Right? And I literally had a conversation with ChatgPT one day I was like, can you explain what deficit spending is to me like I am a five year old, right? And I keep using that example, but like you could just as easily say, explain it to me in a story, explain it to me with different examples and bullet points, or go one by one through this concept or whatever it is, because sure information's out there. You can Google search and find things, but it's not necessarily going to be presented in a way that is easy for you to consume and understand the concepts. I think one of the coolest things about all this information being available to people is you can get it [00:08:00] explained to you in a way that's going to resonate with you and makes sense given the way that you like to learn.
And I think that's the kind of more like nuanced, like powerful thing about that, that is really useful, especially when people like really understand how to prompt and talk to these things differently and it becomes more widespread,
Dax Miller, hapily: Yeah, it's the
two way, It's the two way, man. Everything we, the way we talk about it, or the way I talk about it is... Everything's an extrapolation, up to speed. Originally if I had to do a book report, you gotta know the Dewey Decimal System. You gotta go look it up. Before that, you had to go to the, there was no library.
You had to go to someone that remembered it. And someone wrote it down at their house, and they have their own manuscript of it. Then we have, we got Google. Cool. Then we had Google on your computer. You're good. You can find your stuff. You had Bing, you had all the funny search things that it would throw out.
But then you had it in your pocket. Then once you have it in your pocket, now it's pretty much, you have the knowledge all the time, everywhere. But now, you have a two way, where I can get the knowledge, and I can build my own university. Right. It's like you used to be able to do that with YouTube. I'll make playlists.
I've got college. Like there's that you can get into SAAS by making a bunch of [00:09:00] playlists and you're good, right? As long as you've got your hands dirty, you are going to be eligible to get the problems fixed that get thrown at you because no one's going to know how to solve those problems. There's just going to throw a match and be like, hey, man, piece together your life and try to solve this problem today because we put you in this position. But with AI and the way that you're able to talk and build, like I can build my own language tutor and talk to it and get the actual you to have a, this accent or that accent or use this and then I can build my own language model, literally, to talk to me and teach me and keep me ready to hearing different things in different language rather than me clicking Duolingo, right? It's just ridiculous the way that works out and how to bring that back to where it is. That was not in the HubSpot ecosystem. To me I don't know if this is fact or not, but I'm going to stamp it as fact. The AI, first interaction with AI in the HubSpot ecosystem was Triple H. Period. To the point where there was, I forgot even the original name, wasn't it HubSpot Helper or HubSpot Harry and it got shut down.
Nico Lafakis, HHH: Yeah, it was HubSpot Harry, but...
Dax Miller, hapily: They shut that down. Everybody's got their [00:10:00] naming story.
Nico Lafakis, HHH: It took about two weeks.
Max Cohen, hapily: Yeah, Yeah, so Nico like I want everyone to know about this tool and I think you're probably the best one to like really explain it to people. Everyone's like super obsessed about Breeze these days, right? But, Hubspot Harry is where it's at, to be honest with you.
Dax Miller, hapily: There's HubSpot, there's Helper Harry. Yo, this is real talk. There is Triple H code base live in effect right now, being used by lots of people that was written by Triple H.
Max Cohen, hapily: I get that. Sure. There's people out there using ChatGPT to do HubSpot stuff, right? But I think there is a bedrock fundamental difference of using just naked ChatGPT versus using HubHelper Harry to help with your HubSpot stuff.
Can you explain the work that went into it and like the benefit of using this custom GPT when you're trying to help assist your HubSpot work versus just using like naked ChatGPT?
Nico Lafakis, HHH: Yeah, no, there is, there's a huge reason. And so the work that went into it is [00:11:00] literally the entire knowledge base, the entire API documentation, the entire CSM documentation, and the entire interface, documentation, so pretty much every type of documentation you can think of.
And then all the Deep Research, the top 100 knowledge base posts that have been made in the last 6 months that have answered the most pertinent questions. So, if you're wondering what $200 worth of these deep research is for, it's insanely invaluable. So Harry's also filled with a lot in the last week, a lot of marketing research data, that's been done, about 300 pages worth now, almost and increasing, covering just a wide array of topics. I'll actually post an update to that later. So, on top of that, he's also got a lot of the freely available academy information. So, what it is an amalgamation of is, knowledge base, API, all of that. [00:12:00] Plus, some stuff from the community where there might be some hardcore answers. And what all of that allows you to do is essentially use HubSpot as the base for everything. So instead of talking to a ChatGPT focused bot, and then having to give it the context that you're working in HubSpot, give it the context of the project you're working on, give it the context of how you work with fields and what objects are and what properties belong to objects. You could go to Harry and you could say something like, "Hey, how's it going?" And then say, " everything there is to know about HubSpot objects and properties for them, right?" And he'll tell you yeah, of course." And then you can say, " Cool. Can you make me a downloadable CSV file of dummy data for contacts and include the most pertinent information that I would need if I was doing like a demo on cross referencing data for insights."
