Betting on AI in B2B Marketing

AI itself isn't new like it's been around since the 1950s. The breakthrough that really led to a lot of what we see today is the ability for machines, for AI to understand and generate language. And so when ChatGPT emerged in November 2022, all of a sudden anyone had access to it. So prior to that moment, it was hard for a marketer to look at AI and say, yeah, I totally get it. I can go use an image generation tool or a video generation tool or a language generation tool. But once we had generative AI that a user with no COD AI ability could show up and get value out of it, that was what created this kind of tipping point.

Paul's here. I'm excited to do this. I've wanted to have Paul on the show for a while. I don't know why I didn't reach out earlier. Maybe I did, and there's an email and spam somewhere, who knows? But Paul, it's good to have you here. For people that aren't familiar with you and your background, how do you talk about who you are and what you do?

Yeah, so, I mean, my. My current life is CEO and founder of Marketing AI Institute, which I started in 2016. At the time I was running PR2020. You and I crossed paths at different times in our careers. When you were at HubSpot and Drift, when I was running PR 2020, which was HubSpot's first partner back in 07, but I sold that business in 2021. And today we're really focused on trying to accelerate AI literacy in all aspects. Not just business, but society, governments, everywhere. And so that's. The vast majority of my time goes to AI education.

Can you talk about. We'll get into the AI stuff, but just your past life and building and running PR2020. Talk about that business.

Yeah, so that. That started. I was 27. You know, I started that. I had been five years at an agency. I graduated out of journalism school in 2000, started PR 2020, 2005. And you know, as you can appreciate, like, the world was changing a lot. Like the technology wise was starting to really transform around 2005. I think that's the year YouTube was created. Google's like 6, 7 years old at the time. Social media is emerging. Twitter's about to come on the scene two years prior to the iPhone emergence. So we were just like, we're in that really origin phase of mobile and social and content marketing and podcasting starting to become a thing, blogging. And so it was A really fascinating time to be building a business. And it was right around the time HubSpot was created. I think they started in summer 06. We became their first partner in fall 07, and we just sort of like, rode that wave. And that's what PR 2020 became, was largely like a HubSpot partner. It was a fascinating time to build a business and build a service company. It was hard. Like, service businesses are really hard. Being an entrepreneur is hard, as, you know, like trying to build something and create a following and a community and eventually a membership model, which is what we do with the Marketing Institute. But, yeah, I mean, it was. It was great. I was just. By 2021, I was so ready to focus solely on AI that I just had to get out of the service business.

It's cool because your story, like, the story of growing PR 2020, just. You seem to have this in your DNA. And I don't know you personally, but. But you were early on the inbound content movement as a marketing channel. You build an agency, then later you. You're early on AI. Where does that. Can you just talk about what type of advantage I try to push people on this podcast, especially marketers. I've noticed that a common thread between the top performers that I've been lucky to, like, work around as peers or, you know, my bosses or whoever, they all have this thing that you have, which is they're early on something. And I want to push more people. And we had talked a lot about AI, like, last year, and like, hey, this is an opportunity to go build, you know, plant your flag in this. How do you. Where does that come from? And how do you push people to do that, to go take a channel or something and go be early on something? Just. You can create such a huge advantage in a world where there's so much noise.

Yeah, you can. It takes fortitude, though. Like, it takes resiliency because it looks like you're wrong for a long time. So, you know, when I bet on AI, I started researching it in 2011. I wrote about it in my second book in 2014. Just like 700 out of 50,000 words was about AI, but it was just like a premise of what I thought was going to happen. Did Some public speaking 2015, started the institute in 2016. Lost money every month for seven years building it. Like it from the outside, it was just like, yeah, that's cool. He's working on that AI stuff. Don't really get it, don't understand it. But as an entrepreneur, you're so like you believe so deeply in the vision, but it's the will to see that vision through until it makes sense to everyone else that really creates an entrepreneur. And that was like that concept actually came from a book I read early in my career called Will and Vision. And they talked about what does it take to basically be a category leader? What does it take to build a great company? And the premise was like a lot of people have a vision for tunity and how people choose to learn in different ways. And some people want that interactive component they want like the more visual. So we sold it through like Eventbrite and like just patch the thing together back in the day I think we maybe use Thinkific as the learning management system back then. So then in 2004, 14, I invented this kind of like point pricing model to sort of reinvent how pricing was done within service companies. So we created a series for that that did probably a quarter million or more. Then we took our, we taught other agencies how to do client services. That series became, you know, quarter million plus. So we were doing something once that was like 30 to 50 hours of my life to create that could make money overnight while you were sleeping. And so when you're running a service company where every dollar basically is another element of human resource needed to deliver that dollar to create something that just perpetuates and generates revenue and value for the customer. That's a really interesting thing. So I always wanted to find a way to do that. Like I never was that interested in building software and kind of diversifying our revenue through software. But I saw information products as like a real key thing.

