At Marketing AI Institute, we are seeing strong quantitative and qualitative evidence that indicates we are at an AI inflection point where the technology begins to go fully mainstream in 2023—with profound effects on…well…everything.
For years, we have hypothesized that AI has the power to change just about everything in marketing, business, the economy, and society at large.
In 2022, we saw that hypothesis confirmed thanks to an explosion of AI capabilities and technologies driven by the stunning power of generative AI tools like ChatGPT.
In a matter of months, even the brightest minds in AI have been surprised by how quickly and effectively AI tools have reshaped what we thought possible in marketing and business.
Episode 28 is dedicated to outlining why we think we’re about to enter a golden age of AI progress that will reshape the work of every professional in every industry.
Paul and Mike have a great conversation on this week’s episode, and you won’t want to miss the links referenced below—there is a lot of ground covered. Listen to the podcast below or in your favorite podcast player.
Timestamps
00:06:54 The AI inflection point
00:11:08 Evidence of the AI inflection point
00:20:07 How it will affect marketers, agencies, business leaders, and investors
00:36:08 What are big tech companies doing about it?
Links referenced in the show
- Introducing the Marketing Artificial Intelligence Institute (2016), Marketing AI Institute
- Art Isn’t Dead, It’s Just Machine-Generated, Andreessen Horowitz
- Marketing Artificial Intelligence: AI, Marketing, and the Future of Business, Marketing AI Institute
- A New Chat Bot Is a ‘Code Red’ for Google’s Search Business, New York Times
- Pathways Language Model (PaLM): Scaling to 540 Billion Parameters for Breakthrough Performance, Google
- Language modelling at scale: Gopher, ethical considerations, and retrieval, DeepMind
- Microsoft and Nvidia team up to train one of the world’s largest language models, VentureBeat
- Democratizing access to large-scale language models with OPT-175B, Meta
- ChatGPT crosses 1 million users in 5 days
- VCs try to parse through the 'noise' of generative AI, Pitchbook
Watch the Video
Read the Interview Transcription
Disclaimer: This transcription was written by AI, thanks to Descript, and has not been edited for content.
[00:00:00] Paul Roetzer: I don't know that anyone is going to remain the CMO of a major company if they're not talking about AI in 2023.
[00:00:07] 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.
[00:00:28] Paul Roetzer: My name is Paul Roetzer. I'm the founder of Marketing AI Institute, and I'm your host.
[00:00:36] Paul Roetzer: Welcome to episode 28 of the Marketing AI Show. I'm your host, Paul Roetzer, along with my co-host as always, Mike Kaput, chief content Officer and co at the Marketing Institute and co-author of our book, Marketing Artificial Intelligence: AI, Marketing, and the Future of Business, which is available in print, digital, and audio. Now, this is our first episode of 2023.
[00:01:00] Paul Roetzer: And the first time Mike and I were back in the office together in two weeks and we actually didn't have a plan for today's episode. And then we got talking and I was sharing with Mike a podcast episode. I'm listening to a Lex Fridman. I was right listening to On the Ride Into Work today, and it just like got us talking about big picture stuff.
[00:01:22] Paul Roetzer: So I'm going to turn it over to Mike in a moment. First, this episode is brought to you by our new piloting Ad for Marketers Online Course series is a series from Marketing Institute's Online Academy that we designed as a step-by-step learning journey for marketers and leaders. To guide them through adopting AI to advance their companies and careers.
[00:01:42] Paul Roetzer: The series includes 17 new on-demand courses that Mike and I recorded in December, 2022. So as everything was happening with ChatGPT and generative ai, so extremely timely and relevant, it actually worked out really well when we planned to release these. But there's dozens if not hundreds of AI use cases and technologies feature.
[00:02:02] Paul Roetzer: A collection of templates and frameworks to help you get started, really understand and apply ai. We've taken everything we've learned over the last decade or so of researching AI and hundreds of hours of planning and production and put it into this series to bring it to life. So, if you're at the point where you're trying to figure all this stuff out, and you want to get a headstart, especially in, in 2023, whenever you're whenever or whenever we are, you're listening to this, visit piloting ai.com to learn more about that series.