Max Cohen, hapily: And it's going to know what that means.
Nico Lafakis, HHH: It knows exactly what it means.
Max Cohen, hapily: Yeah. That's the thing. Like it's ChatGPT with the context of [00:13:00] fully understanding every single weird technical thing about HubSpot that like just ChatGPT isn't going to have such a deep understanding of. And it's coming preloaded with all that information. So it's it's awesome.
Dax Miller, hapily: It's like asking a smart person, "Yo, you're smart. How should I file my taxes this year?" They could figure it out, right? They would have to get the context. Well, I know what year it is. They have to go and look it up. But you actually basically ask the tax preparer. How do I fill my taxes? They're like, duh, like here. Context, experience, life. What's funny though, my man, is nobody can get quote template stuff right. Nothing. Never gets quote templates. Quote templates in HubSpot are that difficult
Max Cohen, hapily: Well, to be honest, it's not like it even has the source material to be able to actually learn it.
So it's
Dax Miller, hapily: That's the problem. There's no source material, but I
thought that was always funny, man.
Nico Lafakis, HHH: If it doesn't exist it's going to be tough. He only goes, he's only got, he's only supposed to go to 3 online resources. So he's only supposed to go to the knowledge base, [00:14:00] assuming that there's something in knowledge base. he's missing, he'll go to the community, and he will go to the documentation.
Max Cohen, hapily: Yeah.
I've been able to, I remember I was having a problem with it and I was able to give it, a knowledge, like a recently updated API documentation and be like, look at this article and based on this, can you restructure what you built and what you showed me and it literally was just like, oh yeah, man. Okay. And by the end, then it did something different. Right. So like it has the ability to go like ingest new information on the fly when you're prompting it in real time.
Nico Lafakis, HHH: Oh Yeah. Yeah. It still has web access. It has canvas. It can do a data interpretation.
Max Cohen, hapily: Yeah. And the best part it's looking at like new API documentation with all the historical context of everything that came before it. Right. So it's going to understand it at a much more nuanced level than something else would, right? That's not HubSpot focused. What's your future plans for it? What's like the next big thing you're working on with it? What are you really excited about with all the new AI stuff coming out? And is there like anything immediately like Hub Helper Harry can [00:15:00] benefit from this like big time?
Nico Lafakis, HHH: Yeah. Because now I can actually talk to him through Agent AI. So there's like agent AI now has API capability. So there's going to be some really nice back and forth that goes on there where it's, I'm going to try to do it from both ends, but essentially I'm going to give HubHelper Harry users the ability to leverage agents from Agent AI within Hub Helper Harry, and then I'm going to try to also do it the other way around. So if you're using Agent AI, it can go and try to leverage... I'll be building an external assistant. So it won't be, which is fine because it won't need web access or anything like that. But, the difference will be that the Agentic Workflow version is what I'll label it. will be able to leverage higher end models basically. So that's the only thing that really sucks about Harry. And I have no problem admitting it.
Max Cohen, hapily: Don't disparage him. Nothing sucks about Harry. That's my boy.
Nico Lafakis, HHH: Yeah, look, he's good. He's good.
for what
he can do.
Dax Miller, hapily: my [00:16:00] mans like
that? Come on. Well, it's
because
it's his mom. It's his baby.
Max Cohen, hapily: 0 Nah, the best part, the craziest part about this is Nico say, you think that HubSpot Harry is cool? You ain't seen shit. s And I'm like, what? I'm like, what? It gets better. Dude, that's crazy. So you're building, so you're building something on Agent AI right now. Could you explain to me like a five year old, like what that is and how it's fundamentally, different from Hub Helper Harry.
Nico Lafakis, HHH: It's going to It is so fundamentally different. So like the latest stuff that's rolled out for agent AI is really powerful. If you look, if you're somebody like Dax who understands, like how awesome it is to be able to reach out and connect to certain endpoints that you wouldn't normally have access to, just through rest API call.
And, so, and using that, I don't really, I don't really have a problem talking about it because it's going to get released for free. Anyway. So, oh, well, basically, right now, I'm developing, an audit agent, like a HubSpot [00:17:00] portal audit agent. so I don't know how many we already we went over this yesterday and,
singularity.
