45 Min Read

[The Marketing AI Show Episode 52]: 15 AI Questions Everyone Is Asking

Featured Image

Wondering how to get started with AI? Take our on-demand Piloting AI for Marketers Series.

Learn More

After 26 Intro to AI classes since November 2021, we’ve seen quite the gamut of questions from our attendees. So far in 2023, we’ve hosted seven classes, with a collective 147 questions asked. We wanted to take this week’s podcast episode and run through the most commonly asked questions–and answers.

We hope you enjoy this episode, and join us for our next Intro to AI for Marketers class on June 28!

Listen or watch below—and see below for show notes and the transcript.

This episode is brought to you by BrandOps, built to optimize your marketing strategy, delivering the most complete view of marketing performance, allowing you to compare results to competitors and benchmarks. 

Listen Now

Watch the Video

Timestamps

00:07:03 — How do you define "generative" in the context of AI, and how might it be applied in various marketing disciplines, such as marketing, design, or podcasting?

00:08:07 — How can marketers balance the integration of AI in their strategies while maintaining the human touch and creativity in their campaigns? How will the role of human creativity evolve within this AI-driven landscape?

00:11:14 — As AI continues to improve, how can writers maintain their value in a world where AI is getting better at creating flowing and creative content?

00:13:13 — Given the rapid pace of change in marketing technology, what advice would you offer to businesses trying to navigate the influx of AI tools in the market, and how can they build a coherent tech stack?

00:16:01 — As more and more AI-based companies emerge, how can marketers discern which ones to invest in for their company's specific needs?

00:19:09 — How can the marketing department work collaboratively with IT/AI/technology teams to leverage AI capabilities?

00:20:48 — Do marketers need to be concerned with plagiarism when using AI writing tools? Do AI writing detection tools work?  

00:25:43 — How can someone verify the content generated by AI?

00:28:33 — Can AI replace or supplement roles like graphic designers? And how accessible are these AI tools for non-designers to create their own visuals?

00:30:08 — What are the potential impacts and opportunities for agencies as AI technology advances? How can they adapt today to ensure they aren't left behind?

00:32:56 — In terms of language capabilities, how far have AI tools come, and what can expect moving forward?

00:36:01 — Given the rise of privacy concerns with AI tools, what guidelines should companies follow while using AI models like ChatGPT?

00:38:06 — How can companies ensure they comply with copyright regulations when generating images with AI? How do you envision the evolution of copyright legislation in the context of large language model AI? How should companies handle their proprietary data?

00:44:13 —I want to get started TODAY. What are some steps I should take immediately to learn or to identify ways I should get started?

00:47:03 — As we move toward an increasingly AI-driven world, what are your thoughts on the future role of humans? What advice would you give to those who might feel threatened by the rise of AI?

Summary

We have a special show today! We’re running through some FAQ, specifically from our Intro to AI classes. And we used AI to help us consolidate and surface the most representative questions from our community.

How was AI used?

  • I ran all 147 questions through ChatGPT in batches (30-40 at a time) due to input limitations.
  • For each batch, I used this prompt: Consolidate these questions into 10 questions that cover all the main topics discussed in the following questions. I'm preparing a podcast interview for our CEO and would like to focus on specific questions from our customers.
  • I then had fifty (five outputs of 10 questions each) questions from each of the batches. I then wanted to combine those lists into another shorter list.
  • For that, my prompt changed slightly: Could you please take your generated responses from each batch to come up with 10-15 questions consolidating all of your numbered responses to provide me with a podcast brief where I can interview our CEO, Paul Roetzer, on these customer questions? Our goal is to answer these questions and provide Paul's expert point of view and guidance.
  • I then wanted to make sure the podcast had a good flow, so I asked this question last prompt: Would you change the flow of the above questions if we want the questions to seamlessly segue from one topic to the next?
  • Next, I read them all, cleaned up some, removed a bit of duplication, and added in some context.
  • We had still to reorder some of the questions (a human in the loop is critical!) and adjust the language slightly.
  • Paul and I then reviewed the final list, and it was time to record.

The questions covered are listed above in the timestamps.

Links referenced in the show

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: for me, AI gives me some creative ability, or at least the ability to interact with creative people in a better way because I can just, in my words, say, here's what I'm thinking, and you give me some concepts, and then I can take that to designers and say, here's roughly what I'm envisioning.

[00:00:16] Paul Roetzer: So I think that, when you look at it as truly a way to augment human, creativity, human innovation, that's where you see it really shine as a valuable tool.

[00:00:26] Paul Roetzer: Welcome to the Marketing AI Show, the podcast that helps your business grow smarter by making artificial intelligence approachable and actionable. You'll hear from top authors, entrepreneurs, researchers, and executives as they share case studies, strategies, and technologies that have the power to transform your business and your career.

[00:00:46] Paul Roetzer: My name is Paul Roetzer. I'm the founder of Marketing AI Institute, and I'm your host.

[00:00:55] Paul Roetzer: Welcome to episode 52 of the Marketing AI Show. I am your host, Paul Roetzer, here with special guest host Cathy Mc Phillips, our Chief Growth Officer. Welcome to the show, Cathy. Well, thank you so much. I think this is big shoes to Phil, big Shoes yourself.

[00:01:11] Paul Roetzer: This first time on the show, right? Like you and I do the intro AI class all the time together, but this is your first time on the podcast, right? It is. So Cathy behind the scenes brings the podcast to life every week. She is our, our, our person that is in script every week taking our transcripts, summarizing transcripts, cutting up videos for YouTube, getting the podcast posted.

[00:01:33] Paul Roetzer: So if you're a regular listener, you have Cathy to thank for your weekly episode of marketing. I show it wouldn't happen without her. But today I'll explain the format in a moment. First, I want to recognize our sponsor. So BrandOps is our sponsor of this episode. Did you know brand health, media monitoring, social listening, competitive intelligence share a voice and review tracking could all be done with the same tool?

[00:01:57] Paul Roetzer: When we sat down with the BrandOps team, it was remarkable to see what one AI-based platform could do, having a complete view of brand marketing performance and instantly knowing where to focus to increase impact helps businesses unlock faster growth. Visit BrandOps.io/marketing ai show to learn more and see BrandOps in action.

[00:02:21] Paul Roetzer: Okay. Thank you. BrandOps. And when

[00:02:22] Cathy McPhillips: I saw that demo, I was like, we need all these things. It was so

[00:02:27] Paul Roetzer: good. So good. So yeah, definitely check out BrandOps and Sonata kind of this special edition. So if you, if you're a regular listener, you know, Mike and I are traveling in, I'm in, I'm doing some European travel.

[00:02:40] Paul Roetzer: Mike is in South America, I believe. So we have, we have a number of talks on this international speaking tour going on. And rather than just not having a show for a couple weeks, we thought, all right, what are some really valuable, practical things that we could do that would be really interesting to our audience?

[00:02:56] Paul Roetzer: And so the one that immediately jumped to mind for me is, Every few weeks since November of 2021, I've been teaching an intro to AI for Marketers class via Zoom. It's a free class. Cathy, moderates that for us, and at the end of every class, so we do about 30 minute presentation, I then spend 30 minutes answering questions that Cathy curates.

[00:03:16] Paul Roetzer: So we thought we have hundreds of unanswered questions throughout this series. We've had over 11,000 people register for the Introed AI class. If you've never taken it, you're welcome to jump in and sign up again. It's on our site every few weeks. It's, it's ongoing. So, and it's a live thing. We adapted each, each time based on, you know, what's going on.

