258 AI’s Impact on Marketing and the Future of Business with Paul Roetzer, Founder & CEO, Marketing AI Institute & Co-Author of Marketing Artificial Intelligence | Partnering Leadership AI Global Thought Leader

***DISCLAIMER: Please note that the following AI-generated transcript may not be 100% accurate and could contain misspellings or errors.***

[00:00:00] Mahan Tavakoli: Paul Razer, welcome to Partnering Leadership. I'm thrilled to have you in this conversation with me. It's 

[00:00:04] Paul Roetzer: so great to be here. I'm looking forward to it.

Paul. 

[00:00:08] Mahan Tavakoli: I'm really excited because. After getting deeper into ai, I've been studying some aspects of it for a few years, but seeing chat G p T made me understand that this is going to be transformative and getting deeper into it. . I ended up finding your podcast.

Which is an outstanding one, artificial intelligence podcast. Enjoy your conversations and the business guidance you provide. Then also found your book Marketing Artificial Intelligence and the Marketing AI Institute. So there is so much that I can't wait for us to get to talking about. But first, Paul would love to know whereabouts you grew up and how your upbringing impacted the kind of person you've.

[00:00:54] Paul Roetzer: I grew up in Cleveland, which is where my business still is. We're in downtown Cleveland, probably about seven miles from where I grew up. We didn't have much, it was a good upbringing. We didn't have a lot of money. I never saw a house worth more than a hundred thousand dollars until I started caddying in high school.

But we had everything we ever needed. So I was happy, very wonderful upbringing. And so I learned very on early in life that money doesn't buy me happiness. I loved my life and didn't know anything different growing up. My father worked for the Cleveland Plain Dealer, so I learned about newspapers media and journalism to a degree.

And then my mom was a preschool teacher until my junior year in high school when she. A small business. So that's when I learned entrepreneurship was a thing. Like I'd been around entrepreneurs before, but I didn't really ever know you could be one. Maybe from that, I probably took away that anybody could be one.

My mom had no training to be an entrepreneur and start a business, and yet here she was in her forties, starting a business after being a preschool teacher for 20 years. So I think I just grew up believing anything was possible. Anybody could build anything. I was naive to the fact if that wasn't true, and I think that probably carried through a lot in my life where I just never really cared for whatever, guardrail someone might put around what you're supposed to do in life.

I just have ideas and they go do 'em. , I think that's part of the reason why the institute exists. Just saw an opportunity and someone needs to figure this out and nobody's doing it. So let's start talking about it. And here we are seven years later, now everybody's listening.

[00:02:20] Mahan Tavakoli: How did that happen, Paul? A lot of people, their awareness of AI's impact was raised through interactions with chat G P T. Everyone had heard about AI and algorithms whether what Facebook uses or LinkedIn or wherever else ways, but still most people hadn't reflected much on it.

But you got deep into ai. Six, seven years back, what was it that got you to focus on artificial intelligence? 

[00:02:48] Paul Roetzer: The initial thing was actually in 2011, I was running my marketing agency, so I came outta journalism school at Ohio University in 2000, started my own agency in 2005. We became HubSpot's first partner in 2007, which sort of threw us into the marketing technology and about marketing, social media, move.

And then in 2011, IBM Watson won on Jeopardy, and that was the year I wrote my first book, the Marketing Agency Blueprint. And as soon as I finished the manuscript, I actually became fascinated with ai and what was this Watson thing and could it be applied to marketing? So that actually led me to just start trying to figure out artificial intelligence.

So it was out of a curiosity initially. And then I became obsessed with the topic and was reading every book I can get my hands on and trying to figure out what it meant. And in my second book in 2014, I wrote a theory of a marketing intelligence engine, how eventually marketing would have ai and it would be applied to marketing strategy and budget allocation and all these other things.

And it was just a small piece of the book. It was a like 700 words out of 50,000. But that became the thing everybody wanted me to do a talk on. So I started traveling the world, doing talks about artificial intelligence in 2015. I said, all right, let's start sharing what we're learning. And we launched the institute in 2016, and it was just to tell the story really initially, and then it evolved into an conference and online education and everything we see today.