Dax Miller, hapily: All my friends that bother me like, Oh, am I doing this right? I'm going to click a button.
Nico Lafakis, HHH: That's what it's going to be
Max Cohen, hapily: Dude, so this is, so. I've been saying as we see these like deluge of, breeze updates and AI stuff getting built into HubSpot, the thing that I keep coming back to is thinking of like, all right, like all this content creation stuff is cool All this enrichment stuff is cool. This, agent stuff is like pretty cool, but I'm waiting for the day where, you know, me as like an admin coming into a HubSpot portal to clean up a mess, right.
Can be like Hey portal, how you doing? What's wrong with you and that portal can go. Hey, dude. Thank God. You're here. There's a lot wrong with me Here's what it is, right and it almost sounds like you're building this
Dax Miller, hapily: I've got 9,000 address fields that keep getting written. Here's the history. These haven't been written. Should I just archive these? That would be great. Hey, can I, you want to archive this [00:18:00] property? It's being used in nine places. I can obliterate all those for you.
Cause I know it's bogus to click through all that.
So you want me to hook you up? Yeah, man, please
Max Cohen, hapily: They're still using, full questions as their property labels. Kill me. Kill me.
Save me. Please.
Nico Lafakis, HHH: Please delete all this shit and
Start over you guys got, they got workflows that they used one time in 2004. Still
Dax Miller, hapily: sitting here, man. My memory is clogged up.
Nico Lafakis, HHH: so that's the thing, right? Is like this, and that's why I don't really mind, I've been telling people since the beginning that like, I don't really mind talking about use case because there is something that I've discovered that even if you tell somebody use case. They're like, they're super reluctant to follow through on it.
It's almost like telling somebody, it's like giving somebody a map to go find never ending gold For some reason, they just,
because they have it, they're like, I'll get around to it. I have it. So, I'll put it away and I'll get to it, in a few years or something. Right? There's this weird thing like that.
So I don't, I say that only to say that I don't mind talking about this [00:19:00] stuff a, because there's not too many people that are going to be able to put it together as fast. And then B, well, I should say a, if you ain't first, you're last B, you're not gonna be able to catch up. C, you have no idea the complexity that this is going to in terms of like how powerful or like leverageable.
and actionable of a like audit response is going to be, it's to your point, it will be very much what you're, whatever you're expecting in your head, just
Dax Miller, hapily: Yeah. I expect an agency in a box,
Nico Lafakis, HHH: Yeah, something like that, because it's going to be like, if you don't so, if you're trying to encompass what I mean, when I say disruption, it's going to audit your portal, then it's going to talk to you about what best practices you could employ that are like HubSpot then it's going to talk to you about.
Tactics that you could take in order to better your content based on the content that you already have. It's going to do an email optimization for you and talk to you about your metrics. Look at your industry, do research on your industry, compare your information and give you [00:20:00] the analysis that you need in order to actually make like actionable changes to your emails, as opposed to just, Oh, let's AB test some And then at the, and then it's going to give you the content calendar for your social based on whatever your, ad targeting is and whatever your ad performance is.
It's going to give you a suggested calendar for that. along with suggested content based on again, industry, space that you're in competitiveness, does a background SWOT analysis and then. Throws that into a, another, competitor analysis takes that with all of your audit data and pulls together a 30, 60, 90 of, roadmap to level to measurable KPI effectiveness.
Max Cohen, hapily: You
said this was free?
Nico Lafakis, HHH: It's going to be
Dax Miller, hapily: Yeah. So what are you talking
about?
Max Cohen, hapily: Wild.
Dax Miller, hapily: No, it doesn't even matter because it's not about that. It's about what's next. There's this whole concept, man. What you're doing.
Nico Lafakis, HHH: See, like, that's the thing is love DAX for this. Go ahead.
Dax Miller, hapily: I get what you're, this is the thing. All right. There was a colonizer [00:21:00] named Christopher Columbus. He came and took a bunch of stuff.
Wait in the name of the queen. Right. He didn't do it for free, but it didn't matter because he didn't get paid. Anyway. And when you come and you've charted this land, like literally charted this land with AI, whenever I see AI anything, I'd be like, Oh man, they probably called Nico and asked him like two questions minimum. There's nobody that's charted these waters with AI better than you. This is not where you're going to make What happened in the back of my mind was like, how can I find 5 million and give it to him? So he can just like not have a job and just do this stuff. So he could just be like, yeah, I'm going to get me, I'm gonna get them rare Funko pops and then I'm just going to do this.