[00:03:34] Paul Roetzer: So, hundreds of unanswered questions and we thought, well, what better special edition episode? Like, let's answer the frequently asked questions. So I threw it to the team and I said, let's figure out, let's not curate this human. Like, let's use some ai. So, Cathy, walk us through how this all played out. Like how have we decided on the questions we're going to answer today?

[00:03:55] Paul Roetzer: Okay,

[00:03:56] Cathy McPhillips: so we started with, we've had seven intro classes since January. Cause we thought the questions from this year are going to be the most relevant. There were some last year that were outdated, so we just wanted to make sure that they were comprehensive of, you know, what people are asking today. So I took all the questions through the chat and through the q and a box and it was 147 questions, and I kept in the ones that were like, what are tools I should use?

[00:04:21] Cathy McPhillips: There were a lot of mul D duplicates. Yeah. And I thought, you know, the AI needs to see that this question was asked numerous times. So I kept 'em all in there and I knew what was going to happen. But the first thing I did was I put all 147 of them in ChatGPT, and it was like too, too much, too much information.

[00:04:36] Cathy McPhillips: So I knew I had to chunk it up. So I did. Batches of 30 to 40 questions. Cut and paste them in the ChatGPT. And I said, so my prompt was consolidate these questions into 10 questions to cover all the main topics discussed in the following questions. I'm preparing a podcast interview for our CEO and would like to focus on specific questions from our customers.

[00:04:59] Cathy McPhillips: So I did that five times, four times, whatever it was, and it gave me those batches. And then I took those 40 questions that remained and I said, could you please take your generated responses? And I always say, please,

[00:05:13] Paul Roetzer: which I've actually heard that that helps. Helps. That is nicer. Yeah.

[00:05:16] Cathy McPhillips: Yeah. So could you please take your generated responses from each batch to come up with 10 to 15 questions, consolidating all of your numbered responses to provide me with a podcast brief where I can interview our CEO Paul Roetzer on these customer questions.

[00:05:30] Cathy McPhillips: Our goal is to answer these questions and provide Paul's expert point of view and guidance. So it got it down to 15 questions. And then I was like, okay, the next thing I need to do is I need to go and order them so that we're going through this. We're not asking old questions. Having this, you know, disconnect in the, you know, so then I said, would you change the flow of the above questions if we want the questions a seamlessly segue from one topic to the next?

[00:05:54] Cathy McPhillips: So we read them all and we went through them. We just went through them before we got on today. And what we realized is, as used to repeat numerous times, the human is still very much needed. There were some questions that didn't make sense, some that people

[00:06:07] Paul Roetzer: weren't asking, seemed like had too complicated in some cases, like there's no reason to go into that kind of stuff, right?

[00:06:13] Cathy McPhillips: So we removed some duplication, added in some context, and flipped the order, readjusted some of the questions. Like I said, so I think we're in a good place now to.

[00:06:22] Paul Roetzer: To get through some of these. Yeah. And so we'll share those prompts in the show notes. So if you want to see what Cathy was just kind of walking through, we'll put those in.

[00:06:29] Paul Roetzer: So it was just a cool thought experiment. I actually did a event recently with Yext, with Christian Ward, their CMO, and he did something similar with theirs where they kind of categorized them. So again, like segmentation, categorization, summarization, these are really interesting use cases that a lot of people aren't thinking about with, you know, what can be a free version of ChatGPT.

[00:06:48] Paul Roetzer: So yeah, just kind of cool. And so with that, Cathy, let's go ahead and get into the 15 questions. We have 15 questions, again, FAQs from our intro to AI for Marketers class. And hopefully you find a lot of value in these. Let's go. Alright.

[00:07:03] Cathy McPhillips: Number one, how do you define generative AI in the context? Or how do you define generative in the context of AI and how might it be applied in various marketing disciplines such as marketing, design, or podcasting?

[00:07:15] Paul Roetzer: Yeah, so certainly ChatGPT ushered in this generative AI era. It wasn't, you know, the term wasn't invented when ChatGPT came out, but I think probably going back to Dolly two in spring of 2022 is when generative AI really started being talked about as a phrase to define this category. And so the simplest way to think about it is the generation of any form of content, like multimedia content.

[00:07:37] Paul Roetzer: So the five core uses that we talk about are language, image, video, audio, and code. So as marketers and in all cross discipline, all the different areas of marketing, communications, advertising, analytics, social media, email, wherever you are using ai, generative AI is the ability to generate content, by the machine.

[00:08:00] Paul Roetzer: Easy enough? Yeah. Yeah. We'll keep these concise. We're going to try and hit these at a high level.

[00:08:06] Cathy McPhillips: Okay. Number two, how can marketers balance the, in integration of AI in their strategies while maintaining the human touch and creativity in their campaigns?

[00:08:15] Paul Roetzer: Yeah. We really see AI as an amplification tool and augmentation tool for humans.

[00:08:21] Paul Roetzer: I have found GPT-4 in particular to be exceptional as a strategy assistant. So I don't use, AI writing tools to write, oddly enough. I like writing. I like thinking I, that's how I do my thinking is by writing, and so I don't really need AI there, but I love it as a strategy tool, you know, building out personas, thinking through business plans to do lists.

[00:08:45] Paul Roetzer: And so for me, a lot of the way we think about AI and a lot of the way we teach it is it enhances human creativity. You know, it really does take you to different realms and gives you the possibility. One of the things I love about it is I have no creative ability outside of writing. I'm not a designer.

[00:09:00] Paul Roetzer: I'm not a videographer, and I have, I have always struggled to even work with graphic designers because I can't explain what's in my head. It's like I'll know it if I see it. The anyone in the agency world knows that like client reaction of, yeah, this isn't it. Well, what do you want? I don't know. I'll know it when I see it.

[00:09:15] Paul Roetzer: That's how I work with artistic and people, including my wife who's a painting major and she has all these amazing ideas and it's like, I draw it for me, like I can't visualize it. And so for me, AI gives me some creative ability, or at least the ability to interact with creative people in a better way because I can just, in my words, say, here's what I'm thinking, and you give me some concepts, and then I can take that to designers and say, here's roughly what I'm envisioning.

[00:09:41] Paul Roetzer: So I think that, you know, when you look at it as truly a way to augment human, you know, creativity, human innovation, that that's where you see it really shine as a valuable tool.

[00:09:53] Cathy McPhillips: And that's one thing you and I have talked about when I was working with JK Kalinowski, you know, me being able to get out of my head when I'm, what I'm envisioning and giving that to him instead of me trying to explain it and going round and round.

[00:10:04] Cathy McPhillips: Like, what a great, what a great tool to be able to help us do things like that. For sure. He's even talked about using it to help have him design or not design some things, but brainstorm. Yep. And just get, if he's stuck on something, how can, you know, what can, what will that give him that he can then take and make, make his own?

[00:10:22] Paul Roetzer: Yeah. Even the experts, like the design experts, the, we all suffer from that blank page syndrome where you're just like, you're just staring at the page and just can't get the first ideas out. So using it as an assistant and ideation tool is a, a great way to think about these tools.

[00:10:37] Cathy McPhillips: Do you think human creativity will evolve

[00:10:40] Paul Roetzer: now?

