Welcome to Episode 150 of The Artificial Intelligence Show—a special milestone that marks the launch of a brand-new series: AI Answers. In this episode, Paul Roetzer is joined by Cathy McPhillips to debut a fresh format to systematically answer the best questions we get during our live AI education sessions.
Over the past few years, our free Intro to AI and Scaling AI classes have attracted more than 32,000 learners—and they’ve asked hundreds of smart, tough, practical questions. This new series tackles them head-on.
Listen or watch below—and see below for show notes and the transcript.
Listen Now
Watch the Video
What Is AI Answers?
AI Answers is a biweekly bonus series that curates and answers real questions from attendees of our live events. Each episode focuses on the key concerns, challenges, and curiosities facing professionals and teams trying to understand and apply AI in their organizations.
In this first episode, we address 19 of the most important questions from the May 15 Scaling AI class, covering everything from tooling decisions to team training to long-term strategy. Paul answers each question in real time—unscripted and unfiltered—just like we do live.
We organized them into five key areas: shifting mindsets around AI, making smart tool and strategy decisions, building internal training and leadership support, preparing for the future of work (and helping the next generation do the same), and looking ahead at what’s coming—from custom GPTs to the future of search.
Whether you're just getting started or scaling fast, these are answers that can benefit you and your team.
Timestamps
Introducing the Series
00:00:00 — Intro
Mindsets, Myths and Momentum
00:08:32 — Question #1: How do you explain AI as a tool for transformation to someone who’s unfamiliar or maybe even a little afraid?
00:10:44 — Question #2: Do you see learning to use AI effectively as the modern version of learning to type?
00:13:03 — Question #3: How realistic is it to create an actual AI roadmap?
00:16:29 — Question #4: Once you build a roadmap, should it be shared with the entire team?
Strategy, Tools, and Tough Decisions
00:18:48 — Question #5: Is it better to invest in ChatGPT or Microsoft Copilot?
00:20:22 — Question #6: How do you make the case to leadership that a paid license to ChatGPT is worth it?
00:22:03 — Question #7: I’m using multiple AI tools—but each one only does a few things well, and the costs are adding up. How do I better train and support my agents so the company becomes more AI-forward without overwhelming them?
00:25:49 — Question #8: In two years, how many GenAI platforms do you think will dominate the enterprise landscape?
00:27:40 — Question #9: Do you have any thoughts or concerns around using open-source LLMs in the enterprise AI stack?
Learning, Literacy and Leadership
00:30:39 — Question #10: How involved should the CEO be with an AI council? What kind of role makes the most impact?
00:33:25 — Question #11: Once you have an AI policy, where should you begin to use it to educate your team?
00:35:28 — Question #12: What’s a solid KPI to track AI literacy or adoption?
00:38:42 — Question #13: If you were building MAII from scratch, with what you know now—what would you do differently?
Careers, Custom GPTs, and the Next Generation
00:41:19 — Question #14: How do you actually bridge the gap between current capabilities and future roles? What’s the smart move for career future-proofing?
00:49:15 — Question #15: What courses should kids in school be thinking about if they want to be prepared for an AI-infused world?
00:53:20 — Question #16: What are three things you’d suggest for helping teenagers use AI to accelerate learning, without just relying on it to do the work for them?
Agents, Search, Industry Predictions
00:56:07 — Question #17: Is it better to create a specific GPT for each job task, or one mega-GPT that does content, strategy, internal reports, sales writing—all of it?
00:59:09 — Question #18: What do you think AI will do to the search marketing industry, especially paid search?
01:01:15 — Question #19: What excites you about AI?
This episode is also brought to you by the AI for B2B Marketers Summit. Join us on Thursday, June 5th at 12 PM ET, and learn real-world strategies on how to use AI to grow better, create smarter content, build stronger customer relationships, and much more.
Thanks to our sponsors, there’s even a free ticket option. See the full lineup and register now at www.b2bsummit.ai.
Read the Transcription
Disclaimer: This transcription was written by AI, thanks to Descript, and has not been edited for content.
[00:00:00] Paul Roetzer: I feel like we finally have arrived at a point where companies, and more largely, society is, is just understanding the moment we find ourselves in and the significance of what's happening. And so there's far more, urgency from people to play a role in this, to have some agency in what happens next.
[00:00:21] Welcome to AI Answers, a special Q&A series from the Artificial Intelligence Show. I'm Paul Roetzer, founder and CEO of SmarterX and Marketing AI Institute. Every time we host our live virtual events and online classes, we get dozens of great questions from business leaders and practitioners. Who are navigating this fast moving world of ai, but we never have enough time to get to all of them.
[00:00:45] So we created the AI Answers Series to address more of these questions and share real time insights into the topics and challenges professionals like you are facing. Whether you're just starting your AI journey or already putting it to work in your organization. [00:01:00] These are the practical insights, use cases, and strategies you need to grow smarter.
[00:01:05] Let's explore AI together.
[00:01:11] welcome to episode 150 of the Artificial Intelligence Show. I'm your host, Paul Roetzer along a special co-host today. If you're a regular listener to the Artificial Intelligence Show, I would usually introduce Mike Kaput this time, but this is the first episode in a new series we are doing called AI Answers.
[00:01:32] So Mike will be back every Tuesday. Don't worry, the weekly format isn't changing. This is a new series, as I said called AI Answers, and the premise here is, every Cathy do two free classes together. We do Intro to ai, which we started in fall 2021. Cathy if I'm mistaken, we have had over 32,000 people register for that class.
[00:01:56] We have done 40, how many Cathy [00:02:00] 40. Nine I think around 49. That sounds about right. and then we also do a scaling AI class every month for free. And these are both run through Zoom webinars. You can register for either of them, both of them, whatever you'd like. We have people who come every time. It's, it's kind of wild to us.
[00:02:18] But, those two classes are a key part of our AI literacy project. So we're constantly trying to find ways to accelerate AI literacy for as many people across as many career paths and industries as possible. And so the intro class, which the next one is, June 10th. So we'll put it, we'll put links in the show notes.
[00:02:40] You can go and check these out. for that one, the whole idea is to try and provide like this very fundamental understanding of AI very quickly. So like 30 minutes is the class I present for, and then Cathy and I do q and a for 30 minutes. Well, we normally will get, I don't know, 12 to 1500 people [00:03:00] registered each class, and we will get dozens of questions, sometimes over a hundred questions.
[00:03:06] And they're all amazing and we usually get to like five to seven of them. And then the same thing happens with scaling AI each month. That one we get maybe five to 800 people each month that come to that class or register for that class. And then again, dozens of great questions. Now, scaling AI tends to be more for like director level and above.
[00:03:24] It's definitely more of a strategic approach. It's like five steps to scaling AI in a company. but again, dozens of questions and we get to maybe five to seven of those questions in the time allotted. So we had this idea to take those two classes as well as our virtual events. So we have like our a AI for B2B marketer Summit is coming up June 5th.
[00:03:45] we will get hundreds of questions 'cause that one's gonna have thousands of people registered for it. And so we have all of these questions, which one we want to be able to answer as many as possible. Thus the idea for this series, but two. What we realize is [00:04:00] each of these live events is a window into what people are thinking about the challenges that they're having with AI adoption, the pain points they have, the strategic questions they have.
[00:04:12] And so by doing these AI Answers episodes, which we're probably, we're planning on about two a month, so the idea is we, we do an intro to ai and then the next week we answer questions from that one. The, we do a scaling ai. The next week we answer questions to that one. So this one today, episode one 50 is based on the May 15th scaling AI class.
[00:04:32] So we have taken questions from that course, and we've curated them. I have not actually looked at them. I did this, the way we do it live, which is Cathy picks the questions and asks them to me. And so that's kind of how we're gonna roll here, is it's gonna be like real time. Cathy asks me things and we do our best to answer as many as we can.
[00:04:50] So Cathy has curated, I don't know, about 20, 25 questions that sound about right, Cathy
[00:04:56] Cathy McPhillips: I boil it down to 19.
[00:04:57] Paul Roetzer: Okay,
[00:04:58] we're down to 19. We'll see if that [00:05:00] expands or contracts as we go. but I'm gonna do my best in like, maybe, you know, one to two minutes. Try and be as high level as I can and just get through as many of these as possible.