And that's what you do because there's no way that you would be able to, that monetizing it is almost against the creed because
Nico Lafakis, HHH: Thank you.
Dax Miller, hapily: your, because. You're just being like, all right, you thought this was the foundation. Now this is the foundation. And now this is the foundation because again, max for event•hapily specifically to call it out.
Max [00:22:00] wanted some cool stuff. He asked this, he asked you visa, like you, let me cut your head off. And you're talking like Futurama like. You're out. He asked you in a version of you like a symbiotic version of you of how to do this. Told him he did it. It's live. It works. He got better. He improved it.
And now he's elevated. And Max has a new foundation for his knowledge and grasping of what's possible because he Saw it happen in action. And when people look at this and they go in there, or screw anybody, I want to look at HubSpot. I wonder if HubSpot is good. We've been using it for three months.
Are we using it good at all? Click the button. It does the thing. Now they're on a new foundational level so that they can ideally raise the consciousness of everyone and do better for who they're trying to do better for, right? It's Oh, I made free water. I wonder how much I can make off this. That's not the point.
That's not the point and because you're on that mission and you wield that power, it would be mad if you're like, I'm gonna get $2.99 a month for this. What? $2.99 month? [00:23:00] Really? It's not gonna do anything. It could be a million dollars. It's not gonna do anything because what you've been doing, it's been charting these waters, paving the path, building the bridge for people to understand to not have to do the things that people shouldn't be doing.
We as humans just need to be able to set the robots in action and go live our lives or do better for the people around us. And bring positivity to the people around us because then the playing field is level and that's that. Oh, I know what you're doing, man. That's crazy. And no one's going to do that.
Sorry. Cause they're going to be like, we're going to monetize it. How are we going to get this? They're going to be doing the bureaucratic double dutch
of how they're going to get paid and how they're going to support this and how they're going to raise. You're already done.
Nico Lafakis, HHH: I gotta tell you like the number one reason that I'm with harvest is because Dave is every bit as disruptive about this as I am, and every bit as eager about doing it as I am, and yeah, we are, we're absolutely going to be like the number one disruptor in this space. And,
Dax Miller, hapily: I don't see
any other [00:24:00] HubSpot GPTs. What are you talking about? You already, it's not even disrupting because you're just the first one at the party.
You didn't show up in a purple suit, you were the first one, and you're like, well, let me, might as well put some music on here.
Nico Lafakis, HHH: I thought it was like hilarious, just this past year at inbound when, just throwing together the custom for hapily, right? The front end version of that.
Max Cohen, hapily: Yep.
Nico, I want to get your thoughts on the, we were talking about this during the Hub Heroes podcast, this morning. There's this Cause we did a whole episode on like Breeze and stuff, right? And I think whenever we're having these conversations around AI, there's this whole, there's this whole sector of folks that are, I'm in fear of it coming for my job and replacing my job.
And then there's this whole other section of people that are just like, no, it's not meant to replace you. It's meant to enhance and elevate you and make you do more. And like this, like very optimistic, looking, thinking, I'm like. I want to hear from you what's your honest take on the whole balance between that, right?
Because like I totally understand the argument and I believe in the idea of [00:25:00] yes, like this is meant to Get you to do your job better not necessarily replace things that you're doing, right? But then I also look at something like prospecting agent and go, huh, what about BDR is coming fresh out of college?
Sounds like we're not that many years away from that being not an entry level position anymore, right? Given the type of work that it's doing, right? And so there's situations where I agree with like both sides of that argument and like it probably is going to end up being like a both sides thing, but I want your wisdom on the balance between that and like how you approach that sort of a paradox.
Nico Lafakis, HHH: so the way that I see it, there's a couple of different things that get taken into account. The first is people who are already at their job that may not be using these tools already. You're really the riskiest, right? You're the riskiest for the reason for the third reason. The, or the third thing, Second thing is that if you're at your job and you're using these tools and you're fairly good at your job, you're probably [00:26:00] a lot better at it now because of using these tools and you're going to do well, you're going to do really well for every field that's been tested that.
Where people pick up any type of generative tool and use it to their advantage, they have all seen at minimum 65-percent increase, both in productivity and just overall knowledge, right ? Just a better understanding of what it is that they're working on better understanding of the context for what they're doing.
So you'll be fine. The third thing is that these, this, it's going to evolve, right? It's going to evolve from just working with the generative platform to working with an agent. And this agent. What we have now is just agentic workflows. Do not confuse what we have now for what's coming. If you want to see what an agent looks like, go check out, singularity soothsayers from last week, we did a demo on a operator.