[00:10:41] Paul Roetzer: Yeah, I think we're already seeing it. Certainly. I think, you know, people who are professionals, who are professional designers, professional writers, videographers are, when they embrace this technology, are realizing they can do incredible things with it. And then people like me who don't really have those abilities or realizing like, oh, I can actually create concepts or storyboard things that previously I would've had to use an outside person for.

[00:11:07] Paul Roetzer: So yeah, I think it's already evolving. Some people are more accepting of that than others, certainly in different disciplines. Okay. Number

[00:11:15] Cathy McPhillips: three. As AI continues to improve, how can writers maintain their value in a world where AI is getting better at creating content?

[00:11:22] Paul Roetzer: Yeah. So again, I think this goes to using it as an assistant and an augmentation tool, but also realizing that I think more human content wins in the end.

[00:11:34] Paul Roetzer: So we've, I've written about this, we've talked about this on a previous episode of the podcast. the authentic human content that's e that that is difficult to fake it, I think becomes more valuable. So in-person events, podcasts, Editorials with opinion pieces, interviews with people, live events, you know, in person.

[00:11:56] Paul Roetzer: Those are the kinds of things I think people will come to crave when AI generated content is a commodity, which it basically already is. We can all write the listicles, the how-tos any brand can have anybody go in and generate that stuff. I think people are going to crave actual human content, human creativity, human imagination.

[00:12:14] Paul Roetzer: Now it's going to be a blended thing. You know, AI's going to be assisting in those areas in some ways, but I've found personally, lately I think people just want to be back together. They want to see someone answer something unscripted. They just want to hear, you know, people's thoughts and opinions and points of view.

[00:12:32] Paul Roetzer: And I see that kind of content becoming more valued and that carries over into the art world. You know, I think that at some point, while AI art is interesting, people are still going to want to see humans create things and they're going to hear music that comes from people's soul and their own experiences and.

[00:12:48] Paul Roetzer: Know, I just, I feel like they're, at some point's going to become a bit of a, I don't know if our backlash to AI generated stuff, but I just don't think it'll carry the same value as human generated content. I was

[00:12:59] Cathy McPhillips: talking the other day with someone about e greeting cards. Yeah. Remember when that was a thing?

[00:13:04] Cathy McPhillips: Yep. Like everyone's like, I don't want an email

[00:13:06] Paul Roetzer: of a card. Doesn't feel natural. Doesn't, doesn't feel right. Authentic. Yeah.

[00:13:13] Cathy McPhillips: Okay. Number four, given the rapid pace of change in marketing technology, what advice would you offer to businesses trying to navigate the influx of AI tools in the market and how can they build a coherent

[00:13:24] Paul Roetzer: tech stack?

[00:13:25] Paul Roetzer: I'll keep this one really simple. I would start with your existing tech stack. So all the major technology companies. You know, if I think on design side, Adobe, I think on the CRM side, HubSpot, Salesforce, I think about, you know, the work productivity tools like Office, you know, Microsoft 365, Google Workspace, they're all infusing these Geneva capabilities right into the platforms.

[00:13:48] Paul Roetzer: So your first point, is going to be looking at your existing tech stack and saying what smarter tools exist within these platforms right now or in the very near future and that we should be infusing into our workflows. And then you can go out and start really evaluating some of these other, you know, more startup type companies that are getting some VC funding and starting to build smarter solutions.

[00:14:10] Paul Roetzer: But I would always probably start with your core tech stack and see if the answer already lives within software your team is familiar with. And is that like

[00:14:17] Cathy McPhillips: a self-serving you? Like just go and see what you're missing out on? Or is there like CS teams in place? Or does it really depend on the size of the organization?

[00:14:26] Cathy McPhillips: Probably just depends what questions could,

[00:14:28] Paul Roetzer: should, should they be asking. Yeah, I think it just depends on which companies you're working with and how far along they are in their AI product roadmap and how. Aware their customer success teams are, that the technology is there. What we've seen a lot of is these SaaS companies are trying to kind of play catch up and build gender AI capabilities into it.

[00:14:47] Paul Roetzer: And much of their marketing and sales teams don't even understand ai. So they have a hard time explaining it, messaging it. So in some cases you may have to be proactive and go to, you know, your CRM company or your email marketing company or social media, you know, management tool company and say, what are you doing with ai?

[00:15:05] Paul Roetzer: Like, if they haven't been very public about it, what are you doing? Anything? Do you have any features in beta that we could get involved with testing? Are there options, you know, even if it involves upgrading your package and ins, you know, increasing what you're spending each month to have access to these kinds of tools, I, you know, it's not going to be much longer that these companies can exist without being extremely public about their AI plans.

[00:15:28] Paul Roetzer: So I don't think you'll have to search very hard in the months ahead because these SaaS companies can be obsoleted very quickly if they don't build smarter solutions that drive efficiency and enhance creativity for their users.

[00:15:41] Cathy McPhillips: Right. Or going to them and saying, do you, you know, we're trying to accomplish this.

[00:15:44] Cathy McPhillips: Here's our goals. Have you, do you have anything? Yeah.

[00:15:48] Paul Roetzer: Help us. Yeah, for sure. Cause it's not, we have to look elsewhere. Yeah. We're, we're aware AI exists. We hear it can drive efficiencies, you know, how can you help us do that? Cool.

[00:16:01] Cathy McPhillips: Number five. As more and more AI-based companies emerge, how can marketers discern which ones to invest in for their company's specific needs?

[00:16:08] Paul Roetzer: Yeah, this can get tricky because there are literally thousands of these generative AI startups, and many of them are offering really interesting technologies and solutions. Many of them have zero customer success teams. They don't have onboarding materials, they have no partner programs with vendors that are trained to help implement these tools.

[00:16:26] Paul Roetzer: So you really, you know, the way to think about AI is it's just smarter technology. You still have to critically assess the companies, the products you're buying, the impact it's going to have on your team. Do you need to train your team to use these tools? Do they have any knowledge base or customer success teams that can help you?

[00:16:45] Paul Roetzer: You could get in trouble just going and buying a bunch of AI startup tools that. Don't really work that well. And that's again why I tend to guide. Start with the companies you trust that are building these tools, that have the infrastructure to support your team, integrating them and getting your team onboarded and trained.

[00:17:03] Paul Roetzer: Like I built my, my marketing agency that I sold in 2021 on the back of HubSpot and we did it starting in 2007 when HubSpot was like a year old. But HubSpot very early on, invested heavily in their customer success team, their partner program. And so I was willing to take a leap of faith with my agency and build around HubSpot cause I believed in the company and the people and the infrastructure they were putting in place.

[00:17:26] Paul Roetzer: Right now, a lot of these gender value tools don't have any of that. and sometimes if you drill in a little bit, they don't even have a vision for it. They're just trying to scale as quick as they can and get some VC funding and then they'll figure that stuff out later. So, if you're in a bigger enterprise, you're probably not taking many risks with companies like that.

[00:17:41] Paul Roetzer: You're, you're going to lean pretty heavily because you gotta get through procurement and they're not going to let you, like, buy a lot of these little tools. So I think approaching this the way you always have, you know, be thorough in assessing these companies. Think about the impact on your team. Think about how it's going to fit into your tech stack and integrate.

[00:17:58] Paul Roetzer: And then from there you'll make sound decisions. And

[00:18:01] Cathy McPhillips: like anything, you wouldn't just buy a tech because it's cool and you like Right. You know, let you, like with, it has to fit into a business need in a, in your

[00:18:08] Paul Roetzer: goals. Yeah. You can demo it and experiment with it. Like that's the beauty of ChatGPT and Google Bard is like, doesn't cost anything.