[00:05:08] So, again, that's the whole premise of the series. Our weekly episodes aren't going anywhere. Mike, and I'll continue to be with you every Tuesday doing the weekly. but this gives some additional content for everyone and hopefully just a ton more value because for Cathy and I, we both say this all the time, our favorite part of doing those classes each month is hearing the questions that people have because it does just truly give us a snapshot into where we are overall, just in terms of understanding and adoption of ai.
[00:05:36] And to watch those questions evolve over time is so fascinating. Like I, Cathy you see more than I do, but like there, I mean, you start getting questions around like the environmental impact now and geopolitical stuff. And like six months ago that wasn't even on people's minds. And now it just is like, it's totally evolving.
[00:05:53] So anything.
[00:05:54] Cathy McPhillips: on, you know, the audience mix that day.
[00:05:57] Paul Roetzer: Yeah,
[00:05:57] Cathy McPhillips: on what happened in the
[00:05:59] Paul Roetzer: it's [00:06:00] true. What?
[00:06:00] Cathy McPhillips: know. So there's so many variables that come into play when the questions are come out. It's like, that's such a great question. And sometimes we use them in our daily Slack, like question of the day just to get other answers from other community members. But there's still dozens, we don't, that don't get answered.
[00:06:14] Paul Roetzer: Yep.
[00:06:15] Cathy McPhillips: this is exci exciting way for us to connect with our community and say, Hey, we asked your, or we answered your question on the podcast.
[00:06:20] Paul Roetzer: Yeah, and that's true. We'll do our best to like, let people know, but you know, again, and these will be, you know, available. It's, this is on, if you're a regular listener on the podcast networks, apple Podcast, Spotify, we also have these on YouTube. So we publish all of our podcast episodes on YouTube.
[00:06:37] Claire and our team does a great job of cutting up, clips so you don't have to watch the whole thing necessarily. She puts 'em into shorts and breaks 'em off into separate videos. So yeah, we, we try and do this multimedia and wherever you're at, we want to kind of make the, have the content meet you where you are.
[00:06:52] So. Yeah, always go back and check 'em out. and again, if this is your first time listening to the podcast, join us for the Weekly, every Tuesday. We've been [00:07:00] doing that since, October-ish of 2022. We launched the Weekly right before ChatGPT. So we are, you know, we've done 140 some episodes probably of that.
[00:07:11] So, we do this all the time and, we appreciate everybody who's joining us for the first time and all of our loyal list, loyal listeners who are back. So with that, Cathy unless you got something else to add up front, I would say let's,
[00:07:23] Cathy McPhillips: that Claire, you know, this was Claire's great idea to do this.
[00:07:26] Paul Roetzer: Yep.
[00:07:26] Cathy McPhillips: So she, after intro, after scaling last week, she's like, let's use these questions. So she
[00:07:31] Paul Roetzer: Yeah.
[00:07:31] Cathy McPhillips: this idea on how to do this, how to format it, and then the other day I took all the questions through Claire's formula. Put them into AI and said, okay, help me prioritize these, help me do these in an order so it makes sense for the listener. So it's like segueing from one thing to the next. And then I went through, as we always do, and made sure that it was actually correct, you know, and that it did flow
[00:07:51] Paul Roetzer: Cool.
[00:07:51] Cathy McPhillips: and did a little bit of cleanup. and I feel guilty taking episode one 50, a milestone away from Mike Kaput, but I'm happy to be here.
[00:07:58] Paul Roetzer: Well, that's funny. Like we had said, I don't [00:08:00] know, a few episodes back, we were like, oh, we should probably do something like special for episode one 50. And then it just works out that this idea sort of came up and it's like, all right, let's do it now. And oh, that would be episode one 50. It's like, all right, great.
[00:08:12] Launching a new series is I guess a good way to spend episode one 50.
[00:08:15] Cathy McPhillips: Mike gets 200,
[00:08:16] Paul Roetzer: There you go.
[00:08:17] go.
[00:08:18] Cathy McPhillips: Okay, so I broke this up into five different sections, mindset, tools, team building, future proofing, and then where this is all headed. So that's kind of the block of questions I, was thinking about. So let's jump in,
[00:08:31] Paul Roetzer: Cool.
[00:08:32] Question 1
[00:08:32] Cathy McPhillips: Starting with mindset, you know, as much as we, as we talk about capabilities. Fear and misunderstanding are all like huge blockers in this. So how do you explain AI as a tool for transformation, not just another shiny tick to someone who's unfamiliar with it or even a little bit afraid or wary of this?
[00:08:49] Paul Roetzer: Yeah, this gets into that bigger idea of change management that we talk about so often and not allowing AI to fall into the tech and data silos of the company. you [00:09:00] know, really understanding the overall impact this can have, not only at the organizational level, but down to the individuals and their teams and the departments overall.
[00:09:08] So, you know, it really is just taking a strategic approach. That's part of the thought process behind our scaling AI class is like teaching five steps. And so, you know, going through and building an AI council, developing generative AI policies, responsible AI principles, doing impact assessments, building a roadmap, building an internal AI academy, so to, to truly approach it from a transformational standpoint.
[00:09:31] You have. To have a framework for how to think about that transformation moving forward. and so that would be, to me, like the first step is set some, some very tangible, elements to the plan and then a realistic timeline to get started.
[00:09:47] Cathy McPhillips: That's one of the things I love about Katie Robbert session at the ad, at the AI for B2B marketer Summit coming up. It's all about, okay, you can do all of these things, but if there's not someone in in that spot or some change change agent or [00:10:00] whoever in the organization, it's not gonna go anywhere. So I'm really excited for her session to hear what she has to say about this.
[00:10:05] Paul Roetzer: Yeah. Someone to own it. And then like if it's not an executive that's owning it, that you need an executive sponsor that's gonna give them the kind of support and resources they need to see that through.
[00:10:14] Cathy McPhillips: Yeah, and sometimes a third party is the one that kind of cracks that nut when people are, you know, not really
[00:10:20] listening,
[00:10:20] you know, getting that outside perspective to come in and do
[00:10:22] Paul Roetzer: Yeah, some, someone to come in and, yeah. And I understand, like, I get the frustration, especially for people that work at big brands, it's like, you could be saying this from the rooftop for six months to, you know, a year or more, and then you get the consultant to come in and say the exact same thing and all of a sudden, you know, everything, everything starts moving.
[00:10:39] So I empathize with that. But you, you know, sometimes you just gotta do whatever you gotta do to get the ball rolling.
[00:10:44] Question #2
[00:10:44] Cathy McPhillips: Alright, here's another one that stuck with me. Do you see learning AI learning to use AI effectively as the modern version of learning to type? Like are we headed toward a world where not having AI skills leaves you behind?
[00:10:56] Paul Roetzer: I really like this one. It's, super practical [00:11:00] in terms of like, some advice I've been giving of late, which is, yeah, I mean, I think on resumes when you're doing interviews, it really is where you're starting to say like, okay, what's your familiarity with using like, you know, Microsoft Word or like, project management tools.
[00:11:13] Like, it's just these fundamental skills that we require. Of any professional in any career. and I do think that learning to work with these tools, that learning to prompt that, those are just gonna become very fundamental to every job right now. I think the challenge is a lot of maybe the hiring professionals or the executives that, you know, guide, which, you know, the kind of people they want to bring in, they don't know to build this into what they're looking for in job descriptions and skillset.
[00:11:41] But I do think that within the next year, across most industries, familiarity with these chatbots, the, you know, your prompting ability, those things are, are going to matter and just be required for any job.
[00:11:55] Cathy McPhillips: And conversely, you've even said, you know, if the company that you're interviewing with doesn't [00:12:00] ask those sorts of questions, you might wanna reconsider actually sticking with that.
[00:12:03] Paul Roetzer: Yeah, and again, I didn't look at any of these questions, so I don't know if we're gonna get to any of this stuff in advance, but that is one of the things we say a lot, which is, you know, up until now it's been okay to work for a company that maybe didn't give you access to ChatGPT or Gemini or, or didn't train you on these things.
[00:12:21] but moving forward, you're gonna start to put yourself behind your peers if you're at an organization where you cannot use AI in your job. And so I do think not only is it important that we look for these skills as employers, but as individuals who have these skills or developing these skills, you're going to want to work for organizations that allow you to apply these skills and thrive and not spend the next few years not learning these things.