And if you want to see something that might scare you out of your job,
that's definitely
Max Cohen, hapily: want to talk about that after this,
Nico Lafakis, HHH: So in watching that, a lot of people, their [00:27:00] feedback was like, Oh my gosh, this thing is doing my job. I'm watching it do my job. Okay. Well, the reason that the first group is. In worrisome that the first group says that when they see the third thing, right?
And the problem with that is because you're not using the technology, you're not able to augment yourself. So what's coming through this year is I'm going to be Nico's. It's going to be Dax's it's going to be Max's right? Because even though there's the operators, you need me to operate the operators. I have to direct them,
right? So literally everybody who's at your position. Now, you technically become this manager on a digital scale where you will be in charge of not only yourself, you'll be doing your work, but you'll also have an agent that does work. Beside you and works with you and does things with you. but you'll be directing initially, but over time it'll, you'll just be a matter of directing it and it'll understand what it needs to take care of.
Max Cohen, hapily: [00:28:00] The
individual, will
become a team.
Dax Miller, hapily: of teams.
Nico Lafakis, HHH: So that's really where this goes, right? It goes to a point of what, people are trying to put together well, how does commerce explode based on all of this? If there's going to be such massive job loss, how's it going to be job gain and this, that, and the other? Well, that's the thing is, if we're going to have this orchestra of workers, we're always going to have that thing in the back of our minds that if a human didn't verify it, we don't really trust it.
So there's always going to be a human operator somewhere in the mix, even with the driverless cars, there's somebody behind the screen in traffic scenarios, it's watching it all. So that's how it's going to work out. Will there be less people overall at a company? I don't know. The way that I see it this year will likely stay stagnant and increase based on what this year produces.
Right? So I think through the end of this year, people like, cause you're going to see the agent pickup, right? So one person's going to become two or [00:29:00] three which is going to lessen the need for new hires. But that also increases the want for what solopreneurs entrepreneurs, new companies, startups.
Dax Miller, hapily: Also, the thing is, I
think,
Nico Lafakis, HHH: You're going to have an explosion in all directions, really,
Dax Miller, hapily: the biggest thing too, is we live in this weird vacuum where all of us know too much. Not to toot our own horns, but we know too much. think that we are the ones we are not a representative of the world. Or the, excuse me, let's just take it to the nation. It could be the world scale.
We're not a representation of that. Our knowledge is not a representation of everything that's happening in the world. Most companies are like, AI, yeah, one day. AI agents, what's that? I've talked to multiple people that own businesses that are like, I haven't even gotten into that yet. We don't have a, they're like a GPT for my, tech support, so that they can just ask questions that I have and to go call some.
We haven't gotten to that and we don't have any of that. It's not normal. We are not normal. Nico, you're beyond not normal alien. As
You are. So there's a long standing [00:30:00] opportunity. Just like I think there's people still nowadays going through a digital transformation. Meaning we're getting off of our fax machine.
We're getting into the digital world. Like all the people that own businesses and create, like I go by if my cup, if there's a competitor that has a fax machine, I'm going and buying them and I'm going put. A CRM, in I'm gonna go take this storage place that uses pen and paper and Imma put a digital system in that they make, these digital transformations are still going on in 2025.
And AI agentic transformation like it's a ways out. The technology is here, but the implementation is a ways out because there still needs to be adoption. People used to give, people used to have jobs, cutting out backgrounds of images. That's all I do. I just cut the backgrounds out of people, pictures or whatever.
That's all. You don't have to have that anymore. And
now
Max Cohen, hapily: a demon with that lasso.
Dax Miller, hapily: I hit him with the lasso, but now even so like photo P and all that made it. So I don't need to have 300 bucks a month. And if I just go on, click the button, I got Photoshop-ish I got editing capabilities and now I have AI and I [00:31:00] can say, cut the background out and put fairies and unicorns, please.
It just does that, picks it up a little bit and I'm done. They used to be someone's paid hourly position to do so. Also horses used to run around and carry people. So I think that there is okay. Opportunities for humans to learn the technology, learn how to use a power drill versus a screwdriver and go take that to other companies and expand their horizons.
Because people move slower than us. We work in this LinkedIn vacuum where all we see is the same people in our same service, saying the same things about the same AI. As soon as you step out of that into the real world, they don't understand the concept of letting robots loose. They're so it's still a,
it's still a foreign
concept.