[00:04:10] Mahan Tavakoli: So with everything that you're doing today, Paul, marketing a critical aspect of it over the past at least 20 years, has been content and content creation. How do you see AI impacting content and content creation? 

[00:04:26] Paul Roetzer: After the Marketing Intelligence engine idea in 2011, the next catalyst for us was actually a 2000 16 2015.

I started an internal project called Project Copy. And the actual question was can AI write content? Can we scale the production of content? Cause that's what my agency largely did, was created content to grow audiences and businesses. And so my question was, is AI at the stage where it could write content?

Because at the time there was this technology called Automated Insights that was being used to write earnings reports and fantasy sports stories and things like that. And I was like, oh, this is amazing. I can do this. And so then I bought the technology and you stared a blank screen. I'm like, where's the words?

How does it write? And I realized it was actually. Template driven and formulaic where you actually had to teach the machine how to write everything. It was just like taking data out of a spreadsheet like this is not what I thought it was. So it fast forward seven years and the answer became, yes, AI can write content, but at the time, in 2015, it could not.

But that kept us at the forefront of every major advancement in language generation. We were , oh, maybe now's the time. Maybe now's the time. And so again, it was that curiosity and that very specific pain point that we were trying to. That led us to just stay. And then around 2018, 19, you started seeing major leaps forward in this area and now today, yes, we have chat, G P t and g P T four, and these amazing technologies that are able to do these things.

with that, 

[00:05:50] Mahan Tavakoli: I would love to understand from your perspective, how will marketing change. Right now content drives a lot of what goes on, whether in social media or search. A lot of companies are spending effort and time in writing and creating content. How do you think this shift in AI writing content will impact marketing?

[00:06:16] Paul Roetzer: It changes everything because as you pointed out, there's so much language generation at the core of everything marketers do, whether it's ad copy or social media copy, emails blog posts, articles, press releases, video script, all of it, it's all language. And these advancements have made the AI very good at writing and they're getting more creative in their writing style.

They're able to infuse logic into it to actually understand what it's writing and whether or not it makes sense and improve on it. You can prompt them to make this better, find the errors and what you wrote it's really quite shocking if you haven't gone in depth what these things can do to see what they're capable of doing.

And so I think that it's becoming a almost indispensable writing assistant. You have to be really careful with how you use them because the US Copyright Office, so if we're talking to the US audience the copyright office on March 16th, 2023, issued new guidance that said you cannot file for a copyright for anything that an AI generates.

And that's written word, that's images, videos, anything. You cannot get a US copyright on that. So if you're a brand, if you're a leader at a company, you're hiring an agency to help write or using a freelancer. If anyone is using AI to write the content and not significantly editing it, you cannot protect that content.

So that might be okay for emails and things. You're not gonna worry about filing. But if you're talking about blog posts and video scripts and things like that it's a really important distinction that most people aren't even aware of. So I think it's gonna have a transformative effect. It's gonna be infusion in all aspects of writing and content creation, but with the caveat that you have to be really careful that you can actually protect the content you're creating.

[00:07:59] Mahan Tavakoli: So with that Paul, I heard the conversation Lex Friedman had with Sam Altman. Yeah. And he was talking about some of the implications of AI tools such as G P T, and one of the things he was mentioning is, for example, he was saying there is gonna be need for drastically fewer coders, so one coder could get the job of 1,000 coders done very effectively.

Part of what you're saying is these AI tools can be used as assistance in creating great content. Will that also be the case for creating content in. One person can scale and in essence do the job of 2000 content creators through the AI 

[00:08:45] Paul Roetzer: tools. I believe, yes. I do think that our general guidance to people has been that, using AI tools will give you the advantage and hopefully protect you as these, tools come on.

That the tools themselves won't replace you, but that, people using the tools will. And I do believe that to be true. Let's say you had a hundred writers on your staff or in your, media company and 10 of them adopt AI and infuse it into everything they do, those 10 can do the job of a hundred.