Yeah, everyone, everyone says that to me now. Everyone says it to me now. They're like, you need to be built. Like if you were really smart, you'd be building a SaaS product to then sell all these B2B markets. I'm like, look, I agree that that could be a path. I have no interest, I don't have any interest in do that. I'm not a great product person. I don't want to get in with developers and engineers. Like I'm very intentional about like just doing the, just doing the knowledge and education side of this. I think that's interesting call out when.

I think it's really, it's a great business. But as you know, the key is you have to have an audience first. Like if you show up and start selling memberships or selling courses and you're not, you don't have an online audience or you don't have a subscriber base of newsletter subscribers. It's really hard to have success.

Yeah, I didn't mean to hit, I've hit this raise my hand button in Zencastr for the first time. I don't know why I've ever done that.

It's okay.

Um, so, so, yeah, so, so it, and, and it's interesting to think about what, what's possible. So you've always been, you've, and you, you, you have a lot of common threads. Like you built an inbound agency, you've always been creating content. What was the catalyst to get you to say you started researching AI in 2011? When did you realize that there was an opportunity to like fill a gap here and, and create content and educate people? And what was the first piece of that for, for your business?

So we looked at. So if you go back to 2014, the second book comes out 2015. I don't know that I ever did a talk about the core of that book. So the second book was the Marketing Performance Blueprint. It was Talent, Tech and strategy, how to build an optimal like marketing program. Basically the AI piece was a segment of the tech section of the book. That is the only thing anyone ever asked me to come and speak about. The U.S. government called. I keynoted the Mark Tech Conference. The next year I did south by Southwest. So I realized by 2015 that the forward thinking organizations and events were interested in the AI topic. And then 2016, Mike Kaput and I, my chief content officer who was working for me at PR 2020 at the time, we got talking over a happy hour and it's like, well, this AI stuff's kind of interesting. Like what should we do with it? And so we looked so well, we can tell the story of it. Like that's what we're uniquely capable of doing. We're not the engineers, the data scientists, but we were journalists by trade. Like we, we can tell a story. So we created Marketing Institute as a blog that was part of our agency. It was just a separate domain under the agency. And we, the two of us just basically started researching and writing about AI. And my belief was if we build an audience, like if we get people subscribing and we get enough of them. At the time my belief was 10,000 was like the minimum I wanted before I committed to anything else, then we would maybe consider building a conference. So my friend Joe Polizzi had built Content Marketing Institute, Content Marketing World. We'd seen it done. I kind of knew what the blueprint was. And so that was the vision of like, well, we could build something like that if we get enough people interested in it.

And. And when you said you, you started writing was the initial distribution. What was the initial distribution of that to get those first 10,000 subscribers?

It was just a blog and then a weekly newsletter that just kind of recapped whatever we were writing about. And the blog was. There was a few components to it. One we wanted to look at use cases and tools if they were out there too. I wanted to interview the entrepreneurs who are actually building the stuff because I was still trying to figure out what is the story. Like is. Is a real. Right now. Can we be integrating it into our businesses? What does that look like? Should I be thinking about it for our clients and bringing services to them around AI?

And everyone says that to me now. Everyone says it to me now. They're like, you need to be built. Like if you were really smart, you'd be building a SaaS product to then sell all these B2B markets. I'm like, look, I agree that that could be a path. I have no interest. I don't have any interest in do that. I'm not a great product person. I don't want to get in with developers and engineers. Like, I'm very intentional about like just doing the. Just doing the knowledge and education side of this. I think that's interesting call out when.

I think it's really. It's a great business. But as you know, the key is you have to have an audience first. Like if you show up and start selling memberships or selling courses and you're not, you don't have an online audience or you don't have a subscriber base of newsletter subscribers. It's really hard to have success.