Max Cohen, hapily: On that,
Nico Lafakis, HHH: Yeah no, I realized that I'm in, I even, I tell my wife that it's the same way, we live in Dayton, Ohio, and I think the odds that I would find someone else in Dayton, Ohio, that I could have this conversation with or have a conversation, whatever. I'm not trying to say anything.
God, I'm like, I really don't measure this but to have to hold a conversation with somebody the same way I [00:32:00] would with you guys, I don't know. I think there's maybe like 1%. Maybe half a percent of the population. I don't know that I'd find more than a handful of people, basically, that would have the interest, let alone want to go into depth about it.
And more than likely, they would all be actually, no, we do have a really baller university nearby that does AI studies. So I could go to that call, go to the university, but outside of that. probably not so I understand what you mean in that regard. And I totally get it because look, they tried to go AI on mainstream media for a while and it didn't really work like people were just not
Dax Miller, hapily: They're just like, they're like, we'll stop talking about this. It's not even worth, it's not riling people up. It's not polarizing it. It's over their head.
They're like,
Nico Lafakis, HHH: Exactly. Yeah. Because it's something that you have to learn, right? And that's so, and that's what gets me like the most is people look at it so intimidating in like this intimidating view, and you really have to [00:33:00] stop doing that, like the sooner that you actually, and I don't, I really don't care what any of the other data scientists say.
I know they all say the opposite. I don't care because it's the easiest way for you to get used to these platforms easiest way to understand how this stuff works. If you just treat it like a person. And you just talk to it like it's a person, you will almost immediately understand not only like the usefulness of it, but what you could do what you could learn, like, how you can use it and fold it into your everyday life.
There's, even for my brother, who's a mechanic. Was able to show him a GPT advanced voice with, being able to use camera and the fact that it can take a screenshot. Anytime you're asking it a question about something, it's yeah. Okay. I might not know how to work on cars, but this thing does like, that's what I tell people nowadays to try to get them to understand it.
It's I don't know. Not that I'm going to do it. I don't know nuclear physics, but this thing does. I don't know [00:34:00] Python, but this thing does. I don't know how to build apps. But this thing does, I have, I've been over it since before the technology came out, the idea that like, yeah, there's going to be a greater intelligence than us.
And you're just going to have to deal with that. I'm overtly happy that it's already here because of the fact that I'm no longer barred. That's really what, how you guys have to think about it is like the amount of times you've had to turn to a coworker to ask them a question, the amount of times you had to go to Google to go look something up and then you hated it because you had to spend so much time on Google, actually looking into it, finding the right link, finding the article, finding the piece of the article that was relevant, probably took you at least five minutes on an average search, right?
Where you could just ask, Any generative model doesn't even have to be GPT. It could be Claude could be any of these models and especially perplexity. If you're doing a search, you get it in half a second, you get exactly the answer that you're looking for with the citations to the pages that it got it from
Dax Miller, hapily: And you can follow up with the context. Well, here's my actual thing. That's Oh, well, if you're doing, yeah, it's, it is people. If you're afraid of the tools and you don't want to learn the [00:35:00] tools, then you should be scared for your life. You probably already feared for your life anyway,
because again, this is going to what it is.
Max Cohen, hapily: I Speaking of fear for your life, Nico this is what I want to ask you. So, the other day, I hadn't heard of GPT operators until, I think, was it Dex, was it like Connor that posted a video in our
Slack that
was like, Hey here's operator booking plane tickets for me. And I literally just see the thing just like going through a travel site and like booking plane tickets.
And then immediately I just went, oh, oh, low level HubSpot implementations are about to get sniped. Right? and my question for you to tie this into, The HubSpot universe is like, where do you see operators being used practically within the context of HubSpot usage?
Am I overreacting? Thinking oh, the implementation specialist job is gone? Or the, are you gonna Is the HubSpot [00:36:00] admin going to go extinct because you've got this really intelligent thing that can push all the buttons and set a lot of stuff up. Right. And already clearly understands the nuances of HubSpot as you've demonstrated as possible, right?
Like with operators, what's stopping me? From having an operator that has the knowledge of HubSpot Harry for me to say Hey, I need you to go build me these dashboards and set up these pipelines and create a bunch of dummy data. And then it just does it on its own, right? Without a human doing that.
Dax Miller, hapily: Guess what?
Max Cohen, hapily: is this a
Dax Miller, hapily: Yo,
Max Cohen, hapily: is this
a
reasonable fear? I, yeah,
Dax Miller, hapily: You just said to do it. That's the human. You said, can you go do this thing? I made, cause you took concept that somebody told you or sent you an email. Granted, you could get that intercepted, but dude you were there
Max Cohen, hapily: Hold on. Hold on. That's true until HubSpot builds a chat prompt for the customer to just say, this is what I want to Do Right? And then it just does it. Now
I understand
Dax Miller, hapily: with
[00:37:00] you.
