Paxton Gray (00:00)
Hello everyone, I'm Paxton Grace, CEO of 97th Floor, and this is the campaign.

Thank you for joining us today for another episode of the campaign, where we talk with marketing leaders about better knowing your audience, innovating beyond best practice, and converting visitors into customers. The campaign is produced by 97th Floor, a digital marketing agency designed to build world-class organic and paid channel strategies for mid-level and enterprise organizations. You can find past episodes of the campaign on YouTube, iTunes, Spotify, and at 97thFloor.com.

Today's guest is Pete Larkin. Pete started in film running his own production company before making the move into marketing. He has built a career as a marketing executive, entrepreneur, and creator. Now he's made another leap into GTM engineering and AI orchestration, helping teams and individuals figure out where to start and how to put AI to work. In this episode, Pete will break down what's actually keeping people from adopting AI and shares how to push past each one.

He's also gonna share with us some of what he's building today with AI. Let's get into it.

Paxton Gray (01:06)
Pete, thank you so much for joining us today. We're excited to have you on.

Pete Larkin (01:10)
Thanks Paxton, I'm excited for the conversation today.

Paxton Gray (01:13)
Okay, all right, well, we're gonna talk about AI adoption within teams and organizations. And to start, why don't you talk a little bit about some of the benefits that you've seen for individuals or for teams from going through this, a bit of a journey working to adopt AI.

Pete Larkin (01:33)
Yeah, absolutely. I think first and foremost is just my own way of working. And in my personal life as well. It enables me just to do a lot more, more efficient, more effective in my work, which maybe this is a little bit philosophical, but

Paxton Gray (01:56)
Yes.

Pete Larkin (02:00)
Paxton, think maybe in conversations past, we've discussed a little bit about what do do with that? What do you do with those efficiency gains? And for me, instead of just running at 150, 160 % capacity with these new superpowers, if you will, for me, I'm able to reinvest into the things that I enjoy the most. And so for me,

On a personal level and my own work, it allows me to focus on the things that I care about the most and the things that I enjoy the most on an organizational level. My teams and the people that I work with, we are also more efficient and effective, but it's also transforming the way that we think and approach our work in such a way where, yes, like there are automations, but it's a cause for reinvention where the

Paxton Gray (02:41)
Thank

Pete Larkin (02:55)
It's not just how do we take AI and kind of fit this into a box of how we already do things, but how can we fundamentally step back and rethink the way that we do this work and to recreate it or reinvent it without the box of how those things were done in the past.

Paxton Gray (03:15)
Okay, so like an opportunity to reformat what we're doing, chase efficiency and reinvest that into that time saved from efficiency into better execution as we rethink what we're doing. I love that. So I conducted a small survey. Survey is a strong word. Among, it's about 40 people asking them what they see as the biggest roadblocks to their adoption of AI.

Personally and I think it might be good to go through as we read these like I think they're pretty consistent Of what people are feeling so I want to go through you haven't seen these yet and I'm gonna read some of the most common answers and get your opinions and maybe ideas on how to overcome as an individual and then maybe how to overcome it as a leader of a team who might have these roadblocks so the first is

Pete Larkin (03:57)
No.

Paxton Gray (04:13)
By far the most common, it's like not even close, time. I feel like I don't have enough time to invest in learning how to do this thing. It's like a big time barrier. have to do all the things that I normally have to do. You've had a couple of LinkedIn posts, I think, that address this specifically. So tell me your thoughts on that from individual as well as a team perspective.

Pete Larkin (04:19)
time.

Absolutely. I love this one. It's investment. when you look at say like planning for retirement and investing in your retirement now, like it can be hard to do that upfront, but then later on you get to you get to enjoy that. And it's very similar with how we invest into AI. You don't see the gains immediately. You see it over time. And as you build out your new automations and workflows,

those gains that you get back from that investment, like they come back in multiples, right? And so the question then, well, how do we do that? Right? Because first it's actually realizing and just mentally getting over the hurdle of like, okay, I believe that if I actually do this,

that I will see gains. It will be worth the investment of time. So it's making the decision first. And then second, practically how do you do that? For me, it's time blocking. I have to time block. I have to put it into my calendar and I try to go, I'm on average around 45 minutes during the work day. I'm a little bit unique in the sense that it's fun for me.