Yeah, I didn't mean to hit this raise my hand button in zencastr for the first time. I don't know why I've ever done that. So. So, yeah, so. So. And it's interesting to think about what's possible. So you've always been. And you have a lot of common threads. Like, you built an inbound agency. You've always been creating content. What was the catalyst to get you to say you started researching AI in 2011. When did you realize that there was an opportunity to like fill a gap here and create content and educate people? And what was the first piece of that for your business?

So we looked at. So if you go back to 2014, the second book comes out 2015. I don't know that I ever did a talk about the core of that Book. So the second book was the Marketing Performance Blueprint. It was Talent, Tech and Strategy, how to build an optimal, like marketing program. Basically the AI piece was a segment of the tech section of the book. That is the only thing anyone ever asked me to come and speak about. The U.S. government called. I keynoted the Mark Tech Conference. The next year I did south by Southwest. So I realized by 2015 that the forward thinking organizations and events were interested in the AI topic. And then 2016, Mike Kaputnai, my chief content officer who was working for me at PR 2020 at the time, we got talking over a happy hour and it's like, well, this AI stuff's kind of interesting, like what should we do with it? And so we looked so well, we can tell the story of it. Like that's what we're uniquely capable of doing. We're not the engineers, the data scientists, but we were journalists by trade. Like we can tell a story. So we created Marketing Institute as a blog that was part of our agency. It was just a separate domain under the agency. And the two of us just basically started researching and writing about AI. And my belief was if we build an audience, like if we get people subscribing and we get enough of them, at the time my belief was 10,000 was like the minimum I wanted before I committed to anything else, then we would maybe consider building a conference. So my friend Joe Polizi had built Content Marketing Institute, Content Marketing World. We'd seen it done. I kind of knew what the blueprint was. And so that was the vision of like, well, we could build something like that if we get enough people interested in it.

And when you said you started writing was the initial distribution, what was the initial distribution of that to get those first 10,000 subscribers. Subscribers.

It was just a blog and then a weekly newsletter that just kind of recapped whatever we were writing about. And the blog was, there was a few components to it. One, we wanted to look at use cases and tools if they were out there. Two, I wanted to interview the entrepreneurs who are actually building the stuff because I was still trying to figure out what is the story. Like is, is AI real right now? Can we be integrating it into our businesses? What does that look like? Should I be thinking about it for our clients and bringing services to them around AI and, and the only way to do that was to go talk to the people building and try and figure out where are we? And I think 2016, in retrospect, we were very, very early, like adoption, honestly, in enterprises, as you Know, it's still extremely early. So back in 2016 it was almost non existent.

And then fast forward to today. What's the full view of the business? You have content, you do an event. What, what's the, what's the mix?

Yeah, so media, event and education are kind of the three pillars of the business model. So the media side is published content, build an audience. We have about 65,000, I would say subscribers, like opt in contacts, however you want to define them, that grows at about 3 to 4,000amonth. Now the top of the funnel is largely the blog. The podcast has, you know, really become a driver and I can get into kind of how that happened. It was sort of by accident that we built a massive growing podcast. And then we have an intro to AI class that I teach every three weeks. I started doing this in November of 2021, that it's free on Zoom. We've had almost 17,000 people register for that class. We've done it 33 times since November 2021. And that was again, this resiliency thing. What I told the team is like, we have to think about this like, like you're building a rock band. Like, we're going to show up and there's going to be 10 people in the room and I'm going to keep showing up every three weeks and I'm going to do this class until there's 300 and then there's 3,000. Like, and, and over time it. But Chat GPT just literally changed everything. We went from 300 on average to 900 on average in the months following ChatGPT.

Well, I mean, yes, you, you timing matters, right? Talk to any entrepreneur and it always seems like there's like somebody toiling away for seven years and then boom, that your moment happens and you're ready. Like you had to be ready to capture it, right? You, you had already been building. And so when the Chat GPT wave came out, you know, last two years or whatever it's been, you were ready, you were ready to capture it. But there's also something there. I love that from a speaking and content creation and learning standpoint, if, you know, I bet that you would. You have this talk that you give, you've probably edited and updated that 50 times because you give it to 10 people. You kind of feel the response. You okay, that slide needs to go here. We gotta tweak that thing. Like you just must. You, you get that flywheel going. And so just taking that mentality of like, I don't care if 10 people are here or 10,000. Like, while we grow this audience, I'm also trying to nail this pitch. Was that not a part of the learning process?