Nico Lafakis, HHH: okay. okay let's go timeline, right? Short term. What happens is what is what I was talking about before, right? Like operator rolls out to agency first and it rolls out to AI ready and capable agencies first, and then all the rest of them catch up. Who knows someday, and in an AI, so in that scenario, what's happening is, again, there's two maxes, there's two Daxes, there's two Nicos, there's two Chris's, two Georges, right?
Everybody just gets doubled. And what it looks like is you able to work on essentially two client portals and spin up two client portals at the same time, instead of one, right. Or Maybe you're a solopreneur, maybe you're on profoundly and there's plenty of jobs to take, but there's only one of you, you take one, your agent takes the other.
Dax Miller, hapily: Rush it.
Max Cohen, hapily: Yeah.
Nico Lafakis, HHH: not even like Russia. It's just we in this world's economy, most people need two jobs to facilitate. So why not now you've got, you've got two employees for the price of one when it comes to a [00:38:00] business. So
Max Cohen, hapily: the question is, did it take that job away from someone else?
Dax Miller, hapily: Yeah, I'm going to take it regardless. I'm going to take two if I didn't even have an agent because I'm crazy and I did take it away from somebody, but that's the mentality that people can't live on. They can't say I took
anything
Max Cohen, hapily: Not true. I'm just playing Devil's Advocate, that's
all
Nico Lafakis, HHH: no, no. And I fully understand that, but it's okay, but if that's true, why are there so many? How come I can go on Fiverr and find so many HubSpot jobs?
Dax Miller, hapily: Yeah. Why is everybody got that badge hiring
all the time?
Max Cohen, hapily: to go around.
Dax Miller, hapily: There's so much opportunity and people are just sleepy and scared. But if you just get a sword, you can go out and hunt.
And now we're giving you like the Ninja Gaiden where you got the three ninjas behind you, like doing the swab. It's Gradius, you got the option.
You need to have the option.
Max Cohen, hapily: You know what I think is like a really interesting use case for that? Is imagine a day Where let's say, I don't know, X amount of years, maybe it's a month in the future, but X amount of years in the future, you, you open up a HubSpot ticket and it's like Customer is complaining, X, Y, Z, isn't working.[00:39:00]
And then HubSpot tells you, we recreated the issue and we were successfully able to do it. And here's an example of us doing it with the operator. Right. And dude, that's going to be crazy. And we already have screen recorders. We already have operators We already have things that understand your portals and your customers really well and like that It doesn't seem that far off
as crazy as that
Dax Miller, hapily: he's thinking
agentic,
Nico Lafakis, HHH: Yeah, that's the thing is that's to me, that's this year. I know what Dax says, but like to me on an agent. So when I guess when I say this year, from our from an admin standpoint, I think it's definitely this year going into the end of the year. HubSpot is about it.
They're not. They're at the pace that you're talking about. Right? So they're not a year behind. They are a year behind in terms of educating people and like getting them onboarded to what's going on. Right? Like them talking about like agentic workflows and breeze and content creation and stuff.
This past year was just like, all right, cool. We've been talking about this since January, but yeah, sure. That's cool. But it's fine because [00:40:00] I, that's what I learned at inbound was exactly what DAX is talking about is like how small the percentage was how small the percentage is of like people who were actually this close to it and like actually keeping up with it.
and actually building stuff, right? Like the amount of builders on agent AI, there were, there are plenty of people on the platform. Not a lot of not a lot of builders that are like, I don't know, taking stuff too seriously in terms of there's, I don't know, eight different content engines, right?
And I'm sure that they're all special for some particular thing, but it's like when it comes to using different generative models, there needs to be like some distinction as to like, why this content creator is better than this content creator is better than this other one. Right?
and
Dax Miller, hapily: it's just a bunch of
to do lists at that point, they're like, oh, we got to do list, this to do list, that's the building block of, that's again, what HubSpot's doing, what's the building block, the easiest entry point, lowest barrier of entry of AI, I could take your video transcript and turn it into a blog.
That's like the, that's the AI bum on the corner. Yo, I can just, I could, Will convert content [00:41:00] for tokens.
Nico Lafakis, HHH: so so that's a perfect example, right?
You Oh,
at the end of the day, that's the system that they're going to end up running is like a credit based system, essentially. The you'll end up, what you'll end up doing is creating like agentic workflows or agents overall, and that'll end up leading into this sort of.