[00:18:14] Paul Roetzer: Just go play around with it like that. That's fine. That's great. Well, can you scale ChatGPT in a fortune thousand company? No. Like it's not, it's not set up that way yet. There's no enterprise features really. So that's what I mean by you can play around with this tech, but if you're going to truly pilot and scale it, you, you, you need to trust the vendors you're buying from.

[00:18:34] Cathy McPhillips: Right. But to that point though, if you're, if you work for an enterprise and you are just starting to get into this and tinker with some things, using ChatGPT on personal stuff and just seeing what it's capable of will really help benefit you understanding what could be bigger in the

[00:18:48] Paul Roetzer: enterprise.

[00:18:49] Paul Roetzer: Definitely help you make the business case internally. Sure, yeah. Experimentation is critical. Yeah, just don't, you don't want to bet your career on recommending a vendor that doesn't exist in six months because someone built a smarter version of their company overnight. Right.

[00:19:03] Cathy McPhillips: So, kind of segueing, you know, from what you were saying about, you know, talking to procurement and other departments, how could marketing departments work collaboratively with it, AI technology teams to leverage the AI capabilities?

[00:19:16] Paul Roetzer: Yeah. So if, if you're in a big enterprise, certainly the sooner you, you collaborate, the better. In small to mid-size businesses, it may be an outsourced function. You may not even, you know, talk to it regularly or even know who your IT person is. So it really depends on the size of the organization that will kind of dictate this.

[00:19:34] Paul Roetzer: There are many use cases where it's going to have minimal to no involvement. So, you know, I've said it before, like if you just want to jump in and start getting an AI writing tool, you may not need a IT to do that. You know, your own governance guidelines in your company. That may not be true for you. But I think that the more we work cross discipline, we're going to need each other.

[00:19:54] Paul Roetzer: And we often advise companies build an AI council. You know, a great way to think about AI moving forward and help build a roadmap and principles and policies is to bring cross-discipline people within the organization together and talk about the impact this is going to have. It is certainly going to have their concerns around privacy, security, cybersecurity risks, and those are very valid, especially as you work on larger scale AI projects that require the access and use of data.

[00:20:21] Paul Roetzer: So I think, you know, you know your organization and you need to really figure out the best way forward to involve people in and kind of get buy-in the early pilot projects. Sometimes it's just not going to be needed to be that complicated. But as you start to think about scaling this and driving digital transformation in your company, you're going to need your peers bought in and you know, the people across the diff different elements of the C-suite, right?

[00:20:48] Cathy McPhillips: So this one, we, I seem to be, it comes up more and more lately. Every time we do this do marketers need to consider. The marketers need to consider plagiarism when using the AI writing tools and the writing detection tools work.

[00:21:02] Paul Roetzer: Yep. Very similar vein. These two questions, plagiarism, not really an issue.

[00:21:07] Paul Roetzer: I mean, realistically, if you generate something using an AI writing tool, it's not plagiarizing it, it's not copying and scraping content from somewhere. It's predicting words in a sentence, a large language model. It's basically what it does, it learns from a corpus of knowledge. Like imagine it ingesting Wikipedia and Reddit and c n n and you know, your corporate website and all of these different things.

[00:21:27] Paul Roetzer: It consumes all this information and it learns how humans write and the frequency with which words appear near each other. So, and I don't know if that's overly complicated, I'm trying to simplify this, but it's basically learning and then it just predicts words when you ask it a question or give it a prompt.

[00:21:42] Paul Roetzer: So it's not plagiarizing anything. It's, it's truly original content in that sense, it's synthesizing information the same way you or I would this morning. You

[00:21:51] Cathy McPhillips: described that as you. You know? I think that was a really good under, you know, that made us understand more about what this means. Like you are, if someone asks you to talk about a topic, right?

[00:22:01] Cathy McPhillips: You read about the topic. Do you want to talk about that for a second?

[00:22:03] Paul Roetzer: Yeah. So like, the example I've often used is if, if I say to Cathy, can you go write me, an article about the 10 ways, large language models are going to affect marketing and en and enterprises moving forward. Cathy's going to go and do research the way a human learn to do things.

[00:22:19] Paul Roetzer: You go and Google the topic, you find a bunch of sources. You copy and paste stuff over. You read through it. You highlight you bold phase, like you do all this stuff to like learn. Then you synthesize that information and you write a brief that says the 10 ways large language models are going to transform marketing and enterprises.

[00:22:36] Paul Roetzer: Now, sometimes within that you may quote something, but generally speaking, you're going to take information, synthesize what you learned, and you're going to write about it. And so if Cathy walked in my office and said, okay, and I said, you got three minutes. Like, gimme this synopsis. She's just going to explain what she learned.

[00:22:51] Paul Roetzer: She's not going to cite everything she says because it's just synthesized knowledge. That's how these things work. They go out, they learn a bunch of stuff way faster, in a much larger scale than the human mind is capable of. And then you ask it something and it synthesizes what it learns. It's not citing anything.

[00:23:07] Paul Roetzer: It's not taking anything from any specific spot unless it's an actual quote or a statistic or something like that. It's synthesizing knowledge. And that's part of the thing, like people just don't really realize that that's what it's doing. And that's the argument honestly for when all of this comes to legal cases about whether or not it's infringing, copyrights and things like that.

[00:23:27] Paul Roetzer: This is the argument that attorneys will make. How is this any different from a human synthesizing information? So there will be counterpoints, but this is basically the premise of, you know, it's just learning at a different scale than we're capable of as humans, but it's doing what we do basically. And

[00:23:44] Cathy McPhillips: you know, if something is wrong or you know, it's because it's taking the, it's get not guessing, but it's predicting the right answer based on the information that it has.

[00:23:54] Paul Roetzer: Which humans, right? It, yeah. It doesn't actually, unlike us, it doesn't know people, places, things. It doesn't really, it's not very good at assessing its own output to say, is this factual? because it doesn't really even know what facts are. It doesn't, doesn't know anything. It's just making, so that's like the craziest thing is like you use G PT four and it seems so intelligent, so creative, so you know, advanced in its reasoning capability, but when you step back you realize like it actually doesn't know anything.

[00:24:21] Paul Roetzer: It doesn't truly understand and have the context the way a human would. It's just really good at synthesizing what a human or simulating. What a human does to the point where it's, it's extremely convincing to where you think it, it talks like a human. It does.

[00:24:40] Cathy McPhillips: When I was doing Macon, I was writing some Macon promotional copy and it said, ma marketing AI conference held in San Francisco.

[00:24:46] Cathy McPhillips: Yep. But if you look at where the tech conferences are held, where marketing conferences are held, I mean, that's a pretty good assumption. That's where it's going to be.

[00:24:53] Paul Roetzer: Right. So it thinks any AI conferences in San Francisco? Yeah. Okay. And then the writing detection tool you mentioned so quick there, they don't work.

[00:25:03] Paul Roetzer: Just point blank. If your, if your, you know, your teachers at your kids' school are using AI detection tools, or if your company is using 'em in some way, they do not work. There is no path for them to work. Google OpenAI, everyone has talked about this publicly. No one knows how to solve this. There's some theories that you're going to be able to watermark the contented outputs.

[00:25:22] Paul Roetzer: There is no scientific proof or clear path right now. To have AI detection tools work at scale because someone will just build an AI tool that remixes the output. So the detection tool can't find it. We are entering a phase of AI versus ai and at the end of the day, they don't work. It's just the simplest way to think about it.