Yeah. And I thought, you know, so many of the things we do, we do because it just feels like the right direction. So, you know, the podcast again is a great example of this. The way what had happened was in October 2022, I felt our content strategy was getting kind of stale. We were doing a lot of republishing. HubSpot had come out with their study about, you know, refreshing old content and how it can drive a lot of traffic. And so we were playing, we were running that playbook. And so a lot of what we were doing on the blog was updating old stuff. And I just, I wasn't feeling it. And so I went to Mike, our chief content author, and I said, listen, let's, let's do the podcast. Because we had started it two years earlier and I just wasn't fulfilling it. I was doing like five episodes a year. So let's go to a weekly format and instead of you and I sharing links with each other on Zoom all week, let's curate those. And each week we'll pick three main topics. You interview me about those three main topics, then we'll have five to 10 rapid fire items of interesting things. I didn't even honestly know what the KPIs of a podcast were. I was literally just trying to drive a different content strategy for our company. And I thought, we'll take each episode, each of the three main topics becomes its own blog post. And then a fourth blog post that summarizes the whole thing. We'll embed the YouTube video. We'll, we'll start getting content on YouTube and it'll force us to synthesize what's happening in AI every week. And I'll put some stuff on LinkedIn as a relation to it. That was it. It was purely the podcast became the driver of the blog. And then ChatGPT hit like 45 days later. And all of a sudden the podcast became the platform. Like, everything started running off of that. And that was again, it was like this happy accident almost. But I mean, it's truly a pure B2B marketing story. Like everything we're doing is just a B2B play.

Well, it's awesome. I mean, clearly you know what you're doing from it, from you have experience with content and inbound to even have those ideas in the first place. But it also speaks to like, if you went in and I think people make this mistake with podcasting. And if you went in, you're like, here's the KPI. We want to hit 10,000 downloads by this month. Like, you would have approached it differently and it wouldn't have happened this way. You kind of had the right ingredients, which is this story, a clear methodology that you wanted to teach, and you were looking for a new platform to distribute that story.

Yeah. And it's. I mean, honestly, like, it's. I know it's had a big impact on our community. Like, we hear from people all the time about the podcast and how it's like the thing they have to listen to every week. But it's done as much for me and Mike, I think, because it really does force me, like, to keep up with everything I'm already reading and listening to and watching, but then to once a week, stop and think about it in a more. In a deeper way. And oftentimes, some of the most important ideas I have for the future comes serendipitously as Mike and I are having that conversation live each week. And it's like, oh, my gosh, you know what, Mike? I hadn't thought of this yet, but I think what this means is this.

Yeah.

And that may actually lead to, like, the next course I build or the next workshop we do, all because we're taking the time to actually have a conversation about it.

That's awesome. And you, you just get that. You have the feedback loop from so many channels. People are messaging you. They're going to reach out. You do a podcast about a topic and you're like, whoa, that was our most listened to podcast ever. There's clearly an appetite for that. That's what our next course should be about. Or that's what the next session should be about. So you, you do this weekly podcast. Are you two, like, sharing notes throughout the week? Do you have an agenda? Are you kind of creating a doc? Like, hey, here are the topics that we're going to cover this week.

Yeah. So we record it every Monday morning. And as soon as it's done, I start the sandbox for the next week. So, like, in Zoom, we'll have a podcast channel. And it just Sundays, like, episode 80. And then as I'm going throughout the week, reading things, looking on Twitter, whatever, I just grab links and drop them in there. And then on Sunday nights, Mike goes in, he curates those links. He selects what he thinks are the three main topics, and then the rapid fires. And then there's stuff that doesn't make the cut, that just goes in the newsletter, he sends it to me. I look at the brief, say, yeah, that looks good, or, no, let's move this to a main topic, whatever. Maybe takes me like three to five minutes. And then he goes in and he builds the brief. And he. He basically uses different AI tools. He'll take the article, summarize it, grab, you know, Q and A like, things to ask, that kind of stuff. And so he'll build a brief out, and then usually about 30 minutes before we record, I'll go in and look at the brief and I'll add any notes that I have. So he. He generally scripts his part of it. My part is completely unscripted.