Network, paid network that you'll just be paying to use agents and getting paid for people to use yours and all that kind of stuff.
Dax Miller, hapily: We're setting the robot in motion and that's like the key. It always, it reminds me of this and I'll leave you cause we gotta wrap it up cause we're going to talk for two more hours. And this is the, what people need to hear. Well, these are the select few. there is, I don't know why I do not know this industry very well.
But the rap industry, there is some random video that always has, that lives in my mind with, free rent of 50 cent being like, Hey man, are there, somebody asked him like, what happened to Lloyd Banks? And [00:42:00] he said, I couldn't support somebody that wouldn't use social media. He said, Pac wouldn't use social media.
You think Pac would be on Instagram and you think Biggie would be on Tik TOK? And he's yeah, because you're thinking of the tool and not the
outcome. And he's I couldn't support that because obviously you want to build an audience. All we're doing is trying to buy attention. These are the tools to get that at scale.
And when you talk about multiplying yourself, which everyone wants to do, there's not enough time in the day.
There's not enough hours in the day. And now there are tools that allow you to subtract your human self from that time and time usage. And you can either be afraid that other people are now the world is going past you where things are just as hard and you can't feel like you can escape the race.
But now there are tools to do. So you need to take them or you don't. You can teach people, right? There's going to be a flood of courses. Well, the courses are already here. I'll spend pay me 2 grand and I will teach you every right. You can leave the water and people are just making [00:43:00] money making those maps, man.
And Nico, you're not making money making maps. You're building bridges. For the good of the humanity, and we appreciate your work, and I'm gonna ask you three questions. The
Nico Lafakis, HHH: Fire away.
Dax Miller, hapily: first question is favorite cartoon or animated character.
Nico Lafakis, HHH: Batman, the animated series,
Dax Miller, hapily: Man, that was dramatic back in the day. Second is favorite AI tool.
Max Cohen, hapily: That's like favorite kid how you gonna do
this one ?
Dax Miller, hapily: Did it. You heard me
I'm going to secretly go look it up.
Nico Lafakis, HHH: Man. It's only because of the flexibility. It's not even, understand this is not like a, Oh, you know how to use it the most, that applies to a lot of them. This is just, it has the most flexibility. It's like an Android versus Apple thing. ChatGPT.
Dax Miller, hapily: Okay. And last one is, you got one system on a video game system. You're on a deserted Island. No you're just stuck. You got one year with one video game system. What is it?
Max Cohen, hapily: This is the one that breaks him.[00:44:00]
Nico Lafakis, HHH: It's it's the classic Nintendo entertainment system and Battletoads.
Dax Miller, hapily: I was just, what, you're just trying to spend the whole year beating the speeder bike level?
Nico Lafakis, HHH: You got it. No, actually, not only can I beat the speeder bike level, I can beat it on two player with using both hands. Cause my brother couldn't get past it. So I, yeah, buddy. Yeah.
Dax Miller, hapily: Alright, well if you're going, speeder bike level is hard, It's doable, but then the snake level with two players because
you're like, dude, jump! Okay, one, two, three.
Blowing up. Well, at least you picked the correct system. Nintendo, I would have also accepted Game Boy. I would have also accepted Super Nintendo.
But I would have not accepted anything else.
Nico Lafakis, HHH: PlayStation Final Fantasy 7
Dax Miller, hapily: I don't know, man you're probably, your disc is gonna get messed up, but PlayStation is all for the collection, you've got the library, but you've got too many discs to hold around you ever have cartridges off
Max Cohen, hapily: Plus sand on a desert island too, I don't know,
Dax Miller, hapily: I'm just saying, it's a question, man. We
Max Cohen, hapily: the [00:45:00] assumption is that it works perfectly the whole time.
Dax Miller, hapily: Well, dude, thank you so much for coming on, man. It's been a ridiculous pleasure. Keep doing what you're doing. I won't be bothering you because I'm trying not to be, I'm not the smartest person in the room with this, but I got to at least catch up because you are leagues ahead of the entire industry.
So keep doing what you're doing. You built the moat, you built another moat on moat. Hey, if you like moats, cause you got moats on your moats.
Keep doing what you're doing, man.
Max Cohen, hapily: heard you like moats.
Nico Lafakis, HHH: Thank you guys and if you want to if anybody wants to catch a lesson for an hour, I'll change your life discoveraiwithnico.com
Dax Miller, hapily: Hey, shout it out. shout it out. He's booking time y'all.