So will people lose their jobs? Probably. I don't want that to be the case, but I've already seen it happening. In the current economy, you have to be realistic. If their opportunities for publicly traded companies or private equity owned companies to save 20 to 30% and improve efficiency, 20 to 30%, reduce cost 20 to 30%.

By infusing these tools, it's gonna get done it's unfortunate, but that's what's gonna happen. Now, if the economy was humming along and life was great, then I think the redistribution of those resources and those human resources into, more fulfilling aspects, hopefully we would've a chance.

But I think in the, next 12 to 18 months, we have to probably be realistic that there's gonna be significant job loss because of this stuff. 

[00:09:56] Mahan Tavakoli: Part of what I see Paul Is that we've been through transformations technological and industrial in the past.

Those have typically taken a lot longer to cascade through society. Part of what I'm seeing is with artificial intelligence, the pace of this change is going to be a lot faster than before.

[00:10:15] Paul Roetzer:  It's a various astute point. It is a hundred percent accurate. There's a lot of people like to say, oh, we've been through technological revolutions before.

We haven't done it in four months. Like it's unprecedented in human history to have technology this advanced happen this quickly. I do believe that in the end, there is probably some net positive on the economy and on the workforce. I think there will be entirely new career paths created.

They're not gonna get created as quickly as we're gonna lose jobs in the near term. So it's like an evolving point of view from me for anybody's listened to me before maybe listening to this. But my perspective has evolved quite a bit in the early parts of 2023, based on some data and conversations.

I honestly think that the infusion by Google and Microsoft where they're embedding this stuff right into Microsoft 365 and Google Workspace, that actually started changing my perspective because you realize how much knowledge work is going to be affected as soon as all these tools are baked right into the core platforms you're using.

Now, it doesn't take a C I O or a C M O or a VP of something to go find a new tool. They're just gonna live right where you. And when you start looking at all the knowledge work that's done, that it's gonna start automating again, C-suite, the board, the stakeholders, they're not doing their job.

If they don't look at that technology and say, we don't need as many people, 

[00:11:47] Mahan Tavakoli: That's where I would love to get your thoughts first on the marketing perspective and then general broader organizational perspective. Paul. Yesterday after an economic Club of Washington luncheon, I had a brief conversation with a dear friend who runs a marketing and advertising agency.

And actually she is more aware than most of the business executives I'm interacting with. Most of them have heard of chat, G P T or might have played a little bit on it, but in my view are truly clueless about the potential impact of artificial intelligence beyond chat G P t writing an acute poem or to, but she said, 

now with a prompt we can create content, brochures, images. So that means what we have done and what we have charged for years and decades in terms of the value we've brought to clients, now all of a sudden is pretty much automated or has become a prompt.

So where do we go? Therefore, the question to you, With all of these things having been scaled, where is the value add that marketers can bring to the process? 

[00:13:03] Paul Roetzer: That's a really good question. Without a very clean answer. The service industry, so agencies, certainly law firms, accounting firms, any professional service firm is gonna be grappling with the exact same issues.

The clients, the people who pay them for their services are going to realize that those services don't cost as much to produce anymore. That you may need one person instead of 10 to do the thing. So I think there's gonna be a lot more questions that are started to ask of service firms of how are you using ai?

How much efficiency are you gaining? Do we really need to spend 10,000 a month with you? 20,000 a month? Can it be 5,000 a month for the same output? I think there's a lot of hard conversations that are gonna start being had between clients and service firms in the very near future. I do think you still have to grapple with the examples you just gave of, can you use AI to generate logos?

No, you actually can't. Because you have to heavily manipulate them, so you can use 'em as a base for ideation and things like that. But you certainly can't just replace the human in that role. Same with content. So there's this balance between. Really, how much can we remove the human from the equation if we still wanna be able to protect the assets we create?