What is the marketing AI guy? Where is he? Where are you researching? That's what I want to know. That's what. That's what the listeners want to know. What are you. What sources do you look at to build your. What's. What do you find worth paying attention to?

So I would say mine is a collection of podcasts I listen to, but I would say probably 90 to 95% of what I read and synthesize throughout the week comes from a highly curated list of alerts on Twitter. So I have a AI list that's public. You could go look at it. It's probably got about 480 or so AI people. It's entrepreneurs, researchers, VCs, journalists, like, anybody around AI that I've followed throughout the years. But then I actually get alerts from a select group of those people. And I want to say there's. There's no easy way to figure this out on Twitter. I can't go tell you the list of people I get alerts from, which is a crazy thing that that's not a feature in there. But I want to say there's probably 70 to 80 people and brands that I get notifications from the second they tweet. And so when you do that, and like, for a lot, for me, a lot of it is the top AI researchers like Sam Altman, Demis Asabas, Mustafa Solomon, like Dario Amade, like the leaders of the big AI research labs, Yann Lecun. If you follow those people and then who they retweet, you can pretty quickly build a knowledge base of, like, everyone who kind of matters in AI. And then you have to connect the dots between everything they're doing and saying. And that's kind of how I stay up on it.

You've talked about what. You've used this phrase, what matters in AI. Let's talk about what matters in AI for marketers today, specifically the marketers that listen to this podcast, which are majority of them are B2B marketers. Why does AI matter?

Because now it's accessible to everybody. So I think like, one of the things that gets lost in AI conversations is that AI itself isn't new. Like it's been around since the 1950s, this idea that we could give intelligence to machines or to software. But it went through like a lot of growing pains over the 70 plus years. But what happened is in 2011, there was a breakthrough in computer vision that gave machines the ability to recognize objects. Fei, Fei Li, who's at Stanford, her team kind of, you know, created this ImageNet competition. And then a couple people who eventually went to work for Google, Jeff Hinton and Ilya Sutskova, who's actually at OpenAI now, him and another guy created a tool that basically achieved like superhuman object recognition that was a breakthrough in deep learning. In 2011, in 2017, the Google Brain team invented what's called the Transformer. GPT stands for Generative Pre Trained Transformer. It gave us the ability to build language models. And so the breakthrough that really led to a lot of what we see today is the ability for machines, for AI to understand and generate language. And so when ChatGPT emerged in November 2022, all of a sudden anyone had access to it. So prior to that moment, it was hard for a marketer to look at AI and say, yeah, I totally get it. I can go use an image generation tool or a video generation tool or a language generation tool. But once we had generative AI, that a user with no coding or AI ability could show up and get value out of it, instant value. That was what created this kind of tipping point. And so now what we see is like, if you think about generative AI as a class, there's image, video, audio, code. All of these things are accessible right now. And so what we're seeing is enterprises that know these tools are there and they can go get it for $20 a month or $30 a month from, you know, OpenAI or Microsoft or Google. But very few companies actually know what to do with this or how to pick an AI writing platform or an image generation platform, things like that. So it's become real to where CEOs are demanding action. Boards are demanding actions of their CEOs. VCs won't give startups, tech companies any more money until they have an AI roadmap. Like it just became this thing that everyone needs to solve for, but no one seems to actually know what to do.

What are some people love, like the practical tactical examples. You're a marketer, you know, you're a CEO now, but you start a marketing agency. You built a marketing AI institute. What are some of the best or your favorite, like real examples of how AI can be used in a marketing capacity?

Yeah, we always, we teach people like a use case model, which is think about the things you do. So if you're, if you're an email marketer, if you do analytics, social media, whatever it is, think about the things you do every day, the tactical things, and look for ways to enhance it. So an example for us would be our podcast. As I mentioned, if we did every tactical thing each week manually with no AI, it would probably take 30 to 50 hours of work. We do it all in probably eight to 10 hours. And there's things we just wouldn't have previously done. So we use AI for, like I said, when Mike's building his script, he's using it. We do it for the transcription of the audio, we do it for summarization of that transcription. We then write four blog posts based on those summaries. We use AI to write the social media posts. We use AI to cut up the shorts for YouTube and TikTok. Like, it's infused into every aspect of the podcast. Probably 10 to 15 tasks of the weekly regimen for the podcast is done by AI or with heavy AI assistance. Descript is probably the primary tool we use within that, but there's a whole stack of tools that are integrated into it. So that's an example for us, which is something we do all the time, that we now use AI and infuse it into the workflow.