[00:12:49] I mean, it would, the best, you know, kind of closest analogy we've talked about is like going back to the early two thousands and not allowing people internet access to do their jobs. Like [00:13:00] just. You would, you would fall behind if you didn't know how to use the internet?
[00:13:03] Question #3
[00:13:03] Cathy McPhillips: For sure. Okay. If AI is moving and evolving as fast as it feels, which it feels fast, changing itself and everything around it, how realistic is it to create an actual AI roadmap or a company's just stuck in reaction mode?
[00:13:18] Paul Roetzer: So I actually think it's moving faster than most people realize. There are weeks, like last week, you know, we talked about this on episode 1 49 of the podcast where I'm not sure that I fully grasp, I. How fast things changed last week and like what that means for the next 12 to 18 months. So I know it feels fast to everyone, but I trust me, like it's actually probably even faster than we're aware of at the moment.
[00:13:44] So in terms of an AI roadmap, my basic premise there, and this is what I teach in one of the courses as part of our AI Academy, is it, I think of roadmaps as like one to two years where you're starting to lay out priority projects and priority [00:14:00] problems to solve. But the core of the roadmap is having a vision for the type of organization or department.
[00:14:05] You want to build this AI forward mentality where everything is always being evaluated. Like, can we do this smarter? Is there a more intelligent process here? And so I think like you look at them in one to two year increments where you're starting to prioritize. So you don't try and tackle too much at once, but then it's really probably like quarter to quarter and you want to be constant re reevaluating, saying, okay, based on the new capabilities of OpenAI or or Gemini from Google, does that change the project we had prioritized for Q3 of this year, it looks like that may have actually been obsoleted and now our current chat bot already does that.
[00:14:43] And so now we don't need to. So based on this, like the way to assume is a new frontier model will move to the leader of the pack every three to six months moving forward. So whatever today's current state of the art is, which would probably be Gemini 2.5 Pro from [00:15:00] most benchmarks. Realistically, GPT five will probably come out sometime this summer and that would likely then leapfrog and then Gemini three will come out and then, and it's, it's just gonna keep moving this way.
[00:15:12] And a lot of times you can ignore it because it's not gonna change your workflows in life. But sometimes entirely new paradigms shift where like a reasoning model shows up and it's like, wait a second, this just changes things. Or like the deep research product from, you know, Google or OpenAI. And so you do have to be dynamic with these roadmaps.
[00:15:33] Cathy McPhillips: Yeah, and they don't even need to be that robust. To begin,
[00:15:38] Paul Roetzer: Yeah,
[00:15:38] Cathy McPhillips: just get started on like, what are some things we can tackle now and what's the implications of that and what, and what do we do next? I mean, we've done that with some workshop, clients that we've worked with that it's like, you don't, it doesn't need to be this robust thing from the start.
[00:15:51] Just get something on paper, have it be fluid and just keep iterating
[00:15:55] Paul Roetzer: definitely. Yeah. A lot of times simplifying it is the way to go until you really start [00:16:00] getting. Traction within the organization. You've dealt with all the, you know, the people in the company who maybe just don't wanna do this stuff, for different reasons. You know, fear of losing their job. It's abstract.
[00:16:11] Like you gotta solve for that before you're truly gonna scale. And that's why we always talk about like, AI literacy is always the first step in an enterprise when you're trying to drive AI adoption and change management. Because until everyone understands what it truly is and the value it can bring, you're just never gonna get where you want to go.
[00:16:29] Question 4
[00:16:29] Cathy McPhillips: For sure. Okay, so you have this AI roadmap, right? So who should know about it? Should the entire team know about it, keep it with the senior leaders, or is there a best practice on how to roll that out to the team?
[00:16:40] Paul Roetzer: Oh, I, so I'm a huge believer in total transparency and democratization of AI within organizations. And part of the reason is because I think a lot of the best uses of ai, the most high, the highest impact uses are gonna be developed by the practitioners who are actually in these tools every day and [00:17:00] finding great prompts and use cases to apply 'em to.
[00:17:02] So, no, I think the roadmap process needs to be, involving people in it from the start, because so much of the roadmap is personalization of use cases. So if you go get like ChatGPT team or enterprise or, you know, get Gemini for everybody, you wanna like, involve them and find three to five use cases where they can build some Google gems or some, you know, custom GPTs to help them in their job.
[00:17:30] So, yeah, I. A hundred percent would like think about the roadmap as a collective experience for everybody in the department or organization?
[00:17:38] Cathy McPhillips: And there is likely some folks that are more well versed in this than you realize who might be able to add some value to
[00:17:44] Paul Roetzer: Definitely. Yep. Yeah, and I think I, so much of what we do is all about empowering people. This is why, you know, we have a consulting practice under our Smarter X brand, but we aren't fully scaling it because we look more at [00:18:00] empowering people through education and events in a one-to-many format as being more impactful right now.
[00:18:06] And so a lot of times we just guide organizations that reach out to us and wanna work together. It's like, no, like let's get you through the courses, let's get you people to the event. Let's run a workshop. Because for me it's the more you empower the leaders and the practitioners to do this themselves, then you actually enable them through this dynamic phase we're going through where the tech keeps getting smarter all the time.
[00:18:28] Unless they understand it themselves and can find their own use cases and continually to improve, then they're always just gonna be relying on the outside consultants to do it. And I don't think that's the way, you know, you get ahead here. It's not how you build an AI forward organization. So yeah, I always about empowering everybody at all levels to do this in a responsible way.
[00:18:48] Question #5
[00:18:48] Cathy McPhillips: for sure. Okay. Jumping into some tools and some strategy, this one came up a lot for, small to medium sized businesses. Is it better to invest in ChatGPT or Microsoft copilot, [00:19:00] especially when your users range from basic to advanced and some may benefit from agents.
[00:19:04] Paul Roetzer: So, I don't personally have experience using copilot. Anybody listens to our podcast, knows that we, we talk quite a bit about po copilot, but it's usually just anecdotal based on what we're seeing, reading, hearing experiences. I can speak to ChatGPT is a great experience, is what we use for our teams, but we also use Gemini and Google Workspace.
[00:19:25] So we have a collection of tools. I would say that from our state of marketing AI report that we just released earlier in May, ChatGPT is the dominant choice, especially for small to mid-size businesses, and individuals. So. Yeah, I, my general guidance is like, if you choose ChatGPT or Google GeminI don't know that you can go wrong, honestly.
[00:19:51] And Microsoft Copilot is built on top of chat GT's models. it's just like a, a customized experience and user interface and [00:20:00] stuff. So, yeah, I don't know. I think you can, I don't know that you can go wrong with any of 'em. You just gotta commit to really using them every day, experimenting with them, and finding valuable use cases.
[00:20:10] Too many organizations just buy one of these things for their teams, and then don't train them how to use it in personalized, you know, use cases for them. So, I, my general guidance is ChatGPT is a safe bet.
[00:20:22] Question #6
[00:20:22] Cathy McPhillips: Yep. So speaking of ChatGPT, someone actually left this comment. my CEO won't pay for the license because there's a free version. So how do you make the case to leadership that a paid license for the team or for an individual is worth it?
[00:20:37] Paul Roetzer: I would first demonstrate. Super practical use cases that show the CEO, the impact chatGPT has or can have. time saved, increased productivity output, increased growth, improved decision making, like whatever the CEO [00:21:00] responds to you, you know, those kind of triggers. I would talk in that manner. And then once you've created.
[00:21:07] The value perception of what it enables. You know, I went from doing this project in 30 hours, I now do it in three hours. So I'm actually able to do these 10 other things I wasn't able to do before. Like keep it simple, but like show the value and then pull in the, by having a business account, here's what we get.
[00:21:25] Privacy, security, custom GPT sharing across teams and like, what does that mean? You just gotta make that business case in the language that matters to your CEO. And each of us have, you know, each CEO is different in terms of what matters. It sounds like this might be the example of a CEO that is probably very numbers driven if he or she's not willing to spend $20 a month.
[00:21:49] so I would talk in numbers most likely and show the impact it can have on revenue or costs.
[00:21:55] Cathy McPhillips: Yeah, that could get paid for, for a year license in a hot second, like
[00:21:59] Paul Roetzer: [00:22:00] Yep.