So that's one component. That doesn't solve everything though. I think right now we have to figure out what is uniquely human strategy is still largely to a degree where you have to know what to prompt the machine to do. And the hard part where this goes and gets tricky is you still have to be the expert to know if the output's any good.

So let's stay on logo designs. Let's assume I can use AI to do logo designs and I can protect it with a us, copyright and trademark. If I'm a designer, I know whether the designer gave me as any good or not. If I as a non-designer go in and ask AI to do a logo, I'm like, oh, that looks pretty.

My designer from no, it's terrible. The composition's off, the colors off, whatever the designers think about. So if you think about writers, designers, coders, the reason they can assess the output from the AI is because they're an expert in what it's outputting. If I ask Chad p t to help me develop a business plan to launch a startup in the healthcare space, it may give the most impressive sounding thousand word businessmen Any of us have ever.

Put someone who's in that space and be like, this is terrible. This doesn't even make any sense. Those five things aren't relevant. I'm like, really? They sound really good and they're confident in it. So where the models are today, they need a human expert to determine if the output's actually any good.

So we still need the human strategy, the human imagination, the human creativity. The question you posed earlier is the right question is like, how many humans do we need to do that? Can one really good consultant do the work of five consultants previously? And that I think is the great unknown that determines whether we lose a million jobs in 2023 or 15 million jobs in 2023.

And I don't know where it's, but it's somewhere in that range. 

[00:16:03] Mahan Tavakoli: It is going to be significant impact. And I am surprised at how few very smart business people. No, very little about what's going on with AI in the conversations that I have, it it's 

[00:16:16] Paul Roetzer: shocking to me, Paul, you and me, both heads of universities, CEOs of major companies, CIOs of major companies, again, you maybe listen to this and think, oh my gosh, like I am so stuck. I know nothing. Join the. Most people know nothing, like just listening to this podcast alone. You're probably going to know more than your peers in the same roles in government and in organizations. It is shocking and it's scary. 

[00:16:42] Mahan Tavakoli: Thanks for that shout out. And I will say listening to the artificial intelligence podcast is the best way to keep up because you have outstanding conversations every week, and I believe you mentioned you might want to go to twice a week 

[00:16:58] Paul Roetzer: yeah, but I've been saying lately on the show is there's so many really messy topics like AI and education and. I can offer some perspectives but I am not the expert in it. And I think there's just so many topics where we're heading into where we need to dig deeper than one of our three topics for the week.

We gotta dive deep. And I think that's where we're probably gonna start adding another one where we start interviewing experts on really important topics to go deep on these. 

[00:17:27] Mahan Tavakoli: So a couple of other areas I wanted to touch on with you, Paul, is you also both wrote a blog post and I believe talked about it briefly on your podcast about the potential implications of an organization like Bloomberg developing a version of generative AI of their own.

Can you talk through what Bloomberg has? And what that therefore means for other organizations. 

[00:17:57] Paul Roetzer: So the simple way of thinking about is they took a model, like a chat G P T or g p T four, the underlying language model, and they trained it on their own proprietary data that public doesn't have access to.

So now they didn't use G P T four, but they built a model like a G P T four, and then they trained it on all its own data. So if you think about if you're going to use chat, G p t, it has been trained on the open internet. So the model goes out they gather all this content. And it learns how to predict words in a sentence based on the probability of the word occurring from this corpus of knowledge it has.

So it's not trained on anything behind your firewalls, anything on your server, none of the stuff it can get to does it know. So imagine being able to take chat G P T and then dump all your C CRM data, your business analytics data, your finance data, your HR data, everything, and be able to control that in a secure way.

Keep all that data private only to you. And now all that knowledge lives within that chat interface. So now if I'm trying to remember that engagement I was on three years ago with that one client in Omaha, I can just run a quick query and rather than getting a bunch of potential files on my server, I'll just get a narrative response from him.

Oh yes, like these three people went on that trip and here was what they talked about. That's what we're talking about is this immediate access to knowledge because it's trained on it. And so Bloomberg took 40 years of financial data, private financial data, and trained this kind of thing for an internal tool.