[00:25:42] Paul Roetzer: Okay,

[00:25:43] Cathy McPhillips: number eight. How can someone verify the content that AI generates?

[00:25:48] Paul Roetzer: So again, this kind of relates to the detection tool. You don't know if it did, but I think this gets more into if it's stating facts, are they actually facts? And the honest answer right now is a human has to verify it. So the way that this works, again, since it's just predicting words, the way that the language model companies like OpenAI and Google and Microsoft are trying to solve this is you give a prompt.

[00:26:15] Paul Roetzer: The AI creates an output, writes an article, then another AI kind of thing. It as like an AI search engine goes, finds sources that seem to verify what the AI wrote. And then it'll, it'll link in citations, quote unquote citations or resources that seem to verify where that content came from. So then, then it'll give you the output and then it'll say, oh, this is, this came from here.

[00:26:38] Paul Roetzer: Well, it's not, didn't actually come from there. So the way it's doing it is by trying to layer over a search engine in real realtime to do this. So Google Bard, again, if you haven't used it, bard.google.com. You can go try it right now. Theirs has searched natively built in and in in realtime data. So you could ask it who's in the b a finals or who is the top pick in the N NFL draft this year?

[00:27:01] Paul Roetzer: Or who's the president of the United States like it, it would know it because it's with a search of realtime database ChatGPT out of the box stopped training in 2021. It doesn't know anything beyond September of 2021. So the way they solve that is by doing a, a browser plugin that then adds realtime data to these outputs.

[00:27:20] Paul Roetzer: So, That's how they're trying to do it. It, it's not incredibly accurate. So if the content you're creating you, you can't be wrong. You can't make errors in, so for example, just recently we heard of the attorney who submitted a legal argument to a judge and had ChatGPT write it, and it made up all these citations and the lawyer didn't know ChatGPT makes stuff up.

[00:27:41] Paul Roetzer: So he submitted this legal argument to the judge and got caught falsifying citations and had to go in front of the judge and admit that he didn't know ChatGPT made stuff up. So if you're working on something where you cannot make a mistake in names, places, facts, books, things like that, the human has to verify any everything.

[00:28:00] Paul Roetzer: And so oftentimes there's really no time saving from having AI do that because you have to spend as much time verifying everything it said. Yeah, that's a lot. People don't realize that like it's, again, so many people are just like experimenting with ChatGPT and they don't know the ins and outs of this, and that's a really important thing to know about these tools.

[00:28:21] Cathy McPhillips: And it's hit the masses. It's not just marketers or, you know, my dad, my 75 year old. Oh yeah. Like,

[00:28:25] Paul Roetzer: what's this ChatGPT. I get texts from family all the time asking me about this stuff.

[00:28:32] Cathy McPhillips: Okay. Number nine, can AI replace or supplement roles like graphic designers and how accessible are these AI tools for non designers to create their

[00:28:40] Paul Roetzer: own visuals? Yeah, so for me, I think the simplest way, and I'll keep this answer extremely concise, is they are super powerful assistance and co-pilots.

[00:28:51] Paul Roetzer: So I think the best way to think about AI is a very powerful intern or assistant or in some cases like data analytics, data analysis. It's going to be like a PhD level assistant. You're going to be able to ask questions of your spreadsheets. You won't have to build pivot tables at a human. You can just ask it and tell to build you visualizations.

[00:29:10] Paul Roetzer: So I just think that there's certain knowledge tasks where it's, it is assisting you, but it's sometimes at a very high level. Right?

[00:29:18] Cathy McPhillips: So, you know, I, with the podcast, I'm using Canva to remove backgrounds from images. I'm making pretty basic images. But for us, that works for that particular need for descript, I'm using it to delete words, to edit theran, edit the video.

[00:29:33] Cathy McPhillips: So, you know, it's a nicer video, but there's things like, you know, you ask me about a speaker reel. I'm project managing, putting together a speaker reel for you, for you right now. I can't do that. I won't pretend to do that. I'm not going to use an AI tool to help me do that. Cause it's not going to give us the output that we want.

[00:29:49] Cathy McPhillips: So the professionals are still needed in some, in most of those instances. And quite frankly, they don't want to be doing things like what I'm doing. Yeah. They want to, I mean, they want to be creative. They want to use their talents for much bigger things. So, There is still a need for all of those pros, not just

[00:30:04] Paul Roetzer: people like me, for sure.

[00:30:07] Cathy McPhillips: Okay. Number 10. What are the potential impacts and opportunities for agencies as AI technology advances and how can they adapt today to ensure they aren't

[00:30:15] Paul Roetzer: left behind? Yeah, so again, if people don't know my background, I owned, an agency for 16 years and so I, you know, I think a lot about this side.

[00:30:24] Paul Roetzer: I'm not involved with the agency anymore, so I don't spend every day of my life thinking about this. But I do see AI having a massive impact on agencies, both, the financial models. So if they're still charging billable hours, that gets very challenging when you can do things 10 times faster to how do you make enough money if you're doing things that efficiently.

[00:30:41] Paul Roetzer: The service mix, evolves their own, how do I build a smarter operational engine for my company using ai? There's lots of those con considerations. But I, so I think there's a lot of threats to agencies that don't quickly evolve and figure this stuff out. But the opportunity is, ev literally every business in every industry across the world has to solve poor ai.

[00:31:04] Paul Roetzer: There are very few of them capable of doing that. So I think there's an enormous ecosystem to be built of agencies that develop specialties in advising and consulting on ai, and then leading integration projects. So for example, every enterprise is going to have some form of a, a custom, large language model that's trained on their proprietary data, their proprietary knowledge base that no one else is going to have access to.

[00:31:30] Paul Roetzer: That helps marketing, sales, service, all of different elements of the organization. Who do you go to for that kind of guidance? You want to share what language model to work with, which application company to work with, how to integrate it into your tech stack. So there's, I mean there's going to be billion dollar consulting practices built just on advising on large language models and doing the integration work.

[00:31:51] Paul Roetzer: So that's, if I was running an agency, that's what I would be thinking about is, yes, this is going to be disruptive. It might be a little painful for the next year or so, cause it's going to cause some maybe uncomfortable changes in the organization. But if you look out ahead and say, but what are the big opportunities?

[00:32:06] Paul Roetzer: They're massive. Wait dwarfs. When I, when I saw the opportunity with HubSpot and sort of was the origin of that whole ecosystem, which, I mean HubSpot's a, what, I don't know, 25, 30 billion company, their, their channel partner program at one point was making up 45% of that revenue. That was the thing we were the originator of.

[00:32:23] Paul Roetzer: And I look back now and think this is 10 times that size, like a hundred times the size of the market potential that could occur in the services ecosystem. So yeah, I mean it's, in my opinion, it's a hard time to run an agency, but it's a, it's a phenomenal time if you have the vision to do, to do this, to integrate AI into your organization.

[00:32:45] Paul Roetzer: Yeah, it's so much bigger than billable hours. Ah, don't get me started on billable hours.

[00:32:52] Cathy McPhillips: Okay. Skip past that real quick. Number 11, in terms of language capabilities, how far have AI tools come? What can we

[00:33:01] Paul Roetzer: expect moving forward? Yeah, I mean, in terms of how far they've come, just go use GPT-4 yourself. You'll, you'll see like it is, I mean, I truly, there's almost daily, there's a use case for GPT-4 that I'm, I'm in shock that it does it, like I know how these things work.