[00:22:00] Cathy McPhillips: just one example. One. One use case.
[00:22:02] Paul Roetzer: Hundred percent.
[00:22:03] Question #7
[00:22:03] Cathy McPhillips: Okay. A practical question that I really liked. I'm using multiple AI tools, but each only does a few things well, and the costs are adding up. How do I better train and support my agents so the company becomes more AI forward without overwhelming them.
[00:22:16] Paul Roetzer: Yeah, so I've found myself, I don't like I'm the CEO, I don't, I don't mind spending money on different tools, 20 bucks here, 20 bucks there. But I have definitely found myself, yeah, I think last month I got rid of perplexity. It's like I haven't logged into perplexity in three months. Like I just got rid of it.
[00:22:36] I think I got rid of my Claude license a week or two ago. And again, not because the 20 bucks really bothers me, it was just noise and like. What I was finding is every once in a while I would duck in and I would try something in perplexity or Claude, and maybe it would be like incrementally better for one use case.
[00:22:54] But then like I would go to find the thread. I'm like, God, what chat bot did I do that in? Like I can't even [00:23:00] remember if it was my personal account, my business account, my ChatGPT. And so I'm finding like for simplicity sake. I get 99% of what I need from Google Gemini and ChatGPT. And so I'm centering my uses there and then trying to solve as much as I can through custom GPTs and gems.
[00:23:20] And now with like the coding ability that's baked into these things, like I'm finding I need less and less tools. And so for my own personal adoption and what I, what I would hope to guide our team to do is maximize one or two of the tools. And then if we need specialized use cases that require, like descrip, we we're not getting rid of Descript.
[00:23:41] Like Descript is fundamental to what our team does, for audio, video production, the podcast, webinars, cloud, all that stuff. So that is like a, we absolutely are gonna keep, you know. Pushing on that and learning all those tools. HubSpot for us here, I'm like, we're gonna have these other tools, but when I think about our core [00:24:00] chatbots, they're able to increasingly do a lot of general things very well.
[00:24:04] And so hopefully we can, you know, reduce our tech stack there.
[00:24:07] Cathy McPhillips: Okay, so not everyone is a CEO like you, who is very willing for us to go in and try new tools. So, you know, is that the best place to start? The best pitch is like, get the ChatGPT or a copilot and start there. I.
[00:24:19] Paul Roetzer: I do think so. Yeah. I think finding your core chatbot and then spending a lot of time with it and finding its full capabilities. Is gonna solve a lot of, op you know, challenges. Open up a lot of use cases and create a ton of value. So that is definitely, like I recently posted about that on LinkedIn.
[00:24:39] I'm actually gonna probably work that into my B2B summit opening talk. I do think that if you just pick a chat bot and then you get really good at prompting with it and then learn to build a few gems and GPTs and then use a couple of the ancillary features that are baked right in like notebook, lm, and deep research, like I think you have solved.
[00:24:59] More than [00:25:00] 95% of your peers like what they're able to do with these things. And so, yeah, if you're a graphic designer or a video producer, like you're gonna have specialized tools that these things can't solve for. But for most knowledge workers, regardless of your profession, that on its own is going to dramatically increase your efficiency and productivity and your creativity, and thereby your ability to make an impact on growth and innovation and things.
[00:25:27] So I think that's for most people where you start.
[00:25:31] Cathy McPhillips: Yeah, I think your point about, you know, we need to go in and use these tools and if you're, you know, the tools are so similar. I mean, obviously very different, but they're so similar that it really, since the beginning of time, tech's only as good as the person using it. So if we are mastering certain tools, let's just go with
[00:25:47] Paul Roetzer: Yeah. Yep.
[00:25:47] Cathy McPhillips: with that.
[00:25:48] Yep.
[00:25:49] Question #8
[00:25:49] Cathy McPhillips: Okay. out a bit, in two years, how many gen AI platforms do you think will dominate the enterprise landscape? Are we looking at many or just a few? I mean, just looking at even ai, this isn't all ai, [00:26:00] but MarTech, Scott, Brinker's landscape's at what? 15,000 tools right now? No.
[00:26:03] Paul Roetzer: Yeah, I think there's only a few, honestly, like there, there's gonna be a hundred thousand, like his landscape could literally be a hundred thousand because everyone's a developer now. Anyone can go in and build apps and that's only gonna get easier in the, in the months and in year or two ahead. So you, you could build apps for anything in the future.
[00:26:24] So I do think that it's largely gonna come down to Microsoft, OpenAI and Google for like what Sam Altman would call the operating system of the future. I. I, you know, just we've already seen Anthropic is shifting away from the chatbot competition like they've accepted, they just cannot win in that game.
[00:26:44] And they're pushing now very hard on coding and the safety and alignment side because they can't compete. And I don't see like X AI might try and compete on, you know, with Grok on the consumer chatbot, but [00:27:00] they have no chance in the enterprise. So I do think that, that's where a lot of it's gonna live.
[00:27:05] And then I think a lot of people's experiences will actually be through like a Salesforce or a HubSpot where you're already doing everything within it and your chatbot experience just lives on top of, you know, that data layer and applications. So, yeah, I think that's, you know, it's gonna play out where you're gonna pretty much choose between Microsoft, OpenAI and Google.
[00:27:25] And then you're gonna have your core chat experiences, within your other tech, you know, essential pieces of your tech stack.
[00:27:32] Cathy McPhillips: Yeah, and I do like the smaller tech because most of the smaller tech is built by people that had a problem that no one could fix, so they built it themselves.
[00:27:40] Paul Roetzer: Yeah,
[00:27:40] Question #9
[00:27:40] Cathy McPhillips: they're just making other tools better, you know, the big tools better. Okay. And what about open source tools and models, do you think, do you have any thoughts or concerns around using open source LLMs in the enterprise AI stack?
[00:27:54] Paul Roetzer: the, you know, there, there's a couple of ways to look at this. I, [00:28:00] I am not like the most well-read person on the benefits of open source, despite all of my efforts to understand it. I have probably more concerns about open source than I do, like my ability to advocate for it. But I wouldn't say that that is a highly educated opinion that you should share.
[00:28:23] I listen to both sides all the time and I'm constantly just trying to, I think my. Predisposition is that I'm not the biggest open source fan, but I try to always have a very open mind in that that might be the incorrect way to look at this. So the way I think about this is, my concern with open source is I think that the understanding of the security risks
[00:28:56] is so low that there's going to be ways [00:29:00] that these open source models are used in very nefarious, nefarious applications that the business world is not anticipating right now.
[00:29:09] now that being said, the opposite opinion would be you're trusting OpenAI, Google, or Microsoft to control your intelligence stack and. They each have their own challenges. So if it's proprietary and closed weights, you can't do anything with it. And so I think that more developers prefer the open source side because then they have more control over that model.
[00:29:38] So, I don't know. I think that it's gonna be a mix. I think a lot of enterprises are going to build on open source because it gives more flexibility and they might actually be able to argue more security. And I think a lot of 'em are gonna just accelerate faster by building on top of the APIs of, you know, Google and,
[00:29:57] And OpenAI and Microsoft and [00:30:00] others. So I don't know. This is one I would, I would lean into, like I am, I would not consider myself the foremost expert on open source versus closed. And where I would lean into your it, you know, people and your security people, and, you know, listen to their opinions as well and make sure you get a variety of perspectives on this because there are people on both sides.
[00:30:22] It's kinda like politics. There are people on both sides who are 100% convinced they're right. and I'm sitting in the middle, like I can kind of see arguments for a lot of this on both sides. Like, and so I'm not like, a purist in either sense on, on either side of it.
[00:30:39] Question #10
[00:30:39] Cathy McPhillips: Okay. Let's dig into learning, literacy and leadership. So let's talk about AI councils. So if someone has an AI council or is looking to build an AI council, should the CEO be on the council? What role should be involved in a council?
[00:30:56] Paul Roetzer: In an ideal world, the CEO is at minimum [00:31:00] very supportive of the building of the council. If it's a small to mid-size business, I could imagine there's a potential role for the CCEO, to be a part of it if it's a bigger enterprise. I mean, I'm the CEO of a small business and I don't have three minutes free every day, so I can't imagine like also sitting on a council that's meeting, you know, a couple times a month and stuff, but I would a hundred percent want to be involved in, in that.