So I believe that's where we're going with all companies is you'll eventually be able to just upload all your data. Here's all our videos, here's all podcasts, here's all our knowledge base, here's all our docs, here's our server. Like here have it all. Text, image, video. And the AI will just learn it all and then you'll be able to interact with it.

And that's the direction of custom language models, 

[00:19:48] Mahan Tavakoli: we would call it. That by itself is worthy of hours. Long conversations, Paul, because it's incredible for as long as I've been in training consulting. Knowledge management for organizations has been a real major issue and large consulting firms have gotten paid lots of money to try to put together knowledge management systems, which have been very clunky, inaccessible, and not usable in most instances.

This potentially makes the knowledge in any organization accessible. To any individual to have a conversation with and understand from their day one in the organization, 

[00:20:31] Paul Roetzer: and likely with minimal to no human annotation, because what you're describing is thousands of hours of categorizing tagging, building taxonomy, figuring out all the structure, which then breaks all the time.

This is just give it the data and it learns from it. It's able to actually identify objects and people from images and videos. You don't have to teach it anything, it just learns when you feed it data, it's oversimplifying the possibility, but that's where it's all heading. 

[00:21:00] Mahan Tavakoli: And the process is somewhat simpler than I had initially thought.

I've tested a couple of tools with my podcast. Any of the visitors can have a conversation asking questions about leadership based on all the conversations I've had in the podcast. So it's not a real cumbersome tool for these chat bots to train on the content you already have. So this is not just something that would be impactful for the Bloombergs of the world because I know I have listeners of organizations of various.

All size organizations can take advantage of an opportunity like this. 

[00:21:38] Paul Roetzer: Yeah. I've seen some friends, like author friends who've been feeding at their books, their podcast, their blogs, and just creating a chat interface like you just explained, just ask questions of this whole knowledge base. So If I took everything I've written over the last 20 years, there's probably a few million words and everything I've said in video.

If I just took all those transcripts and all the manuscripts and all the blog posts and just here you go. You could in essence have a conversation with me so there's gonna be a entire ecosystem. I actually hadn't thought about this until this moment, but there's gonna be a subscription-based ecosystem to experts.

So if you're Tony Robbins, let's. Tony listens to this show. Here you go, Tony. Here's the next billion dollar business. And you just, you train an interface. You could even do it a synthetic avatar, like you could actually talk to Tony if you want, but it's trained on everything Tony's ever said. Then I could, in theory, just go in and I could pay $49 a month to be able to have Tony as a on-call advisor.

Damn, we need to build that. Actually really smart that is, it's totally gonna happen. 

[00:22:45] Mahan Tavakoli: Somebody's gonna do that. That is an outstanding thought. And Tony is running right now. He's runs really fast and sleeps only four or five hours a night, so he's going to be creating that. I wonder, along with that, Paul, here is one of the things that happened in the music industry as a result.

Streaming and as a result of globalization, where at one point or a time there could have been outstanding singers in every little town around the world, they would've made substantial amounts of money. But because of the transition to the internet and digital music, you have a few people on top Taylor Swift or whatever else. You like a few people on top making lots and lots of money and very few locals, so you don't have the local Taylor Swifts that existed all around the globe. So I wonder if that's where content is headed. Yes, Tony will have lots of people going to his site to have Tony as their coach and.

But that will reduce the opportunities for thousands, if not tens of thousands of other people that are serving as guides in their local communities and with their 

[00:24:10] Paul Roetzer: organizations maybe. , I totally agree with that perspective, and I'm just thinking out loud here. Think about the influencer movement on social media and the big media companies missed it Mr.

Beast is the biggest media brand in the world, and this is just a dude who started putting out YouTube videos in 2012, and six years later, people started watching 'em. Maybe there is still this creator economy where people, take all their TikTok scripts. They've been doing a TikTok video every day.

Cause they follow Gary V and Gary v told him to post eight times a day on TikTok. So they've got, five years worth of TikTok videos they can train on cuz they were a creator. Maybe there is a whole ecosystem around. I have no idea. You and I just created this whole concept in the last five minutes.