[00:33:17] Paul Roetzer: I know they were trained. I have been testing them since the beginning. And there's still times it'll do something and I'll sit back and try and figure out how in the world is it doing that? Like just as an example, if it's trained just to predict the next word, and it's trained on all the content on the internet, a lot of the content internet's terrible, poorly written, no vision for it.

[00:33:37] Paul Roetzer: No, no solid tone, no good style, all kinds of grammatical layers. How does GPT-4 write almost perfectly and creatively? Like if it's just learned from the median of all of our best, all of the content on the internet, how does it all of a sudden become in the top like 0.1% of its creative ability? I don't actually know the answer to that.

[00:33:59] Paul Roetzer: I'm not sure that Microsoft and Open AI know the answer to that. Like it just somehow does this and it is a bit. Magic. Like it's, it is certainly science, but it does seem to have some magical elements to it. Like, I don't know how it does it sometimes. So I think it's come extremely far, but I also think this is it.

[00:34:17] Paul Roetzer: Like this is just the beginning. This is the dumbest version of AI we're ever going to experience. And so it's fair to assume that every 12 months or so double its capabilities. And like with the current language models, you can't inject an image or a video and ask it to learn from it. That's going to change.

[00:34:35] Paul Roetzer: These things will be able to learn from videos and images and audio and code, and they're going to be able to output, that stuff like in Google Bar, it will now output an image with the text. So you're seeing the ear, what's called multimodal. You're seeing the early phases of this, all these, you know, multimedia components being baked into a single engine, and that's where it's going next.

[00:34:57] Paul Roetzer: As well as being able to take action, not just output something, be able to go take action, like book a trip. So if I ask it. Taking a family of Ford, Italy, here's the city I'm staying in, what should we do? And it comes back and it says, oh, here's the five main attractions and here's the best family friendly things.

[00:35:14] Paul Roetzer: And then I say, okay, go book this for me. That's where we're going next is it'll actually be able to do the things once it, it's created. And so that's all very near term. The tech exists right now. It's being fine-tuned. And so in 2023, you're going to, you're going to start experiencing this outside of just the labs and kind of the people who are out in the frontier testing these tools.

[00:35:35] Paul Roetzer: When I was in Cincinnati

[00:35:36] Cathy McPhillips: two weeks ago, speaking of that social media show and tell, the Matt, Matthew Dooley, who r who ran the event, he said, we should watch this in like six months and see if our presentations are still, I'm like, I don't want to watch my presentation again, but I could just imagine that what we're like, this is amazing, is going to be so mainstream.

[00:35:53] Cathy McPhillips: Yeah. And so, you know, like you said, it's the dumbest version of everything that's going

[00:35:57] Paul Roetzer: to be happening moving fast.

[00:36:00] Cathy McPhillips: Okay. Number 12. Given the rise of privacy concerns with AI tools, what guidelines should companies follow in using AI models like ChatGPT and adjusting content

[00:36:09] Paul Roetzer: for q and a? We always advise people they need generative ai, policies in the organization that shares with their team.

[00:36:17] Paul Roetzer: How do, like what these tools are, how and when to use image generation, language generation, video generation, what they are and are not allowed to do. And I would make that a, a very high priority in every company because people are using these tools, maybe not with approval internally, but where we've seen a lot of people get in trouble is they don't know that you can't put confidential proprietary information into ChatGPT.

[00:36:42] Paul Roetzer: They don't know that that stays with open AI and potentially is part of the training of the next foundational model they release. So you'll have people take a confidential meeting, say and export the transcript out of Zoom, and then drop it into ChatGPT and ask it to do a summary of it or to find action items from it.

[00:36:58] Paul Roetzer: Well, you just gave confidential information to a third party company that you have no agreement with. It's going to stay confidential. So I think just education is so critical here to avoid missteps that gets it and legal upset and hinders your ability to do this stuff moving forward. So as soon as you can possibly get generative AI policies in place and responsible AI principles that, you know, talk at a higher, more macro level, here's how we view overall AI and our humored centered approach to it and things like that, but get policies in place as quickly as possible.

[00:37:30] Paul Roetzer: And policies

[00:37:31] Cathy McPhillips: when it comes to your partners and agencies.

[00:37:34] Paul Roetzer: Yes, you want those to carry over because again, if you hire an outside agency or an outside in, you know, freelance writer or freelance designer, videographer, and they're using AI to generate things and then passing them to you under a work for hire agreement, your assumption is you have a copyright to those.

[00:37:49] Paul Roetzer: And in that instance, you in fact don't because you don't own what they created.

[00:37:55] Cathy McPhillips: And I don't think this is, is one of our questions, but do you think the US copyright laws will, will evolve with everything that's been

[00:38:01] Paul Roetzer: happening? Yeah, I think, well, what are we on, are we on privacy? Gets into the next one.

[00:38:05] Paul Roetzer: You're, yeah, we're on number 13. We're like rolling right into number 13 here. Oh. And I am just, wow.

[00:38:10] Cathy McPhillips: Okay. This is our number 13, our two part question. How can companies ensure they comply with copyright regulations when generating images with

[00:38:17] Paul Roetzer: ai? Yep. And then how does, how's it evolve? So, US Copyright Office on March 16th, 2023, released updated guidance that said a prompt is not authorship, that a human must author something in order to own a copyright on it.

[00:38:30] Paul Roetzer: This applies to anything you would apply for a copyright for including articles, you know, anything you've written, website content, brochures, images you generate, including your logo if you used AI to generate a logo, videos, audio code, all of it. So currently their guidance is if you submit something and want to copyright to it.

[00:38:51] Paul Roetzer: And you cannot, claim it if AI generated it. And they, they're, they're pretty strict on it at the moment. They don't even give much leeway of, yeah, but I edited it a lot or, you know, 20% of it I edited. Then they're basically saying, okay, well the 20% you edited, you can get a copy it on, but the other 80% you can't.

[00:39:10] Paul Roetzer: So our assumption is this is going to take years in the legal system. There'll be lots of battles fought over this. It will likely evolve. I think the first thing that would evolve is what level of prompt does equal authorship? I think I could, I could see a scenario where the more extensive the prompt, the more detail given the more human strategy and creativity that goes into the prompt, that they would listen to an argument that that is in fact a form of authorship.

[00:39:38] Paul Roetzer: And so I think that's probably the first thing that would evolve. I am not an IP attorney, though. I always tell people, talk to your IP attorneys about this. And again, if you're using outside agencies, you have to know if they're using AI to generate what they're giving you. So that's the safest bet right now is assume if you use AI to generate something and you publish that, put it out into the world, it becomes fixed in the world.

[00:40:01] Paul Roetzer: You don't own it. So again, the trick here is you can go into Mid Journey, create a logo for your startup, and it may be amazing, and you create that logo and you have commercial rights to it under your agreement with Mid Journey. But if AI created it, and I know you use data to do it, I can take it, put it on a baseball hat and start selling it for 20 bucks, and you can't do anything about it because you don't actually own the output.

[00:40:24] Paul Roetzer: Nobody owns it. So it's free reign for me to commercialize it if I want to. So that's where it's just like you really have to be careful here, especially for higher profile things that you really want to protect with a copyright. But there are some

[00:40:37] Cathy McPhillips: things, you know, we've talked about it before, outlines emails that you don't care if you can't

[00:40:44] Paul Roetzer: copyright it.