[00:31:25] So, I don't know, I think it just depends on your organization. But the CEO.
[00:31:29] you
[00:31:30] AI forward company CEO has to be a hundred percent involved in the overall AI vision and strategy, and needs to be supportive of whatever the council or councils are doing. And then ideally there needs to be a csuite executive sponsor or co-chair or whatever, like you, you have to have the C-suite involved and the bigger the company gets, the more important that becomes.
[00:31:57] Cathy McPhillips: So would you say someone from every [00:32:00] department, people that are very AI forward, some skeptics? Like should there be a mix of all those folks?
[00:32:05] Paul Roetzer: I mean, in my mind, the ideal is you have a leader from each of the core departments of the company represented because too ear, you know, early on in like 2024 is like AI councils becoming more normal to see an organizations. There was just too many times I would meet with companies and they, they wouldn't have invited the CMO or like anybody representing the marketing organization to the council.
[00:32:28] And I'm like, how is that possible? And I think what was happening in most instances is it was being treated as an it, you know, technical solution. And so it was living under like the CIO or that department, and then legal was involved and they were, they were basically just treating it as like risk and security, and technology.
[00:32:48] And they weren't thinking about business, challenges and business outcomes and products and services. And that to me is like a, just a. A very narrow-minded view of what the [00:33:00] role of the council should be. Now, if there was AI councils for the marketing team, they had their own and the like, fine. But you know, I think that that they, you have to lead, you have department leaders involved because this isn't just a technical thing.
[00:33:13] Cathy McPhillips: And especially when marketing has so many use cases,
[00:33:16] Paul Roetzer: Yeah. It's usually the tip of the spear when it comes to like value creation, marketing, sales, and customer success are like the first three things that you should be looking at.
[00:33:25] Question 11
[00:33:25] Cathy McPhillips: Yep. Okay. Let's say your company has only has an a basic AI policy. Where do you start to begin to use it to educate your team?
[00:33:34] Like what, how do you get that off the ground and enforced?
[00:33:37] Paul Roetzer: Yeah. So for an AI policy. Yeah. I mean, I, so we're in the process of updating ours. Like Cathy had put together, you know, initial version of an AI policy for us. back last summer, you know, I think in, in 2024 and just in the last couple days, we were taking a fresh look at it and starting to realize all the [00:34:00] complexities that have been introduced that we weren't accounting for in our own generative AI policy.
[00:34:05] such as like computer use agents, you know, there's, those didn't exist until in, you know, in the last five months. Like now you have agents that can take over your screen and see and understand everything you're doing. It's like, we didn't anticipate that, or like, we didn't think it was gonna happen enough that we'd need to put in the policy.
[00:34:22] So you're starting to look at, you know, usages like that, what. You're allowed to connect it to, if you go into ChatGPT right now and, you know, and you have a, a team account that asks you, do you want to connect to Google or Google Drive, it's like, well, what are we allowed to connect these things to? So I think, you know, it probably helps to take a fresh look at these policies every three to six months based on where the models have advanced to what the new capabilities are, and make sure that the policy is still, up to date and that the team is still properly trained on that policy.
[00:34:53] You know, probably needs to be part of all your onboarding processes for new, new hires. You teach the interns, like [00:35:00] the policy is becoming more and more important because the risks associated with this stuff are becoming much greater and so we really need to make sure people are doing this in a responsible way.
[00:35:11] Cathy McPhillips: Yeah, and there's those slides you put, I think it's in scaling, definitely one of your presentations where it's like 70% are using AI and 64% or for companies don't have a policy. So people are using the tools whether you tell them how to or not.
[00:35:26] Paul Roetzer: Definitely.
[00:35:28] Question #12
[00:35:28] Cathy McPhillips: Okay. Once a company starts that journey, with their, AI policy and usage, what's a solid KPI to track AI literacy or adoption during those early pilots?
[00:35:39] Paul Roetzer: Ooh, this is a good one. I actually have been thinking about this. So again, if you're kind of new to what we do, we have an AI academy we have for five years, but it's primarily existed of two core series and certification programs, piloting AI for marketers. Originally it was targeted at marketers and then scaling ai, which is more [00:36:00] for business leaders.
[00:36:01] We are completely re-imagining all of that. So AI Academy is becoming far more robust. It's meant to be like a turnkey academy for any size organization that wants like a plugin AI Academy with personalized learning journeys. And so I've been thinking a lot about, as like, let's say a company comes in and they buy 50 licenses for their marketing team, 20 for their sales team, 10 for their customer success team, and maybe their, their, their lawyers and their accountants get some licenses.
[00:36:29] How do you measure the impact that that's having? So there are simple ways like. Completion, you know, certi certificates earned that someone has actually gone through and done the thing. but I think when you start to actually look at applying that, that learning, you gotta then start to look at like, okay, what is the utilization rate of.
[00:36:52] The co-pilot license they have. So like if you start with someone who, and you put them into AI Academy and they go through AI fundamentals and [00:37:00] AI piloting, and then they take AI for marketing and they kind of go through this journey and they had co-pilot before and they used it like 10 times in the month of May.
[00:37:09] And now in the month of July they've used it a hundred times and they've built five, you know, co-pilots. And so that's one way like where you can start to kind of, then you can te level it down to like, well, what is their efficiency? What is their productivity? Which may require establishing some benchmarks.
[00:37:25] Like, here's the thing they do in their job all the time. And they now do this three times faster than they did before. So it's something we're definitely gonna be working on. We're, we're actually gonna hopefully be building into our academy where you'll be able to do this, like do an assessment upfront and then take an assessment as you go that can kind of help you measure progress.
[00:37:43] But yeah, I'm thinking about that a lot as I'm building the courses that will make up sort of Academy 3.0 as it when it comes out.
[00:37:52] Cathy McPhillips: Yeah, because comprehension's one thing, but putting it into action is
[00:37:56] Paul Roetzer: Correct. Yeah. We, so our, our whole thing is we want [00:38:00] to drive transformation, whether that's individual transformation or business transformation. And so you need to be able to measure how you are transforming as a, as a professional or as an organization. And so a lot of what we're doing is really starting to think about that now, like separately for yourself.
[00:38:18] I mean, honestly, you could literally just go into ChatGPT and say, Hey, I'm, I'm gonna start investing a lot of my time in AI literacy. I'm gonna be taking courses, going to events, or I'm doing this for my team. We're gonna be investing X dollars per team member this year in literacy. How could we measure their improvement?
[00:38:35] And you'll probably get some like really cool ideas straight out ChatGPT
[00:38:38] Mm-hmm.
[00:38:38] Cathy McPhillips: Yeah. Okay. Here's a little time for you to reflect.
[00:38:42] Paul Roetzer: Okay.
[00:38:42] Question #13
[00:38:42] Cathy McPhillips: If you were building marketing AI Institute all over again from scratch today, with what you know now, what would you do differently?
[00:38:49] Paul Roetzer: Oh boy.
[00:38:51] Cathy McPhillips: We're
[00:38:51] Paul Roetzer: I, yeah, so I don't know for people who don't know the context. So I started Marketing Institute in 2016 as a spinoff business unit of [00:39:00] my marketing agency at the time. So I sold the agency in 2021. We launched our Mayon event, the, our flagship in person conference in 2019. But the institute lost money for six straight years.
[00:39:16] Like we never turned a profit from 2016 until March of 2023. so that's kinda like the quick background on the institute. I don't know, like I will say like overall in life, I am not one to look back and think I would, how I would do stuff differently because I'm generally, I. I try and focus on the present and like, this is where I am, and I'm, I'm usually generally very content with where I am.
[00:39:45] And so I would say as a business leader, I kind of feel that way. Like it was a hard journey. it was a long time talking into the wind where like people weren't paying attention or like, you know, understanding the significance that was gonna happen. [00:40:00] And I don't, we tried a lot of stuff, like we experimented, we were willing to fail a whole bunch.
[00:40:04] and I would do that all again because like, we're here and I'm content with where we are and we have this opportunity to hopefully make a big impact on accelerating AI literacy. And we've developed amazing partnerships and we have a great team. And so I don't, I don't know, like, I don't, I don't, I think I'm, I'm happy with what we did and the decisions we made, and I'm so focused on the future that I just don't, honestly, I.