[00:02:29] Paul Roetzer: And you can use the Code AIPOD50 for $50 off your registration. So again, piloting ai.com if you want to jump in and take it. It's about eight hours of learning, so you can do it in a day. So if you're trying to figure out AI, lock in a Saturday and just like knock this thing out over the weekend and you're going to be ready to roll.
[00:02:46] Paul Roetzer: All right, Mike, I'm going to turn over, I guess a special edition in essence cuz we are not following the standard format on this.
[00:02:53] Mike Kaput: Yeah, that's exactly it, Paul. Happy New Year. I'm glad to be back. It's, it's going to be a big year as I think we're going to talk about a little bit during this episode. But like you mentioned, We're doing a special edition here.
[00:03:07] Mike Kaput: So typically the format of this is that we talk through, you know, three recent advancements in the news or items in the news related to artificial intelligence. And as anyone who's been following for the past few episodes knows, we kind of joke seriously that every week there. STUNNING new advances in AI a week in AI these days feels like years.
[00:03:32] Mike Kaput: There are so many different developments. There are so many different releases and transformations happening in the world of ai, and with all the past episodes we've done recently, we've. Almost been kind of bombarded by a new transformative development almost every week. And so what we want to do today is actually take a step back and connect the dots a bit for the audience, of the institute and of the podcast, to really show people in aggregate what these types of developments mean.
[00:04:08] Mike Kaput: So, The basic thesis is this, based on a number of things that we're going to discuss in a second here at Marketing AI Institute, we are seeing some really strong quantitative and qualitative signals that indicate we're at what we're calling an AI inflection point. And by that we mean an inflection point where AI in 2023 we hypothesize is going to go mainstream based on these signals and change in the process everything we take to be true about business and careers in every industry.
[00:04:46] Mike Kaput: Now, to kind of tee this up, before I turn it over to you, Paul, back in 2016 when Marketing AI Institute was founded, Kind of put out this idea that AI has the power to change everything. In fact, you wrote in the intro blog post for the institute that, you
[00:05:05] Paul Roetzer: know, the day it was you, you went into the archives here,
[00:05:09] Mike Kaput: the day.
[00:05:09] Mike Kaput: Exactly. And we'll link to that in the show notes to see if people, appreciate our prediction. Yeah. But you wrote. Artificial intelligence possesses the power to change everything. That includes not only marketing, that includes business, the economy, and society at large. And that was back in 2016 and kind of getting in the time capsule here in 2016.
[00:05:34] Mike Kaput: We saw the promise of ai, but it was still fairly limited in its commercial application. You know, deep learning was an important technology that existed and was very promising, but only in research labs. Transformers, you know, a core AI technology that led us to things like G P T G P T two and three that was still under development and consumer AI applications.
[00:06:01] Mike Kaput: While there were plenty of them that were viable and impressive, they were still relatively limited and a lot of them were baked into big tech platforms. Think Accessi, Google Maps, or limited to very big firms that had the. To build these solutions now, we saw really strong momentum and trends of AI investment and development over the subsequent years.
[00:06:24] Mike Kaput: But I think we both agreed that everything fundamentally changed in 2022. And there's a few reasons for this. Since about mid 2022, we've seen this explosion of AI capabilities and technologies in an almost exponential curve. Driven largely by generative ai and in the process generative AI and related technologies have upended what everyone, including some of the top minds in AI, have thought were possible.
[00:06:54] Mike Kaput: So I'm just going to touch on a few very quick points here of why we're at this inflection point. Why we think we might be really on the cusp of a golden age of artificial intelligence development. And then I'm going to turn over to you cuz I want to get your thoughts on this and have you kind of round out why we might really be at a tipping point here.
[00:07:16] Mike Kaput: And so in 2020, Starting just in April, we see the release of Dali two, so we've talked about that at length. Dali two allows you to generate images using text prompts. Now, Dali two got a million users in two and a half months, so that blue G P T three out of the water, it took G P T 3 24 months to reach the same milestone.