And what, what, like, man, I'm just listening to you now and I'm like, there's so much that I'm not doing with, with AI, even in our business. Like, what, what should we, what should we be doing? Is it like just the creating? You know, because I get, I get the value, I get it right. We could feed the podcast recordings and get a transcript, but like, what, what do I then do with that? How do I turn that into something useful versus just like another marketing, you know, thing that we did?

Yeah. So for me, again, it's all about like amplification and activation of a single asset. So, you know, if you think about any business, so I'm the CEO, obviously my time is generally limited in my ability to contribute to like writing blog posts. Like, I don't have three to five hours to done write Blog posts every week. But for one week, one hour each week, we have me dedicated to talk about what is happening in AI. All those points of view, all those perspectives that otherwise I probably wouldn't even tell my own team. It comes out in that one hour. So the question becomes, how do we maximize the value of that one hour and amplify it across as many channels as we can? Because it creates value for the audience, but it also builds the presence and profile of the brand. And so that's how we think about AI. It's like, okay, how do we take that one hour of transcript and turn it into as many pieces of content? Because we realize that it's not like when you and I were coming up. Like, it's not seven years ago where everything just goes on the blog and we just want to drive everyone to the blog. Consumer behavior is changing. The way people consume information is changing. And we have to evolve with that. We have to be on YouTube. We had to build a better presence there. We need to be on TikTok. Even though I don't personally care to use the platform, we need to be there. So we looked and said, okay, well how do we get all this content there? And so that's how we look at AI as a way to really help us do this. We have a team of seven people that includes me and our chief operating officer who don't really get involved much outside of me doing the podcast in the content side. And really the content team is basically one and a half people at this point. So how do we though scale? And if you look from the outside, I think it could appear as though we have a 10 plus person content team. And that's not it at all. Like, I don't. We probably spend less than 50 hours a month creating content, but it looks like way more than that. I used to charge clients $10,000 a month to create less content that we're creating.

Yeah, you should get back in that business. Don't tell them that. You know all the secrets now.

We were charging, I mean, we would do four blog posts a month packages for clients. And I want to say it was like a thousand to fifteen hundred per post or something like that. Just before blog posts. Like, what, what's the.

How do you get good? How do you get good? Like, I think there's, I think there's a difference between, you know, me going to chat GPT and being like, I'm interviewing this guy Paul, who runs the Marketing AI Institute, you know, help me come up with five questions. I Should ask him. I think that's become very standard. I mean, we've reached. We haven't reached peak AI adoption. But like, my, my father, who is 65, he emails me literally last week. He's like, dave, check this out. I'm interviewing. I'm interviewing a new guy for a new role. And I asked ChatGPT, Chad, GPT to write the interview questions. I'm like, man, we've reached. We've reached peak. Like, you know, tech. I don't even know what it's called. And I do think that a lot of people, I do think I've even find myself doing that. Like, I'm not getting the most out of AI other, you know, how do. What's the.

What's.

What are the steps there? Everyone says, oh, you just got to learn how to become a prompt engineer. I don't. I don't know what that means, though.

The weird thing about these models is they're not like normal software. Like, you know, in the software world, we all came up and it did what you told it to do, and it didn't really make mistakes unless there was a bug in the software. It just, you told it these five things and it executed them. These things don't work that way. Like, they'll make stuff up, they'll hallucinate. They'll like, kind of go off the rails a little bit. So they're not like traditional software. And so, as weird as it is, the simplest way to think about using these tools is to treat it like a really smart intern. Like, talk to it like a really smart intern. And so there's like, little tricks. Like, if you tell it who it is, it generally does better. Like if you say, I'll give you an example. I use this as like, an entrepreneur. I don't always want to have to call my attorney or my accountant when I have a question. So let's say I'm building, like, an equity program for our employees. I'll go in and say, you are an attorney that specializes in assisting, you know, entrepreneurs who run LLCs with building employee incentive programs. I want you to consider these five options and build pros and cons for them. And it'll actually, like, somehow or another. Even the people building this don't know exactly how it works. By telling it who it is and what exactly you want it to do, it kind of confines its output. It's. It like, zeros in its expertise on that topic. So I've done this like a friend who's a dentist. Like, you own a Dental practice. Your profits are down because your costs are skyrocketing and insurance controls your pricing. How can you maximize profit using AI? And it'll, it'll do it, it'll build these things. And so it's really just developing this way to talk to it as though you were giving an assignment to a human and you're very clear on what you want it to do. So that's just kind of like one trick to do it. Other than that it is really following best practices and like guidance. Like Ethan Malik, I follow him. He's a Wharton Business School professor. He shares a ton of like really practical takeaways on how to do this stuff. And so I just, I kind of follow those people.