[00:40:33] Ever go back and think I would do something different or, you know, try something else. yeah, I don't yeah, know. It's a really good question, but I think just for me personally, like I very rarely do that. I don't look back. I basically just try and live every day personally and professionally to not have regrets.
[00:40:51] And so if I have an idea, I'll, I'll pursue it. If I have a vision for something, like I'll go after it. Like, I don't fear failure at all in business. I just [00:41:00] try and be strategic enough financially to give us the freedom to fail. and I've all, I mean, like literally my, my marketing agency blueprint book, from that I wrote in 2011, the final chapter was embrace failure.
[00:41:16] So I think I've just always kind of approached business that way.
[00:41:19] Question 14
[00:41:19] Cathy McPhillips: I like it. Okay. This is a little inside baseball, so you might need to explain a little bit of this. in the context of jobs, GPT and AI exposure levels, how do you actually bridge the gap between current capabilities and future roles and what is the smart move for career future proofing? I.
[00:41:36] Paul Roetzer: We will put the link to JobsGPT in the show notes. So if you're not familiar with it, if you go to smarter rx.ai and just click on tools, you can click on JobsGPT I first built that in summer of 2024, and the idea was to help people that you put in your job title or your job description, and it would help you find use cases for AI specifically.
[00:41:58] You know, language [00:42:00] models, chatbots, but it was also meant to accelerate business leaders. being proactive about the impact AI would have on jobs and the displacement of their teams. And what I found and continue to find is there's a lack of comprehension about what these models are capable of and what they're going to be capable of in the very near future.
[00:42:23] And so I devised this impact, like this AI exposure levels like a key that basically today has 11 levels and it just kind of walks through like, okay, today they can do text in and text out and they can connect to other systems. and now they're getting video capability and image capability and advanced reasoning capability and persuasion capabilities and, you know, acting in the digital world like these computer usage, like all of these things we knew were, were coming, like the labs told us this is what we're working on.
[00:42:52] and so. My goal was to try and actually train AGI PT that could project out the impact this would [00:43:00] have. So if you put chief marketing officer, a customer service representative or sales director into jobs GPT, it would actually say, well, here's the ways you could be doing it. Here's how AI's gonna impact you.
[00:43:10] Here's time you could save. And then you could say, okay, but I'm a little worried that as it starts developing the ability to produce videos that I as a videographer may not be as needed. And you can talk to it about that. And so, I think that, and then, and then this year in April, I introduced a new, a new conversation starter and it where you can just talk to it about future jobs.
[00:43:35] Like, here's my major in college. I'm not sure what I'm gonna be doing. I'm a writer. I don't know what like writers are gonna be doing in 12 months. And it'll try and work with you to devise like what the future of that career path could look like. So. Yeah, I don't know. That's what jobs GPT was meant to do was like be a conversation starter and get more people thinking about the near term impact that these things were gonna have on jobs.
[00:43:58] So we could be [00:44:00] proactive as business leaders to reskill and upskill our people to be better positioned for the future of work.
[00:44:05] Cathy McPhillips: Which you went through when we were looking at hiring for all these folks
[00:44:08] Paul Roetzer: Yeah.
[00:44:08] Cathy McPhillips: know, what are the jobs, what could AI potentially be doing? How could we make sure we're not gonna hire people and not need them in 12 months?
[00:44:14] Paul Roetzer: Yeah. And that's like at a very tangible level, like an example of how you would use it. Like we were hiring, a collection of people and so each job description you look at it and say, okay, like this is a full-time employee today, but knowing what these models are gonna be capable of and when GPT five shows up, we can reasonably.
[00:44:32] Assume what its capabilities will be. Is this still a full-time job? Like do we, do we still need this? Or if we're building out a customer success team, what may have taken, you know, five or 10 customer success people a year ago, maybe we can do it two or three. Now we have the luxury of building from the ground up so we don't have to worry about, you know, legacy teams that may not need as many humans in the future.
[00:44:56] So I get that we're sort of in the, [00:45:00] the easier approach here. But my goal was for companies that aren't, that have legacy teams, the more proactive you are, the the better chance you have of being prepared as this impact truly starts coming. 'cause it's coming like we, one way or the other, like it's, it's gonna happen.
[00:45:15] We can't just ignore it and then deal with it when it's here. I would far prefer people were realistic about the impact it's gonna have and then started preparing people. And even if it's not, Hey, we're gonna re-skill you in, in ai, we're gonna, you know, provide this AI education and training and. We hope it's gonna be here.
[00:45:31] You're gonna do this and bring this to life, but if nothing else, we're gonna make you more employable, you know, in the future. Because I don't think any company right now can promise that two, two years from now, they're gonna need the same level of staffing. I honestly think if you, the only way you can do that is if you are extremely confident in your growth trajectory.
[00:45:51] If you're at a company that's growing 10% or less per year, there is no way you need as many people in that business in [00:46:00] 18 months that you do today. If you, if you do, you're not running the company very well. and that's just the reality.
[00:46:07] Cathy McPhillips: so you talked about reasonably assuming and what might change with roles. Not everyone reads as much as you do. Like
[00:46:13] Paul Roetzer: Yeah.
[00:46:14] Cathy McPhillips: Doing or paying attention to, to be able to keep up with that and be able to frame this in a way that actually makes sense for their business.
[00:46:20] Paul Roetzer: Yeah, I don't know. I think the, that's a big part of why we introduced the AI literacy project, and I now mentioned that a couple times, if people aren't familiar with it, we'll put the link to the literacy project in. But the whole premise was we are gonna try and do our part to provide as much free education and inspiration as we possibly can through our podcast, through the newsletters, through blueprints, we publish, through free classes, like all these things, this AI Answer series, so that you can connect the dots in your company, your industry, your career path.
[00:46:55] I, what I encourage people to do is experiment with the [00:47:00] tools every day and. Figure out how you best learn and just go do it every day. Like don't get overwhelmed by the scope of all of this. Just commit like, okay, I'm gonna listen to like three podcasts a week. If you like podcasts, I'm gonna go get like a couple books.
[00:47:17] 'cause I love reading. I'm gonna take courses 'cause I learn really what they're, I need to be inspired by other people. I wanna go to some events like just do it, like find the thing that motivates you and go do it. I totally understand very few people are going to want to consume as much information about a AI as I do.
[00:47:35] Like I some days don't want to do it. Like literally yesterday I was power washing my basketball court in the backyard, spent basketball court for three hours and I just listened to podcasts for three hours. And I wanted with all of my, be my, my being to just listen to music for three hours yesterday.
[00:47:52] 'cause my brain was fried, but I was like, I have, I have to listen to podcast. Like, that's me though. Like that's what I have committed to do in my life. [00:48:00] You just need to find out what your interests are and which thread of all this you wanna, you know, really push on.
[00:48:07] Cathy McPhillips: Yeah.
[00:48:08] Just think you're doing it for all of us, so we don't need to do that.
[00:48:10] Paul Roetzer: Yeah. Total side note, I texted my family, the, my one regret, I said I don't have regrets. My one regret in life is that I didn't discover power washing until my mid forties. That is like the most satisfying thing you can do outside of like, cutting your grass. because it's like this immediate gratification.
[00:48:30] I know this has nothing to do with ai, but like sometimes we need the more human side and like, if you just need to like decompress, buy yourself a power washer and pick like some cement in your driveway or your back patio and just go clean it.
[00:48:44] Cathy McPhillips: I just got a new edger on Saturday.
[00:48:46] Paul Roetzer: Oh, edges are great too.
[00:48:47] Cathy McPhillips: I've like walked out to my tree lawn like four times and like just stared at like that beautifully
[00:48:51] Paul Roetzer: Oh my God, I, every day I cut my grass. Like every time I do it, I'll stand there to my kids. Like, doesn't look amazing. Yeah. Dad looks great. Claire, by the way, you [00:49:00] could cut this part of me just,
[00:49:01] Cathy McPhillips: keep it in.
[00:49:03] Paul Roetzer: talking my sanity.
[00:49:07] Cathy McPhillips: for sure.
[00:49:10] And
[00:49:10] thinking about this next generation you have, you have young kids, I have moderately young kids.
[00:49:15] Question 15
[00:49:15] Cathy McPhillips: What courses should kids in school be thinking about as they wanna prepare for an AI infused world, but really what should colleges and schools
[00:49:23] Paul Roetzer: Yeah.