[00:24:51] Mahan Tavakoli: I'm going to create my Paul Razer version of it and have conversations with. In trying to guide me and guide organizations I work with through transitioning to what ends up becoming an AI-based future. I love the way you talk about some are AI natives. And others need to be able to embrace ai. So as you are having conversations with people that lead organiz, How do you guide them to approach this? How do you recommend for them to test, experiment, and use AI in a way that serves their purpose 

[00:25:34] Paul Roetzer: best? The three main areas we talk about are understanding, piloting and 

scaling. So in understanding, it's all about education. You need a baseline understanding within your organiz. You need to start developing your responsible AI principles, your generative AI policies, like you need to be putting the framework in place for any employee in the organization, understands the point of view, understands their future role, so you're not just rolling out a bunch of AI tools and people are like, when's it gonna come for my job?

You don't want that kind of culture. So I think the education is critical to create the understanding, because once you understand something, you can embrace it. If you don't understand it, you will. So we have to break through the fear factor in organizations and get rid of the uncertainty. The piloting is where you start to look for quick win pilot projects that you can be enacting, where you can build success stories and really start seeing the impact AI can have.

And then the scaling is where you're truly thinking about an infusion of AI across business functions. You're working on solving bigger business problems with ai. And so you're gonna start working on the pilot and scaling in unison. You're gonna continue to stack, 1, 2, 3 pilot projects per business function per quarter, where you're like, continue building that success.

And then you start working on, maybe it's one business problem to start and then you like maybe you eventually build up to okay, hrs got an AI project each quarter finance has an HR project, legal marketing, sales service ops. You really think about it as an organiz. So that's the way we think about it, is that it has to be layered across the organization.

This is not AI for marketing or sales or service, and then you're done. It is an AI first view of the future. 

[00:27:12] Mahan Tavakoli: In doing that, one of the challenges that I've seen is a lot of organizations are on a cadence of reviewing or rethinking their strategic plan once a year. Which I think it's way too slow in the age that we live in.

So how do you recommend organizations approach that, and how do you view AI's incorporation into that strategic thinking for the organiz? 

[00:27:41] Paul Roetzer: Right now what we're doing is guiding people to build three to five year AI roadmaps where you're basically developing a vision of what does a smarter version of our company look like, and then how do we prioritize the use cases and projects over that time period.

And then you're constantly looking at that. So to your point, you cannot wait 12 months, next week open. OpenAI could announce some other major breakthrough or meta or aws or somebody could come out with some breakthrough that just changes everything. So for example, When Microsoft 365 co-pilot launches to the masses, it changes everything.

If you're a Microsoft customer or maybe you wanna become a Microsoft customer because of it, once you do that, you may be able to get rid of 10 generative AI tools that you previously purchased. So you have to stay at the forefront of it. You have to have a team, an AI council, whatever you wanna call it.

You need a team of people who want to stay really close to this constantly looking at the plan and saying, okay, we need to meet as a group because we. Q4 has just changed and we need to revisit what the plan was. 

[00:28:41] Mahan Tavakoli: I love that, Paul. One of the things that I mentioned to some of my clients is that many of the organizations that I worked with did a magnificent job in transitioning through the first stages of the Covid pandemic.

Yeah, in putting together teams that were meeting on a regular basis, saying, okay, this is what we're gonna do with technology. This is how we're going to keep our people engaged. So there was an ongoing focus on that in transitioning their organizations initially. To remote in many instances, and then eventually talking through how they get their people back or whatever they needed to do with AI strategy.

Also, from what I'm hearing from you is that it needs to be an ongoing part of focus because this is truly a transformative moment, possibly more so than the Covid pandemic hitting was. So you can't. Until we do our strategic planning in October, November timeframe, let's think about it then. Yeah, it's 

[00:29:40] Paul Roetzer: an interesting parallel.