And so I invest personal time in this, like after the kids get the kids down for bed, hang out with my spouse. And when she's tired and going to bed, then like AI school is starting and I lock in and I'm learning more. But for those who want to have a better sleep schedule than I do, time block it, put it on your calendar. Find a time of day that maybe is a little bit slower.

Um, or maybe, um, when there's not so many meetings, but if you don't actually put it on your calendar and time block it and put it as a recurring thing, I think it should be daily. Um, it's going to be really hard. It's going to really hard to, to build out your AI fluency and build up those skills.

Paxton Gray (06:52)
What to somebody who's like, okay, I'm gonna block out the time, but I don't even know what to do. What would you point them to?

Pete Larkin (06:56)
Mm-hmm. Yeah.

I'd say start with the courses that are available to you for free from the major LLMs. Anthropic is where I would start. Go do the free courses about Claude. They're great. They're great content. They're engaging, very professional, well-organized. And then OpenAI has got some as well. I have some of my favorite creators that I follow on YouTube and LinkedIn, specifically on YouTube, where I get the most value from the videos that I'm watching.

is Nate Hurk. He used to be local in Utah. Now he's in Chicago. He's got an agency called, I think it's True Horizon. But the content he puts out is so, so good. But here's the thing. Whatever content you're consuming, you can't just be watching it. You have to open your computer and you have to follow along. You have to be doing it as you're learning it or it won't stick. It'll be theoretical and not practical.

So if you want the skills to stick, follow along on your computer.

Paxton Gray (08:04)
Love that. And then from a leader perspective, how do you help people get over this hurdle of time?

Pete Larkin (08:13)
I think it can feel overwhelming for employees because they feel like, okay, not only am I supposed to be learning this stuff, but I'm supposed to be more efficient, more effective. I'm supposed to be doing a lot more with this. But because I'm, I am expected to do more. I feel like I need to be doing more work and executing on these skills that I haven't developed yet. Right.

And so that's a little chicken in the egg here and leaders need to understand that. Like you can't expect your employees to wake up one day and just magically have AI skills. You have to invest in the tools for them to have access. So give them budget. But then you also have to act to invest in the time just as they are investing in the time. You have to invest in enable them to have time learning these skills on like during work time.

You can't expect it to be during their personal time. It's just going to be just going to burn them out. So give them time and encourage them to time block. Maybe even have like your AI hour every day. And you could even say, hey, full company like everyone in the company from like, I don't know, say two to three. We're going to like this is AI hour or something, you know, pick a time.

And if that doesn't work for you for AI hour with the company, find another time that does. Um, but don't just encourage it, but also like in the one-on-ones talk about it and try and make it more of an expectation.

Paxton Gray (09:48)
That brings us you mentioned resources that's this the number two most common response which is cost in resources I think from a team leadership standpoint the answer is pretty clear like provide budget for that provide the Subscription when needed from a an individual standpoint Is there maybe something that's like cost effective or how to think about the cost because I think for people who haven't had a chance to dive in

They hear these horror stories about credit usage and racking up massive bills and all of that. And that could be potentially a little intimidating. So how, should somebody think about that?

Pete Larkin (10:27)
Yeah, I mean, there's a lot of free tools to use. There's so many free tools, Google especially. But I personally think even if your organization, your company isn't paying for any type of paid account, whether that's OpenAI or Claude, I mentioned those. Of course, there are others. To me, those are the two. If you're going to pay for an LLM, those are the ones that you would typically

pay for. To me, it's hands down, no question about it is Claude. Like that's where you should put your money and go to the pro account, pay the 20 bucks a month. If your company isn't paying for it, go to the $20 a month account and it's well worth the investment. And so the starting anywhere when it comes to those tools.

start there first, $20 a month with Claude, and then expand from there, you'll see, as you're building out your capabilities, you're going to want to jump into additional tools as well. And again, you'll build out those efficiencies and those gains, and it will be well worth it. However, then the question is, you mentioned credit usage. You don't have to worry too much.

about credit usage and unless you're building via API or which maybe that's a discussion for another time. Or if you're getting into cloud code and you're just running through tons and tons of tokens, my belief is if you're running through that many tokens, it's probably because the work you are doing merits it.

then you've got to talk with your organization or yourself and decide, do I want to invest more? Do I want to be able to get more work done? Or do I need to step back and settle down a little bit? I don't think that token usage should be too much of a concern for people trying to build out their skills.