[00:40:44] Paul Roetzer: Social media shares like, yeah, I mean, again, the IP attorney may not say that you and I say that, but yes, there are plenty of use cases that have nothing to do with creating a final output that you put out into the world where AI can be used. That's why we often say like, I don't think about AI writing, like true writing from scratch, starting and writing a draft as even a top 10 use case.

[00:41:06] Paul Roetzer: For me, it's just not what I think about. But we use it for transcription, summarization, outlines, ideation, simplification of content, putting into specific formats, improving the ability for it to drive outcomes. Like we use it all the time. That has nothing to do with something we couldn't get a copyright on, you know, so it's just knowledge.

[00:41:25] Paul Roetzer: Like once you know this stuff, it's like, okay, who do I gotta go talk to to figure out how, what this means to us?

[00:41:30] Cathy McPhillips: Simplification was one I just heard about when I was down at Kent State a few months ago doing guest lecturing. One of the students said he took a chapter of one of his history books. He's not, he said, I'm not a history buff.

[00:41:41] Cathy McPhillips: It's a requirement required course. Put it in GPT-4. And he said, can you summarize this for me at,

[00:41:47] Paul Roetzer: seventh grade level or something? Yeah. And he was like, it was great. Yeah, it does. And it's, it's incredible at it. Yeah. Yeah.

[00:41:56] Cathy McPhillips: Do you answer both those, how do you envision the evolution of copyright legislation in the context of large language model ai?

[00:42:03] Paul Roetzer: Yeah, I think I, like I said, I think it's going to evolve. I think there will be a little bit, other things become more lenient, I guess, in terms of what, what is authorship and the significance of the prompt itself, you know, the output. So if I say, write me a 500 word article about language models and it writes it and I publish it, no way.

[00:42:22] Paul Roetzer: Not not getting a copy report. And they're not going to change that law, I don't think. If I say, Write me a article about large language models. Consider the implications on the healthcare industry. Look into how, you know, the threats and opportunities for larger enterprises. Consider this from the perspective of a brief that's going to be supplied to the CEO, and then it does it.

[00:42:41] Paul Roetzer: And then I go on and say, okay, now can you make these more concise? Can you condense this into 10 bullet points? Can you just, like, that to me is way closer to authorship than, you know, me just asking seven word prompt. Write this thing. Right? And I think I, you know, maybe there is some precedent here.

[00:42:59] Paul Roetzer: Again, I haven't really thought this through, and again, I'm not an IP attorney, but some precedent in ghost writing, like, you know, a lot of, if people don't realize this, most books you see written by executives, not ours, I actually wrote my book, but if, if most books written by CEOs aren't written by the CEOs, they have ghost writers, whether they acknowledge them or not, there is someone who interviews the ceo, e o reads everything the CEO is written, said, whatever, all the interviews.

[00:43:24] Paul Roetzer: And synthesizes that into a book that the CEO e o hopefully reads or someone on, the CEO's team reads and approves. And then you publish and you have a New York Times bestseller by a CEO who never wrote a word of it, but they own the copyright to it. So I I think, again, If I put my legal hat on of, I've spent a bunch of money with IP attorneys through the years, and you learned a few things, I could certainly see the argument of we have ghost writing and the CEO owns the copyright.

[00:43:51] Paul Roetzer: So how is this different? So I've always said like the home run career for the next 10 years is being an IP attorney. Like there's, there's going to be no lack of work to do in the generative AI space for IP attorneys for the next decade. Right. Okay.

[00:44:09] Cathy McPhillips: Let's get into the weeds a little bit on this

[00:44:10] one.

[00:44:10] Paul Roetzer: Last two, we're almost there.

[00:44:13] Cathy McPhillips: If you, if I want to get started today, what are some steps, marketers or anybody can take immediately to learn or identify ways to get started?

[00:44:22] Paul Roetzer: The two things we always teach. There's a whole chapter of the marketing artificial intelligence book dedicated to this. We have online courses dedicated to this.

[00:44:30] Paul Roetzer: We teach two frameworks. The first is problem-based. I think of this as the director level and above approach the leadership approach. You look at problems in the organization, whether it's related to acquisition, churn, audience growth, profitability, efficiency, whatever the, the standard problem you're trying to solve or challenge or the goal you're trying to solve.

[00:44:52] Paul Roetzer: You look at those and say, is there a smarter way to solve this? And you go through an analysis that says, here's the tech we're using, here's the people we're using, here's how we're doing this. If we take kind of a first principles, principles approach to this and say, let's start from the ground up and forget everything we know.

[00:45:07] Paul Roetzer: And say, how could we solve this in a more intelligent way? That's the problem-based model can take months to do it. Oftentimes these are a hundred thousand or multimillion dollar solutions. You're looking for things that will create and unlock value in the organization. So a lot of people are involved.

[00:45:22] Paul Roetzer: That's the, it's important, but that's kind of the longer play. The one you can go do after this episode ends is the use case model. That's make a list of all the things you do. If you're Cathy, make a list, sit down, look, think through the last week and say, what are all the things I did tactically? And put an hour number next to how much time you spent on each of them.

[00:45:40] Paul Roetzer: Extrapolate out over a month and say, wow, I am spending 20 hours a month producing the podcast. Can I find AI to, to do 10 hours of that work? And so you just go through your list of the things you're doing and say, can I go find a tool to help me save time here? And then you just force rank these in terms of which ones have the greatest value to you if AI can assist you.

[00:46:03] Paul Roetzer: And then you look at how much time can be saved and you start picking off individual use cases. You want these things to be very narrow, narrow in scope, so very clearly defined. And you want to have a high probability of them creating value for them working. And then you start stacking those successes and then you take those to your leadership and say, listen, last three months we got AI tool for podcasting.

[00:46:23] Paul Roetzer: That saved us, you know, 12% of our time. We got an AI writing tool that saved us 25% of our time and we got an ad management tool that increased the return on ad spend, 40% we would love for next year, 2024 to have budget to start really exploring what we can do with ai. So that's kind of the two ways I think about it.

[00:46:43] Paul Roetzer: I'm used to being an intro where I can

[00:46:44] Cathy McPhillips: drop in that link in the chat, and we're not, we can't do that right now.

[00:46:48] Paul Roetzer: We'll drop it in the show notes, we'll drop it in the show

[00:46:50] Cathy McPhillips: notes, but marketing ai book.com by the book, if you haven't already. And there are about halfway down that page, there is, there are links to the spreadsheet, the

[00:47:00] Paul Roetzer: templates.

[00:47:00] Paul Roetzer: Yeah. For that, the templates.

[00:47:03] Cathy McPhillips: And lastly, this is a good note to end on as we move toward an increase, an increasingly AI-driven world. What are your thoughts on the future role of humans and what advice would you give to those who might feel threatened

[00:47:14] Paul Roetzer: by the rise of ai? I think about the threatened part a lot.

[00:47:18] Paul Roetzer: I talk to a lot of people who are worried, who are fearful, who want to just not have AI be able to do what it's doing. They don't want it to be able to write or to do artwork. And I get that. I sympathize with that. I feel it myself a lot of times as a writer, as a husband of an artist, as the father of an artist.

[00:47:34] Paul Roetzer: Like I feel that, like I get it a lot. I think that we basically, Have a choice. We can accept that AI is here and it's going to be ever present in our lives, or we can ignore it. I don't see ignoring it being a real option for professionals. I think that the best advice I can give is to embrace the technology, to figure out what it means to you to figure out when and how it's going to impact your career, and to just be proactive in trying to solve this.