So I'm just thinking of like two projects that I'm currently working on that I normally would just think and brainstorm and research and write. Which is number one is I want to add a new tier to our Exit 5 community. Like there's a strong pull towards people who want to be future CMOs and we're thinking about like carving off something for that and doing that. Could I use a tool like ChatGPT and write some smart prompts to like help me figure out the packaging and, and what should be in that?

For sure. Yeah. And what I would do there is I would give it like you're an entrepreneur, you run a fast growing media company. My audience is this again, it's like writing a strategy brief as a marketing leader. Basically this is what I want you to do. I want you to compare the pros and cons. Take it step by step. It often works better if you tell it to take its time and think through step by step. The other thing you can do is tell me what questions do you need to ask to get to the final product? What questions should I be asking of you? And so often it helps to think. Think of it as not a single prompt and a single output, but an ongoing conversation where there's questions back and forth. The other thing I will do with, with a use case like that is I will have ChatGPT open the paid edition because you want GPT4.

I was just going to ask you that. So, so sorry to cut you off. Is there going to be a difference in the quality of like I need to go and I need to use the paid version to get the. The better answers?

Yes, a hundred percent. Anything that deals with more advanced reasoning and like chain of thought, feedback, thinking and like you like strategy, I would always use GPT4. It's by far still the best model we have in humanity right now. That may change within a month or two, but at the moment that is the most powerful.

And considering how powerful that is kind of a joke. $20 a month.

It's the greatest value in software history.

And can I get images? Like, can I get better images built into it?

Cool. So what I would do though is I would have anthropic Claude 2, I would have GPT 4 through ChatGPT. You can play around with Google Bard. It'll get better in the coming months. Inflection pie, like I would, I would test it on a couple of them or a few of them, give it the same prompt and evaluate the output. Generally speaking, my experience is GPT4 is almost always the best output. Claude from Anthropic is pretty good and it's getting better. Google Bard will get better once they fix the underlying language model.

Oh man, this is great.

This is.

I can't wait till, I can't wait till my exit 5 team listens to this.

When you build courses, I found it's really good at building like abstracts for courses and like online courses and even for our conferences. We have our Marketing AI conference. We also have three virtual summits. I use it to assist me in building the agendas, coming up with questions to ask or what would be good outputs for, you know, this, this topic, things like that. Cool.

Man, this is so good. Three months from now I'm having you back on. We need more time. So what, what's something that you see? A myth. A myth, an objection, A false belief that you see, you know, people writing on LinkedIn or on Twitter, specifically in the B2B marketing space that you want to correct or wish you could shape that, that maybe something that they believe is true is not, you know, AI is scary. AI is going to take my job or AI can't do this or the quality is crappy. Some, some myth that you see that you want to myth bust.

Yeah, I mean, I, I do think it's scary, honestly. Like I, I think it's going to be very disruptive knowledge work. I think it's going to be a, there's a very good possibility it's going to be a painful few years as the newer models come out. And it, it, it affects, I mean there's a hundred million US workers that are knowledge workers, roughly. That's doctors, lawyers, accountants, attorneys, marketers, consultants, whatever. It's all of them. And I think it's going to be disruptive, not replace roles. Like it's not going to just say, okay, we just don't need this marketer anymore, can't do 100% of their work. But the challenge becomes do we need as many of them? So if you're an accounting firm and you only have so many tax returns you can do each year and you can do them 30% more efficiently, you just need fewer people doing those, those results, basically. So that's, I think, the challenge we're going to face, and that's unknown where that goes, but I think we need to be preparing for disruption. The thing that usually I think about as a myth is that it's overhyped and, and my feeling there is it's completely under hyped. Still like it because if you think about the adoption in these enterprises, it's so early. Like I still have yet to talk to an enterprise and I've talked to some of the biggest companies in the world that have a change management plan in place for if this stuff works the way it's supposed to work. If you go get ChatGPT teams for your company or ChatGPT enterprise, you get Microsoft 365 copilot and it actually does what it's supposed to be able to do. We're talking about a complete restructuring of how organizations work. Their staffing, their tech stacks, everything. Nobody is ready for that. And all the stock prices, by the way, there's almost zero enterprise adoption. When you look at Google and Nvidia and Amazon and Microsoft and people think they're like overvalued in some cases, I just think. But we haven't even hit the adoption curve yet for what they're doing.