Paxton Gray (12:41)
Okay, number three most common is the concept of hallucinations, which I will lump in with another item, which is like trust. I'm not sure I can trust what it produces or what I get out of it. Would you say this to either individuals or teams that have that concern?

Pete Larkin (13:02)
Yeah, the models continue to get better and better. And the hallucinations, I see like the frequency of like significant hallucinations, like serious hallucinations are dropping. And also as you learn more about these tools and these technologies, you'll see what AI is like really good at and what it's really good for and what it's not.

And so one example is math, right? When I'm doing sometimes pretty simple calculations, AI just struggles and it just doesn't get it right. And so I know in what situations I need to be extra diligent in the way that I review and look over the math and the results. But you should always

be reviewing the output of the work. There's another term for this, right? And I'm sure many have heard this as AI slop, right? Between hallucinations and what the result is, is just sometimes mediocre or frankly, inaccurate bad work from AI. And ultimately, before you even use the...

the tools before you even use AI, need to understand and think about is this the right use case? Is this the right time and place to be using AI? And one, I'll just give a quick example. If I'm going to say create a customer facing marketing material before it goes out, I'm definitely going to review it, right? Don't just trust it on its own to be high quality, expect to have some edits and adjustments, but it's a lot of this has to do with the

It's the good input, what you put in, what you get out, bad input, bad output.

Paxton Gray (15:02)
Yeah, yeah. Now I'm going to bring up some less common concerns that I think are actually really interesting and I did not expect to see. Just to, I don't think there's necessarily a great answer these, but they're a little bit more philosophical. One is I'm concerned that if I dive too deep into AI that I'm going to get too dependent on it.

Pete Larkin (15:12)
I'm excited to hear these.

I'll go.

I love this one too. We talk about this one with my kids about like, it going to make us dumber? dependent on it in a sense where like you don't learn skills that you need or your skills start to fail. Right? This one's tricky because like I think about like churning butter.

It sounds a little weird, but like how many people do you know that know how to churn butter? And it was a skill at some point that was like very important skill, right? Or like breaking a horse. Advancements in technology make certain skills irrelevant or unnecessary. And so I like to look like step back and say like, how important is this skill to maintain for myself?

and here's a quick example, like when I'm in, working with AI in Excel or, or Google sheets, like there are certain things that, certain equations or, or, functions within Excel that, I don't really need to memorize or, like maintain those skills and keep those fresh. If the tools can do them for me, I don't, I think it's okay if it goes the way of the butter.

right, of turning the butter and like if those skills like falter. But then there are other skills. Critical thinking, I think, is probably what more people like lean on when they come to this concern or this issue is that they they worry about how it's going to impact their critical thinking and and like keeping them sharp. And I think a lot of us who have seen the movie WALL-E think about, you know, the really giant fat people in the movie who are just like

relying on technology to do everything for them and then just become these like dependent, you know, fat people who just vegetate and don't do much. And in my opinion, I think, I think when you're using AI as a thought companion and to help you enhance your current skills, it's not a dependency where you, you,

like stop thinking and you stop, the critical thinking or developing, you know, skills to keep you sharp. I think it can actually help you, grow and evolve the way that you think. but the caveat, and like, this is what I share with my kids. I think it's important to understand what's happening. Like when you ask a question,

There are certain low profile requests or tasks that I don't think it's that important that you understand. Like show me your work or help me understand exactly how you did this. But there are certain situations where you don't just, like if my kids are, and myself as well, but the example of my kids and their learning, if they're say trying to,

understand we do a lot of like robotics at our house. And so like, you're trying to learn like how to create these robots, don't just tell the AI to give you the answer, but to walk you through it and help you understand how it got there. So that you're, becomes a learning tool. It is actually expanding how much you understand and learn. And I think it can help you move very, very fast or help you learn a lot faster.

Paxton Gray (19:19)
Mm-hmm. love that. Okay, another interesting, uncommon answer, but I thought was kind of, I hadn't thought about it before. One barrier to adoption is this concern or ambiguity around morality. There's a lot of kind of ethics that have been well established that AI comes in and challenges some of the assumptions around.

ethics and what we create as being presented as real or fake and like what to do about that. What's your current take on that concern?