I hadn't really thought about it before, but in many ways what organizations had to go through with Covid is probably the best parallel to what needs to happen now. That was real-time decision making, daily monitoring of news and information. Potentially changing long term, how the organization work, like moving to remote work, for example.

Those are the kinds of things you're gonna be faced with. And yeah, I think a lot of the processes and workflows that were established and the operational manuals that were established to deal with covid may actually be transferrable to what we're about to go through with ai. 

[00:30:18] Mahan Tavakoli: So as organizations think through, Paul, I'm also very actively involved in the community here, and as I mentioned to you before we started recording, I am surprised a lot of people have not reflected on what is going on and its impact on the community. And then on the other side, some have been raising the alarm we need to slow this down. So there was a letter, a thousand plus people signed it.

A lot of people say it's not practical to. Generative AI training for six months. Some of the people that signed that letter, like one of my previous guests, Louis Rosenberg, says he understands that they just signed a letter to try to make some noise to get people to pay attention to this. They didn't think anyone was gonna stop it for six months, so would love to get some of your thought. With respect to the potential impact of ai and what kinds of conversations should we, as leaders of organizations and in the community have in order to try to make the transition forward a better one.

[00:31:28] Paul Roetzer: So if people aren't familiar, the letter was from The Future of Life Institute and it was signed by Joshua Benji, which is a turning award winner, Elon Musk Steve Wazniak, co-founder of Apple and some other big names. And what they were calling for was a halt to AI development of models and systems at G P T four and beyond.

So basically the most advanced AI models, we needed to hit a pause button. I don't think anybody thinks that actually is gonna happen. Maybe some of the people who signed it have some pipe dream that it actually happens. I signed the letter. I have zero belief that it'll actually happen, nor do I actually think it's probably the right idea.

But I agree with your previous guest. I signed it because I don't think there's enough awareness about all the ways this can go wrong, and in that letter they basically said it could end humanity. Is beyond a 0% chance. Like there is a probability that it could occur.

But the more important thing to me is the impact on the workforce, on the economy, on education on the environment. These are very real issues right now and there is not nearly enough awareness of those issues. More or less actual conversational, what to do about. And so I thought if nothing else, the letter was a way to wake society up to the fact that there needs to be more conversations at local governments.

We need more institutions involved, meaning more business leaders involved. We more government leaders involved. We're actually taking the initiative to figure this stuff out. And so that was where I thought the letter could play a role. And I still think that, the media's gonna run with the most sensational aspect of it, and that's fine.

That's what they need to do. But I think once the dust settles, which I hope it is starting to, we'll start getting into the real conversations around the near term impact on workforce, like you and I talked about to start this, and education is a really big one for me. I think there's a lot of really important issues and everybody's been so caught up talking about chat, G p t, understandably because it's accessible, it's a shiny object.

It seems magic I get it. It's. Okay let's accept that. Let's go now, let's say, what does that mean? And I think that's where hopefully with the conversation, we'll start moving as we get further into 2023. 

[00:33:37] Mahan Tavakoli: Paul, where, who, how, so I get together with my Paul Razer bot who knows everything that Paul knows and I say in Greater Washington we've.

Access to the business leaders. We've got access to the political leaders. We've got access to community and nonprofit leaders tell us, what should we do? Who should we bring to the table? And what kinds of conversations should we start with? 

[00:34:07] Paul Roetzer: It probably starts with, bringing together community round tables, some form of listening sessions.

You gotta get the power brokers in the room and make them. Like my effort at the moment is we've been spending so much time trying to educate like chief marketing officers and VPs of marketing, and that is fine to get to the tool and the piloting and scaling within marketing level. It is not gonna move the needle on workforce and how this stuff's gonna be integrated from a responsible way across organizations.

That needs to happen at the C e O level, the board level, and the investor level. So I'm making a conscious effort. Move upstream from an influence perspective to make sure we're having the conversations with the people who can actually do something about this. And I have never been in politics. I have no desire to be in politics.

But at the same time, you have to be willing to have conversations and start moving in circles where, things can get done. I'm not saying I have any interest in running for political office, I'm not doing any of that, but just being engaged with political leaders. So that they can help get valuable perspective and you can say, listen, can you please just have a conversation with these three people?