I don't want to talk about the scary stuff with you because that's not what this podcast is about and there's not enough time for that.

That's a whole other topic as a.

Marketer, like just, man, I'm even fired up just from the way you're telling me to rethink prompts and obviously there's. I haven't gone through any of your courses and training and there's a huge opportunity for me to do that. I just haven't been in the right. I've. If I was like going super hard on like marketer Dave use case, I would be. I need to. I'm learning, man. There's a lot of inefficiencies with my business that I should be thinking about this stuff for. But is this a trend? Like, you've seen inbound, you've seen mobile, mobile, social, local, whatever, all that stuff was this Seems to be the thing that if you're, if you're in marketing, you need to, I don't want to say like, bet your career on, but this seems to be the thing you got to bring along with you that's going to turn you into a superpower. Am I wrong about that?

No, I, I think, like I wrote a few years back that, you know, we were pretty early obviously with HubSpot and that I thought this was 10 to 100 times the opportunity that Inbound was back then. I haven't changed on that. So I'll tell you the quick 10.

This is 10 to 100 times easily.

Yeah, I mean, it's not even, I'm.

Probably as far as impact on marketing.

Everything, marketing, business, the economy, all of it, because it's going to be the underlying thing that's infused into everything. So if you think about, like, I'll often say that I think at least 80% of what knowledge workers do will be AI assisted to some degree in the next one to two years. Now, the reason I'm fairly confident in that assumption is because if you look at HubSpot and Salesforce and Microsoft and Google and Adobe and Oracle, like every major tech company that we use to do our jobs thinks AI is everything moving forward, they're betting everything on it. So every piece of software you use, they're infusing AI into it. You're not going to be able to do your job without AI. It's kind of like your personal life. Like when you use Netflix or Spotify or Amazon or a Tesla or Google Maps or like any of these things you use in your life, none of them are possible without AI. So you're using AI every time you use these tools, every time you go into your social media feed and it curates the news feed for you, it's all AI. That's what's going to happen in marketing, sales, service operations, all aspects of business. The software is just going to get smarter because these companies have to build the money, have to build it into it. If you're a startup or if you've got a seed round or a series A, you're not getting the next round of funding without an AI roadmap for your products. So that's why I think, like, it's just, it's going to be everywhere.

No, this is awesome. I'm taking so much. I have so many notes about this and I was just trying to write the headline and then I'm going to run it against ChatGPT after. So real quick Before I let you go, want to do a quick plug for your upgrade? You're. You're doing this AI for B2B marketer summit in June, which I think is worth talking about before we wrap up.

Yeah, that's a new one. So we have our AI for writers summit we did last year. We had 4,200 people register at the inaugural event. It was kind of crazy. It was a virtual event. We do our marketing AI conference in person in Cleveland. We had 700 last year. We're thinking between 1500 and 2000. This year we have an AI for agency summit. That's a virtual summit in the fall. And the AI for B2B marketers is the new one we're doing this year in June. Awesome.

I think it's a great resource and everything over there. Marketing. Marketing AI Institute. You go to marketingai institute.com. we'll have that in the show notes. I'll have the link to this in the show notes. And you can go to LinkedIn and connect and follow Paul over there. Paul, thank you so much. You've obviously been deep in this space for a while and it shows. I'm excited to continue to follow you, continue to learn more, and I think take this as an opportunity to, hey, how can I get more efficient in my business? And I hope everybody that's listening will find some new creative use cases for AI in their business or at least just take it a little bit more seriously moving forward. Thanks for joining us, Paul.

I appreciate it. Dave, this is awesome.

All right, man. Thank you. Exit.