Pete Larkin (19:57)
Yeah, I think we should be transparent and open about how we present AI work. I don't think we should ever try to pass off AI work as human work in the sense of, like, for example, if I'm going to create an email using AI,

say a marketing email at the bottom of the email, I'm not going to say this was this email was created with AI. I'm not also not trying to say like pass it off as this email was created by a human. Right. Like it's not I'm not going one way or another with it. Right. I don't see it as unethical to not disclose in that sense. However, but like imagine you call in

and to like a support line, know, customer support, and you're trying to get answers and you get on the line and it's AI talking to you. And if you ask the question, like, is this AI? And it lies to you or it says that it's not. I think that's super unethical. also, this is like, this comes to taste, right?

We talked about this Paxton, and I loved your comment about this. And I'd actually like to turn the question back to you regarding taste, but like when it comes to outbound or cold outbound in prospecting, to have an AI call a human to try to then have a conversation and sell something, all that says to me is that you don't care enough about me as a customer.

to send a real human to talk to me, right? And so, don't take my time with your AI, right? I think it's bad taste. And I think going back, we shouldn't try to pass stuff off as human if we're trying to make it look human. That makes sense. I think it's deceitful. So I think that that's where some of the ethics come in. I'd like to hear your perspective on this when it comes to taste.

Paxton Gray (22:15)
Yeah.

Yeah, so we developed an acronym to help guide us on how teams individuals and companies will succeed with AI and based off of the number one roadblocker concern being time As the thing that's preventing people from getting deeper. That's the acronym is time and so it's taste impact matching work and elevating so

T is taste. Basically, like the belief is teams will win in AI as they develop their taste. So if I'm a brand new marketer today, I'm not gonna be as worried about learning the ins and outs of maybe like an analytics setup or converting data from here into a spreadsheet. I'm gonna learn more about taste, like what is going to actually drive action? How do I persuade an individual to

take this next step in this journey. The stronger your taste, the better you'll be able to drive AI in the correct direction. If you have no taste, that's AI slop. So you can produce a lot, but it's not gonna actually do anything. So then impact is self-explanatory. Like I'm looking not to increase the volume of deliverables. Instead, I'm looking to increase the amount of impact that I have ideally on bottom line. Matching is...

matching the correct work to the right intelligence. So we have artificial intelligence and organic intelligence. So matching means developing the instinct to know this task is best suited for AI and this task is best suited for artificial intelligence or organic intelligence. And then last is elevate. So the concept of, you know, I'm not just going to spit this out from AI and not review it. Like AI can take it 80 % of the way there, but humans should polish and elevate whatever this is before it goes out.

I think there's some cases where that may not apply, like you talked about, it's disclosed or it's understood. This is AI generated. In that case, I think it may not apply. yeah, so that's our acronym, TASTE, developing our ability to actually persuade.

Pete Larkin (24:37)
Yeah, I love that man. There's one additional point that I think is interesting when it comes to AI as an art form, because when you look at like painting, when photography started being presented as an art form, I guarantee painters were like, that's not art. Like this is you know, this is technology that's not even doing the hard hand done work, right?

And now we look at photography like, of course it's an art form, right? But then you can see this like in pop culture and popular cases that are going on being debated where artists who are creating art with AI can't get their work, can't get a copyright for that particular work of art. In my opinion,

It's a different medium. It's a new medium, but in my opinion, it's very much a medium or form of artwork.

Paxton Gray (25:42)
Yeah, I agree with that. I agree with that. And in all cases, you could say like the differences, like what kind of heart is the person bringing to this medium? Because, you know, if you take like a still life versus a snapshot, you could argue like that still life, this person's poured hours into this, there's feeling in that painting, the snapshot is like careless, I don't really care.

But then you can bring heart to that same media of photography and you get these beautiful, amazing, world-changing results. The same thing could probably be said for AI. You can have the slop, the quick stuff. I'm doing one prompt to generate this. And then you can have the well-thought-out, intentional. And even if it's AI generated, it doesn't mean that there's no heart or human touch there. Yeah, I like that. I think that's a great take.

Pete, I know that you are doing some really cool stuff in AI, and I think it would be a shame to end without hearing some of the cool things that you're working on and that you've built for yourself or for work. So tell us about two or three of the projects or things that you've built or are building with AI.