[00:49:24] Cathy McPhillips: think about to help these kids?
[00:49:26] Paul Roetzer: step one is teach the teachers you have to drive AI literacy among the gatekeepers of knowledge and experiences in the classroom. So if we're not empowering the teachers and professors to responsibly integrate AI into the classroom, this will fail in your school. And there's, there's very little debating that like, we have to do that.
[00:49:47] the second step is you have to think about AI as a layer over everything else. Like I'm sure that there's an argument to like, have maybe an AI major down the road that's like a dedicated thing, but I don't [00:50:00] know,
[00:50:00] I.
[00:50:01] I would've to think about that, like what the justification would be for that. I think, like we talked on a episode 1 48 or 1 49 about like, bowling Green State University now has like an AI plus program where you can get an AI on top of whatever your major is.
[00:50:15] I love that concept where we're just teaching the AI layer over whatever. So if you wanna go into philosophy or journalism or business or sales or psychology or sociology, like. Understand how AI is affecting that discipline, that career path. And so I just, I feel like there needs to be an AI 1 0 1 everywhere, like every freshman in college.
[00:50:39] I said this back in 2018, every freshman in college to take AI 1 0 1. And then I think there just needs to be elements of AI layered over everything else. And I think it needs to start very early. Like I'm a huge advocate for teaching AI responsibly at the middle school level. Like my kids are going into seventh and eighth grade.
[00:50:58] God, I [00:51:00] can't believe their school's done this week. So they're, they're going to seventh and eighth. And I am very, strongly evaluating how to more aggressively prepare them for the real world because I don't, I don't think it's fair to, to ask their schools to do it fully right now. I understand there has to be this transitional period.
[00:51:22] So as a parent, I'm going to take it upon myself to do everything I can to prepare them. For the future that I am fairly confident is coming.
[00:51:32] Cathy McPhillips: For sure.
[00:51:33] So,back to the AI Plus at Bowling
[00:51:35] Paul Roetzer: Mm-hmm.
[00:51:36] Cathy McPhillips: know, in what, just January of 2023, people were like, I'm gonna be a prompt engineer,
[00:51:41] Paul Roetzer: Yep,
[00:51:41] Cathy McPhillips: we need prompt engineers. And it's like we're all prompt engineers at this point. So
[00:51:45] Paul Roetzer: yep,
[00:51:45] Cathy McPhillips: Plus seems like it makes the most sense versus like just AI as a, as a whole.
[00:51:49] What does that even look like?
[00:51:50] Paul Roetzer: yep. And I do, I get asked all the time about like computer science degrees. I'm not in the camp of coding, doesn't matter. Programming doesn't matter in the [00:52:00] future. I am my son at, you know, 12 is loving learning to code and he is doing it, you know, self-motivated to go learn coding. And I look at it and think, man, you like learn to do repetitive things like, you know, you just grind, you work hard through something, you solve problems.
[00:52:18] You learn the rewards and the benefits of. Working hard on something, coming up with solutions and making decisions. So to me, like coding is all of those things. And if he told me I want to go into computer science, like, I'd be like, great, do it. But I think you need to like get like a liberal arts.
[00:52:38] Like I I would really like it if you also diversified your background into some of these other areas. At least take classes in all these other areas. And so If I had my choice, like right now, I think liberal arts degrees in, in theory will take on far greater value. I think, so I'm, I would actually bullish on [00:53:00] liberal arts, colleges that teach like a diversity of knowledge.
[00:53:04] Cathy McPhillips: Yeah, so my son's in it and he's been sitting through all these meetings as they're doing all this stuff and he's like, why do they keep asking me? And I'm like, because you know how to talk, can communicate. And that's a rare breed right now. So good luck.
[00:53:17] Paul Roetzer: Yeah. Yeah. That's good.
[00:53:20] Question #16
[00:53:20] Cathy McPhillips: okay. Kind of going off of the parenting and kid thing, what are a few things you would suggest to help teenagers use AI to accelerate learning without relying on it to do the work for them?
[00:53:31] Paul Roetzer: You have to teach them to talk to it as an advisor, tutor, mentor, and not as, a cheating aid. So this is, I mean, my daughter at 13 spends way more time, you know, using AI for, for a variety of purposes. and that's basically what I always teach her is like, I don't mind if you use it to help with that, but you, you, you have to say, I'm trying to understand this topic.
[00:53:57] Like, I want to learn it [00:54:00] and deeply understand it. I'd like you to help me learn it. Like, don't go in and say, write this for me. And so that's, you know, I don't know how you do that without being super intentional though, because we were all kids. You all had other things you wanted to do than do the homework.
[00:54:17] And if this is the greatest, like shortcut in history to actually doing the work. And so we still have to develop the next generation who wants to learn, wants to better themselves, you know, wants to develop. Critical thinking becomes strategic. They don't wanna just ask for answers. But that's why I think teaching the teachers is so critical.
[00:54:41] 'cause if these kids are raised to believe they're cheating when they're using ai, even if it's being used in a responsible way, that's a disservice to them because that is not how it's gonna be viewed when they're out in the real world. And so I think we just, we, we have to teach responsible use. And that would mean as a, a learning aid.
[00:54:58] Cathy McPhillips: And you did do [00:55:00] the KidSafe GPT,
[00:55:01] Paul Roetzer: Yeah.
[00:55:01] Cathy McPhillips: might help answer some of those
[00:55:03] Paul Roetzer: Man, I used that this weekend. I should even tell Bailey was gonna be so pissed at me. We were in the car, we were going, sorry, Kid Safe GPT is a thing I use to help parents understand the risks of the different platforms. It's free. You can go to smart rx.ai, click on tools. It's there, helps you like understand platforms.
[00:55:22] talk to your kids and write guidelines. So we're in the car and I'm like, so summer, starting next week we should probably talk about screen time and like digital detoxes and set some guidelines. And he's like, ah, yeah, no. And I was like, let's have a talk with KidSafe GPT. So as we're in the car driving, my wife's driving and I'm like, I'm in the car with my son and I'd like to develop some guidelines for this summer.
[00:55:45] And if my kids could have got out of the moving car, they would have like, it was hilarious 'cause it was giving me talking points and I was just reading them. It was like, this is great.
[00:55:56] Cathy McPhillips: So no
[00:55:57] Paul Roetzer: He,
[00:55:58] Cathy McPhillips: from that yet.
[00:55:59] Paul Roetzer: [00:56:00] I got great guidelines. He's not bought in yet.
[00:56:03] Cathy McPhillips: All right. We're down to our last two questions.
[00:56:05] Paul Roetzer: Okay. Nice.
[00:56:07] Question #17
[00:56:07] Cathy McPhillips: Okay. When it comes to building GPTs, is it better to create a specific GPT for each job task or one mega GPT that does content strategy, internal reports, sales, writing, and all of it.
[00:56:20] Paul Roetzer: Huh. so I built Co CEO GPT. We can drop the link for that. There's a webinar you can watch from December of 24 that teaches you how to do it and has the template to the instructions to build your own. You can build it for whatever your job is. Co CMO, co-writer Co SEO professional, that works really well.
[00:56:44] It's designed to do a general set of things like build plans and talk strategy and complete tasks. I use that all the time.
[00:56:54] But
[00:56:54] Then I also do build a lot of just like custom gpt for very [00:57:00] specific things. And I would say my inclination here, and you can add your 2 cents, Cathy 'cause you also, you know, do with this all the time.
[00:57:08] I think the very distinct use cases can be very helpful to people to like narrow in on what is the value I'm going to get from using this? What exactly does it do? So like jobs, GPT, if I just called it future of work, GPT, it'd be like, what do I do with this thing? Like it's super general and like, oh okay, it does.
[00:57:26] And it might be able to do all those things regardless, but calling it jobs, GPT is like, okay, this is specific to the impact on jobs or like, what do I do with my job? So I don't know. I think that some of it may be perception based, but like all of it is about adoption. And so I would say whatever the path is that is clearest to help people understand the value it's gonna create and drive adoption and utilization.
[00:57:48] That's what I would go with. But I think like specific naming of GPTs is probably the way to go.