[00:48:03] Paul Roetzer: The re the reality is that most organizations have no idea what to do about this. So I think there's unparalleled opportunities in your career to raise your hand and be part of the solution to say, listen, I'm, I'm really interested in this space. I've become curious about it. I'm listening to this podcast.

[00:48:18] Paul Roetzer: I'm demoing these technologies. I would love to help be the one to figure out what to do about this on our team, in our company. So I think there's opportunities for people at all levels to emerge as real change agents within the organizations, and honestly, to figure out what the next career paths are.

[00:48:34] Paul Roetzer: Like if you asked me point blank, what are five new career paths based on ai, three years from now, I would really struggle to envision what those roles are going to be. I could probably come up with a couple, but I think that more often than not, it's going to be you, the listener who embraces ai, thinks about what you do, thinks about the context of the organization and the industry you do it in and says, you know, what I think is needed is this, like, I'll give you one example.

[00:49:00] Paul Roetzer: I think I've told this story before on the show, but a couple years ago I did this talk for, healthcare organization for public affairs officers, like 150 public affairs people. And after the talk, a lady came up to me and she said, is AI ops a thing? Like, is there, is that a role that exists? And I said, not that I'm aware of.

[00:49:17] Paul Roetzer: Like what are you thinking? And she said, well, we're in healthcare. This is going to be disruptive in many ways. There's no one that's thinking about that right now. And that is actually figuring out how to build the infrastructure, how to evolve the team, how to evolve our tech stack. I think there's an opportunity for me in my career, she was an early twenties to be like an AIOps leader.

[00:49:36] Paul Roetzer: And I was like, do it. Like that's beautiful. And I think that's where we're at. We're entering this phase where people who push forward and learn this stuff and figure out to apply it, they're going to identify their own career paths. And then the leaders, like the CMOs who, you know, really latch onto this, they're going to be thinking about, what does a NextGen marketer look like in my organization?

[00:49:55] Paul Roetzer: And what roles am I going to need here? And so I'm, I think that's exciting. Like as much as I do fear disruption to knowledge workers and the potential for job loss as a result of this, I like to focus more on the positive aspects and the ideas that lots of new careers and opportunities are going to emerge.

[00:50:11] Paul Roetzer: And people, you know, potentially early in their career, even later in their career, if they're listening and paying attention, to be able to really kind of redefine their own career path and the value they create in their organization. I think even in

[00:50:23] Cathy McPhillips: November, you know, early December when ChatGPT came out, everyone's like, oh, prompt engineering is going to be a huge thing that's going to prompt engineer is going to be a role.

[00:50:32] Cathy McPhillips: And don't you feel like we're all going to

[00:50:33] Paul Roetzer: be prompt engineers or it's not? Yeah, I think it's just a skill. I don't think it's a role. I've said that publicly before. I do think, like, as we do interviews for the hiring, we're making, we're, you know, we're planning to make a couple of marketing hires.

[00:50:44] Paul Roetzer: That's one of the questions you start weaving in is, you know, talk to us about how you would give a prompt to GPT-4 for this. So I just feel like whether you're doing email or a podcast or building a a, a plan, that's going to be a required skill to like how, you know, if I gave someone a strategy, I think about, again, like all the people I hired and developed at my agency, how much time you spent teaching strategy.

[00:51:07] Paul Roetzer: I always said strategy was by far the hardest thing to teach someone. You could look early in someone's currency. I think they have potential. They're a critical thinker. They connect the dots, they see things other people don't see, but it would take years to develop a great strategist. I really feel like AI can advance that so rapidly because you can just say, listen, here's the problem we need to solve.

[00:51:29] Paul Roetzer: Here's the project brief. I want you to start by developing a prompt to give GPT-4, and we're going to talk through that prompt. Then we're going to give it to G PT four, and we're going to see the output, and we're going to fine tune that prompt. I think that that's a way that AI can really accelerate the development of people from a strategic and critical thinking side, if it's done right, but it is absolutely just going to be a skillset that's required of every marketer moving forward.

[00:51:52] Paul Roetzer: Absolutely. Any final words? Oh boy, I was 15. Oh, huh. So, yeah, I don't, I don't know. I mean, I just, I hope like we, again, we thought like we get hundreds of these questions. I hope this was as valuable to you. You know, I know the questions and answer for you and me, Cathy, at the end of the intro is always our favorite part of doing that every time because we do get incredible questions and some of 'em are just so in depth.

[00:52:13] Paul Roetzer: We can't go into them all, you know, in that short time span. So, I don't know. I mean, maybe like, again, if people like this, like reach out on LinkedIn, like Cathy and I know, you know, leave a comment, if this was helpful to you because it might be something that we do more regularly, like, you know, once a quarter or something.

[00:52:28] Paul Roetzer: Just make sure we pop in and address some of these things and then send 'em off to people who've done these intro classes. So yeah, hopefully this was really helpful for you. We appreciate it. Cathy. Thanks as always for not only. Co-hosting today, you know, pitching in here, but also moderating all of these intros and, you know, helping guide as well.

[00:52:46] Paul Roetzer: Great to see my community. I love, I love our community. Yeah. And if, again, if you're on the, if you're not in the Slack community, Cathy runs that as well. We have, I think we're approaching 3000 people in the free Slack community, so you can join that on our website as well. And Cathy's very active.

[00:53:00] Paul Roetzer: They're more active than I am. I try and be active in there, but that's not your job, so you shouldn't be. I know. So, so I'll leave, I'll

[00:53:06] Cathy McPhillips: put a bunch of links in the show notes to a lot of things we discussed today, and I hope you

[00:53:10] Paul Roetzer: found it valuable. All right, thanks. We will, we'll be back next week with your regularly scheduled weekly.

[00:53:16] Paul Roetzer: Weekly where Mike and I were going to be catching up from three weeks of not having given you that line. So we may have a super episode for you next week with all of the headlines that we missed in early June. All right. Thanks everyone. We'll talk with you again soon. Thanks Paul. Thanks for listening to the Marketing AI Show. If you like what you heard, you can subscribe on your favorite podcast app, and if you're ready to continue your learning, head over to www.marketingaiinstitute.com. Be sure to subscribe to our weekly newsletter, check out our free monthly webinars, and explore dozens of online courses and professional certifications.

[00:53:53] Paul Roetzer: Until next time, stay curious and explore AI.

Related Posts

[The Marketing AI Show Episode 77]: 15 AI Questions Everyone is Asking -  Round Two

Claire Prudhomme | December 19, 2023

Our Intro to AI class has generated great questions from our community. We answer the most-asked on this week’s podcast

[The Marketing AI Show Episode 64]: Top Professional Services Firms Go All-In on AI, New Study Shows AI’s Actual Impact on Our Work, and Major Predictions on Where AI Is Going Next

Cathy McPhillips | September 19, 2023

This week's episode of The Marketing AI Show covers, AI and the workforce, professional services plus AI’s impact on our work, and what’s next for AI?

[The Marketing AI Show Episode 62]: ChatGPT Enterprise, Big Google AI Updates, and OpenAI’s Combative Response to Copyright Lawsuits

Cathy McPhillips | September 5, 2023

On this week's episode of the Marketing AI Show, we break down ChatGPT for enterprise, Google’s big news, and OpenAI’s defensive response to lawsuits.