You have to understand their perspective of what's happening. And so we're trying to play more of a connector role and making sure that the right people are hearing from the people who have something to say that can positively impact the way this goes. 

[00:35:33] Mahan Tavakoli: And then being able to do that effectively.

We also need to be able to. Educate ourselves and tap into sources of knowledge to keep up with this fast pace of change in artificial intelligence. Paul. So I'm curious, what sources do you turn to for more signal and less noise because there is a lot of noise around this too.

I'm. Laugh at the number of companies that now say they have AI incorporated in the tool. And I've been very familiar with the company for years, if not decades, and I'm also surprised at the number of people that all of a sudden have easy hacks and tools and come up with all ways of making money through ai.

What do you turn to and where do you turn for more signal and less noise around? 

[00:36:27] Paul Roetzer: Actually, Twitter, believe it or not, is my primary source. I have a very highly curated list of people in ai. I actually have a public AI list, so if you just go to my Twitter profile and click on my list, there's an AI list.

It's probably about 350 or so people. So I have a curated version that I get alerts from, anytime they publish. And that to me is my main signals. It's okay, if they're talking about a topic, then that's important. I read a lot of AI research papers, believe it or not, they're highly technical and snooze inducing at times.

If you're really not into the, But those are a good indication of where things are going and I'm usually pretty intrigued by who the authors are and then following the career paths of those authors, cuz a number of the language model companies today are being run by people who.

Were the authors of the original attention is all you need paper from the Google Brain team in 2017. So if you followed where those people went in their careers and you just oh wow. Somebody from that paper went and started a company raised 50 million from Andresen Horowitz that's a company I'm gonna follow.

So you really just keep an eye on the emerging areas. And I do a lot of that. It's not. Easily repeatable process, I would say, but you can start to use AI in some of these ways. Like you can feed it a research paper and have it summarize it for you, things like that. So , I am starting to use some AI in different ways.

[00:37:48] Mahan Tavakoli: Along with that I also try to find people who are good, credible, curious sources, which is one of the reasons I've really appreciated finding. You are writing your book and your podcast, Paul, because I love your curiosity and your humility in asking the right questions. You are also able to communicate in a way that the non-technical audience can understand the applications and implications.

So I think finding sources like. Is of tremendous value. How can the audience take advantage of the opportunity to follow you, learn from your podcast and the great content you put out at the Marketing AI Institute? 

[00:38:36] Paul Roetzer: Connected with man, LinkedIn is a great way. Let me know. You heard me on this podcast.

I'll be happy to connect. I published once or twice a day on LinkedIn. And that's usually where my sounding board, like a lot of my early thoughts I'll put there and then sometimes they'll become the key topics in our weekly podcast. The marketing artificial intelligence which you referenced.

So the podcast and LinkedIn are the best near realtime insights into what we're thinking is important. And then our marketing ai institute.com site has the book, it has our conference, it has courses we teach and our newsletter, which comes out every Tuesday. So those are probably the main places to get, and then I am active on Twitter, but probably not as much as LinkedIn.

I'm sure most of my thoughts on LinkedIn. 

[00:39:18] Mahan Tavakoli: I appreciate that and we'll link to all of those in the show notes as well. And most especially, I agree with you Paul. I'm also very optimistic about the. Opportunities with ai, yet mindful that we do need to make the right choices as business leaders and as community leaders.

Really appreciate the responsible AI manifesto that you Yeah. Put out. And you also mentioned in there, in the process of making software more intelligent, AI has the potential to make brands more human by enabling us to focus increased time and energy on communications. Creativity, culture, community, and human connection.

AI should make us better people, professionals, and organizations. However, this will not happen without a continuous focus on the responsible application of AI across all business functions. Love those words. Love the work you are doing. Thank you so much for joining me in this conversation. 

Thanks 

[00:40:26] Paul Roetzer: for having me. Love the conversation. Thank you.