[00:57:53] Cathy McPhillips: I agree. Yeah. The ones that I've built have been for very well. a couple that are very specific for particular use [00:58:00] cases, and I have one that's a broader strategic one, my Paul GPT
[00:58:03] Paul Roetzer: Yeah.
[00:58:03] Cathy McPhillips: that I built, that it's like helps me just think bigger and think about things different in a different way. like you, thinking like your CEO. So that could be across anything,
[00:58:14] Paul Roetzer: Yep.
[00:58:14] Cathy McPhillips: but I guess it is specific, specific in what I want it to tell me. I.
[00:58:17] Paul Roetzer: Yeah, yeah, yeah. I think I would always start with very distinct things, like if you're trying to understand the value yourself, I. I would pick something you do all the time and just build it. And you know, because it's basically like if you're gonna be repeating the same prompt every time, like I'm building one right now, or I built one for Building Academy 3.0.
[00:58:37] So like I already have my co CEO it knows our business model knows all the things we do, but like, I want one that I specifically trained on functioning as like an editor and thought partner for the build out of these courses. And so, you know, I'll go in, I develop my outlines, I'll give it to Academy 3G PT and say, assess this outline, you know, what I'm trying to achieve.
[00:58:57] Like, you get who the audience is, like what do you think about the outline? [00:59:00] What am I missing? So in that instance, like having a very specific GPT or GEM in this case, is really helpful for me.
[00:59:09] Question 18
[00:59:09] Cathy McPhillips: Yeah, agree. Okay. You talked about this, I think on episode 1 47 or 1 48, but, search, what do you think AI will do to the search marketing industry, especially paid search? seen some shifts already. What's happening now and what's on the horizon? I.
[00:59:25] Paul Roetzer: I
[00:59:26] think this space is gonna be pretty dramatically, it's gonna look dramatically different in like 18 to 24 months. I've talked to and listened to some of the leading minds in search and I've yet to find anybody who speaks with very much confidence about what's gonna happen. I mean, just like last week we got AI mode from Google and then they had some research that showed that apparently people are still clicking on a lot of links in AI mode and it's not impacting their paid side.
[00:59:56] I think they even said like they were getting higher [01:00:00] click rates in some cases. So I don't know. I think until this all shakes out, until we see, like if ChatGPT is gonna have a paid function, they're obviously coming after Google on all fronts. I. Until we know what their paid function looks like.
[01:00:16] And until we see how consumer behavior changes around search and whether the next generation ever even goes into a search engine, or if they literally just talk to ChatGPT or Gemini all the time, and that's it. And then like, how do you serve up ads if voice becomes the dominant interface? I have no idea.
[01:00:33] I think this is, you know, of all the conversations I have, this feels like one of the largest unknowns about what happens in search and the impact it then has on paid search and our strategies, you know, as builders of businesses and marketers, like how does that channel change for us? I don't know.
[01:00:52] And I'm pretty convinced the people at these tech companies don't know. So I would say it's a, it's a wide open space. I would be paying very [01:01:00] close attention to. If you're impacted by it,
[01:01:02] Cathy McPhillips: Definitely. All right. I have one more. I love asking this question at the
[01:01:06] Paul Roetzer: you just wanna get to 20, you get 19. You just wanna be able to say you did 20.
[01:01:10] Cathy McPhillips: maybe, but also like I like ending on a super high note
[01:01:12] Paul Roetzer: All right.
[01:01:13] Cathy McPhillips: one didn't seem like the highest note to end on
[01:01:14] Paul Roetzer: All right. Give it to me.
[01:01:15] Question 19:
[01:01:15] Cathy McPhillips: What, what excites you? Like what excites you this week about AI or something you're working on or something you're
[01:01:21] hearing about?
[01:01:23] Paul Roetzer: I would, like at the moment, it's more macro, but I feel like we finally have arrived at a point where companies, and more largely, you know, society is, is just understanding the moment we find ourselves in and the significance of what's happening. And so there's far more, urgency from people to play a role in this, to have some agency in what happens next.
[01:01:54] And I think just seeing people being more proactive about transforming their [01:02:00] own careers and then, you know, the conversations we have every day with company leaders who now understand the importance and are wanting to, Make an impact. And I think that there's still a lot of optimism. Like we saw that in our state of marketing.
[01:02:17] I report where they said like, people overall, the sentiment is still optimism. That they think the future can be bright while they think it's gonna impact jobs. they overall think about the positives. And that's kind of how I choose to think about it. Like there's a lot of ways this goes wrong and if I dwell on those things, then I don't sleep so well at night.
[01:02:37] But I think every day I try and just focus on the fact that, I think for the most part, we are going to be able to guide the impact this has on business and society. And I'm generally optimistic it's going to go really well if we are intentional about it. And I feel like right now we're doing everything we can to try [01:03:00] and drive that positive outcome.
[01:03:02] but I always feel like we could do more. So I don't know, I guess I'm just excited about the unknown ahead. In a good way. And like that means we can kind of reimagine everything. And that's really exciting to me.
[01:03:16] Cathy McPhillips: Yeah, that's what
[01:03:16] Paul Roetzer: yeah.
[01:03:17] Cathy McPhillips: I'm working on a presentation for next week
[01:03:20] and I to say that, you know, we have a responsibility and opportunity to be on the right side of all of
[01:03:25] Paul Roetzer: Yeah,
[01:03:26] Cathy McPhillips: and there's a lot of really great things that are happening and we wanna make, we need to make sure that we are stewards of all of that.
[01:03:31] Paul Roetzer: yeah. Yeah. And I think there's a lot of opportunities for everyone listening to do the same. Like so many organizations are struggling to still even understand this stuff and develop plans, build councils. So I think just the chance to see people at all phases of their career emerge as leaders to drive a human-centered approach to this.
[01:03:54] and then just like hear the stories every day, like the messages I get [01:04:00] on LinkedIn, the stuff you hear firsthand from people, the text messages I get from people who've like. Altering their careers or doing these incredible things in their organizations to try and drive change. That's inspiring to me.
[01:04:13] It's like hear people kind of picking up this message and like going and doing really incredible things. So, you know, I think that gives us, motivation to, you know, really keep pushing with what we're trying to do. because we can't do nothing. Like the alternative is we don't try and make this impact and we don't try and push for the positive outcome.
[01:04:35] And then it's our fault if, if we get there and it didn't happen. And I say that collectively like our fault overall. Like we just, we can't sit around. It won't go well if we don't all collectively do more to make sure it goes, that we drive AI for good.
[01:04:51] Cathy McPhillips: Excellent. Well, if you answered one of these, or if you ask one of these questions on our scaling ai, thank you for your great question. And if you are just listening to [01:05:00] all of these, as you can see, you are not alone. So, a lot of opportunities to join our community and stick with this group and stick together so we can answer them together and learn together. and Paul, thank you as always for, for this. Super helpful.
[01:05:13] Paul Roetzer: Thank you. And just a quick reminder, AI for B2B Marketers Summit June 5th. That is a virtual event. There's still time to join us there. intro to ai, the free class we talked about. The next one of those is June 10th, and then the next scaling AI free class is June 19th. So those are all coming up. And then we'll do one of these AI answers episodes after each of those.
[01:05:34] So we'll be answering a lot of questions in the next three weeks.
[01:05:39] Cathy McPhillips: If you really wanna do this all in person, we'll see you in October
[01:05:43] Paul Roetzer: Come to make on, yeah. It's like 19 weeks away, which is crazy. All right, Cathy thanks so much. Great job. As I think this is your first co-hosting of, oh no, we've done a couple of a, a question and answers sessions together. Alright, well then welcome back and welcome to [01:06:00] the first co-hosting of AI Answers.
[01:06:02] All right, and thanks everyone for joining us. We'll be back with our regular weekly next Tuesday.
[01:06:06] Thanks for listening to AI Answers to keep learning. Visit smarterx.ai where you'll find on-demand courses, upcoming classes, and practical resources to guide your AI journey. And if you've got a question for a future episode, we'd love to hear it. That's it for now. Continue exploring and keep asking great questions about ai.
Claire Prudhomme
Claire Prudhomme is the Marketing Manager of Media and Content at the Marketing AI Institute. With a background in content marketing, video production and a deep interest in AI public policy, Claire brings a broad skill set to her role. Claire combines her skills, passion for storytelling, and dedication to lifelong learning to drive the Marketing AI Institute's mission forward.