Pete Larkin (27:02)
Yeah, the one I'm most excited right now, I'm calling it, I have never shared this like in any type of format outside of very like closed doors, personal discussions. This is one that I'm calling the family operating system or FOS. And basically it is a way to take the conversations that are happening within my home to help me become a better father.

to become a better spouse, to give me coaching on how I can address situations better. This is one I'm really excited for, to see how it's an experiment, it's just happening within my own walls. So I'm excited to see how this one goes. Other past projects.

Paxton Gray (27:52)
Wait,

how far are you along on that? Like, has stuff been built? Are you testing right now?

Pete Larkin (27:59)
This is one of the things I love about AI is it enables me to do things that I in my wildest dreams never thought I would ever be able to do. I now consider myself to be a vibe coder. Never thought I would ever be like, and you know, coding anything besides HTML, CSS, like it's just, I don't know. That's like, I don't have those skills.

I believe that I could learn them if I was going to dedicate time to that. Nevertheless, like I was talking about robotics earlier, I also never thought we would be doing robotics. But now, my kids just using AI and we ask, this is the output, this what we're trying to accomplish, how do we get there? And so we bought a 3D printer and we bought some, like the AI will tell us what electrical components we need to go purchase. And they're cheap. You buy these small electrical pieces,

And then, you know, we follow the instruction and how to build them. And now like, like here I am in my basement, like soldering with my kids, like building out this, you know, little robot where like, I didn't have soldering skills before, right. And I didn't understand, like I'm learning in the doing, right. Going back to some of our previous discussion about, you know, being worried that our skills are going to,

to fail us by being too dependent. Like it can greatly enhance your skills and what you're able to do in the act of doing. So right now you ask like, where are you at? It is in the very early stages. I've been with Claude, I have been learning about different microphone systems and like a radial array and how to pretty much build these microphones.

this system in a way that it will be able to only record when it hears a human voice over a certain decimal point and then be able to transition from one room to another while stitching those audio files together into a single file. within this project, there's electrical engineering, audio engineering, there's coding, right?

the, software and the hardware and like, there's so much into this, but before I would, I would never try something like this because I felt inadequate. never, I didn't feel like I would be able to do that, but now I'm pretty good at following instructions. And I'm also pretty good at communicating the vision of what I'm trying to accomplish. And those are our two important skills with AI. Can you communicate well what the end goal is and, and the different components of reaching that goal?

And then can you follow the instructions well? So yeah, it's pretty early stages. Tomorrow I'm meeting with an electrical engineer to, there's like a 30 page build plan on how I plan to approach this. And I'm gonna present that plan to the electrical engineer to say, hey, what do you think? I think that's a good plan. So that's one project. You and I have talked about my podcast Paxton.

Paxton Gray (31:13)
Cool. Yeah.

Pete Larkin (31:22)
For personal learning, AI, again, is just an incredible tool. The gist of this project is it takes my favorite newsletters and through a daily automation through Claude, it will convert my favorite newsletters into a podcast about 15, 20 minutes per episode using the Notebook LM that

I really enjoy, I learn a lot from it and I enjoy listening to it. So that's one of the projects that I've really loved is my personal podcast. We're on episode 156 today. Got about 3000. What's up?

Paxton Gray (32:03)
Are you okay with sharing

how they can find that?

Pete Larkin (32:08)
Sure, sure, yeah, you can find it on Spotify, you can find it on Apple podcasts. The show is called the AI Deep Dive. The thing is there's like four or five different AI Deep Dive podcasts. And I shared this with, I think I shared this with you, Paxton. The only reason it has that name is because Notebook LLM does not allow you to prompt out certain things or give it more instruction. And at the beginning of every episode,

It always says welcome to the deep dive. And so instead of like fighting against the grain, I just like decided to go with it. And I was like, Hey, this is for my own personal learning. And I don't care. Right? Like I don't care so much about the name. I'm just going to roll with it. So that's the name AI deep dive, Apple podcast, Spotify. You'll have to probably put Pete Larkin in the search to find the right one. so AI deep dive, Pete Larkin and new episode every Monday through Friday.

Paxton Gray (33:08)
Cool, that's awesome. Any other cool, yeah.

Pete Larkin (33:10)
Yeah, those are a couple.

Some personal ones I have. We talked a little bit about business. I do a lot with GTM engineering. I won't dive too deep into that. I think some of the more interesting ones are like the personal uses. I have an agent that will go through searching for any flights.

When you go to Delta has like their flash sales, like that you can find flights using points. And so I have it every day. It will go to the website and we'll search for flights that pop up leaving Salt Lake City, which is my base airport, leaving Salt Lake City that is international, that is under 20,000 points for the flight. And then if there is one, it will send me an email.

that says, here are the details about the flight with this flash sale, which I love that. And it's also helping me find some property. And so it's searching through different sources that I've given it access to to help me find land. So just some interesting ways to use it.

Paxton Gray (34:29)
Yeah, I love that. love that. Pete, thank you for sharing your knowledge with us. I have to say like you more than most people I know, like you have the ability and proclivity to dive head first into some things and not worry about, know, am I going to do this exactly right? Like you're very much like, I'm just going to dive in and learn. And I think as a result, you end up getting, you know, very early into some

really cool stuff. yeah, I appreciate you sharing that with us. To wrap up, one of my like the question that we like to end with is who is like, could you share with us a marketer that has helped shape the way you view marketing?

Pete Larkin (35:17)
Yeah, I've been thinking on this since you gave me that you like there's a little cheating behind the scenes here. I was informed beforehand this question was coming. I've been thinking a little bit about it. This this one's a little tricky. Of course, there's like Malcolm Gladwell or Seth Godin, some of the you know, Neil Patel, other marketers that like I really enjoy their books. I love learning from them.

Adam Grant, this is more on like the philosophy side, I more like personal, a more personal response to that. I think a lot of it has to do with just like the friends that I meet and the connections that I make through networking and through different events and seeing the awesome work that they're doing that like I admire.

the work Paxton, you're one of them. We went to school together. And just like the incredible things that you and others are doing that like I, that I admire inspires me to, you know, continue to learn and figure out new stuff. I, I find I'll, I'll drop Alex MacArthur's name. He's a friend of mine. One of the main reasons I like to, I have a lot of respect for Alex's work is because he

puts it out on the line. Like, we'll put like a lot of risk and we'll put his, you know, kind of reputation on the line to do work that he really believes in. And he has a lot of fun doing it and it's super creative, but I know it's risky and I know what those boardroom discussions look like. And the fact that he continues to do it despite.

you know, the concern, the worry that's coming from the finance team, people worried if it's gonna bring back a lot of backlash and he goes through it anyway and it's like, it's always a home run. He always crushes it and like that's one thing I love about Alex and I'd also say like the early days of the Harmon brothers, like I really loved the work that they were doing but yeah, those are a few.

Paxton Gray (37:36)
Yeah. Great. Great. people to follow. And I agree. Yeah. Alex, he's had a bunch of home runs too. like he has put it out on the line and then he's knocked out of the park, a number of times. So, I think that's a, he's a great, great follow. Pete, thank you so much for taking the time. What is the best, like where could we send listeners to, to either connect with you, follow you or follow along with what you're doing.

Pete Larkin (37:45)
Yeah.

Yeah, absolutely. The majority of people will find me on LinkedIn. I recently created an Instagram account for business purposes. My personal account is super locked down. That's family only. But yeah, find me Pete Larkin on LinkedIn. You can also if you go to theumx.com doing a lot of events. This is typically more for those in Utah. But if you want to come and hang out and

and connect like in person, theumx.com, those will be the events that you'll find.

Paxton Gray (38:40)
Okay, all right, and we'll obviously post these in the show notes too. Pete, thank you so much for taking the time to share what you've learned with us.

Pete Larkin (38:48)
Yeah, thanks for having me. It's been a great time, man. I always love chatting with you and always about AI.

Paxton Gray (38:54)
Likewise, likewise.

Paxton Gray (38:56)
That's it for today. If you enjoyed this episode, please consider leaving us a five star rating and subscribe so you don't miss future episodes. Big thank you to Pete Larkin for joining us today. Be sure to connect with Pete on LinkedIn to hear about what he's working on and check out his daily AI podcast, AI Deep Dive. You can find past episodes of the campaign and examples of our work at 97thfloor.com. There you can learn more about the agency and get in touch with a marketing specialist to get support for your own marketing campaigns. That's it for now.

Thank you for listening. always, keep innovating, keep converting.