[00:07:41] Mike Kaput: In August, 2022, just a few months later, the open source model stable diffusion comes out That also generates images from text. Now open AI's technology kind of had somewhat of a restricted a P I, but stable diffusion allowed anyone to build on top of it. And according to Andreessen Horowitz, it actually had the fastest up.
[00:08:04] Mike Kaput: Of any technology ever by developers that they have seen based on GitHub usage benchmarks. At that time, you're reading a lot of articles about ai, art generation, image generation, copyright, how art is being disrupted, stock photos are being disrupted. That was all well and good and really interesting, but a lot of people missed that.
[00:08:26] Mike Kaput: This wasn't a story about AI generating art. It was a story about AI understanding natural language well enough to turn text into anything you can imagine. And investors, I would say really saw the writing on the wall here, at least the savviest of them. You saw companies like Jasper raise 125 million series a round valued at 1.5 billion in order to generate writing and copy.
[00:08:53] Mike Kaput: You saw runway ml, which has content and image editing tools powered by. Valued at 500 million, raising 50 million, you had stability. Ai, which is powered by stable diffusion, valued at over a billion dollars with 101 million in funding. So there was this wave of funding for generative ai. And you know, according to PitchBook, in 2022 alone, investors bet at.
[00:09:22] Mike Kaput: Over a billion dollars on this crop of VC back companies. This was across 78 different D deals, and that was at a time when I think we were seeing in the rest of the startup world, funding drop off a cliff. Now this all culminates. On November 30th, 2022, and that's when ChatGPT comes out and rocks the world.
[00:09:44] Mike Kaput: And in just five days ChatGPT gets 1 million users, which blows DALL-E two's record outta the water since that day. It's just been 34 days to the recording of this podcast. Since ChatGPT. Launched. We've seen tons of different tools. They use ChatGPT like interfaces to honestly kind of work wonders.
[00:10:06] Mike Kaput: Things like telling AI what you want to code and it generates code for you. There's tools that automatically build pitch decks based on text prompts. There's projects where artists use ChatGPT prompts an image generation to create professional looking characters from scratch in minutes. I mean, we have tools now that.
[00:10:27] Mike Kaput: Videos, narratives, scripts, plans, checklists, even code from simple text prompts. And really where we've gotten right now is this crazy tipping point where what we've hypothesized in the past is actually coming true. And so we really wanted to devote today's episode to outline why we think we are about to enter a brave new world where AI is going to reshape the work of every profess.
[00:10:56] Mike Kaput: in every industry. So I wanted to kick it off, Paul, and just ask kind of, does that track to you with what you've seen in the market, why do you think we're at an AI inflection
[00:11:06] Paul Roetzer: point today? Yeah, I mean there's so much to unpack in, in that opening and it's funny to think back to 2016, you know, when I, we launched the institute and for people aren't familiar like we.
[00:11:20] Paul Roetzer: Researching ai I personally did in 2011. So, I mean, my efforts to understand this space came just basically with a, a, a theory that it would eventually be applied to marketing and sales and we could maybe build an intelligence engine one day that would, predict strategies and, and things like that.
[00:11:37] Paul Roetzer: And that was how I just started researching ai. And then fast forward, you know, 2014 we wrote a little bit about it in my second book, and then 2016 we launched the blog for the institute. So, I mean, we're going back 11 years that we've been sort of researching and writing and theorizing about what was going to happen.
[00:11:54] Paul Roetzer: And I feel like more progress was made in the four weeks to end 2022 than we had made in the, the 10 years prior. Like, it, it was just one of those things where, you know, I wrote. Marketer's first experience with ChatGPT or whatever it was on like December 2nd, I think like, so two days after it came out, and I feel like there was just this crazy tipping point where all of a sudden everyone was talking about generative AI ChatGPT, friends of mine, who I hadn't heard from in years, who knew we were maybe doing some stuff with ai.
[00:12:30] Paul Roetzer: Like all of a sudden we're reaching out, asking questions. , every conversation I was having with people in business and people in the AI community was centering around these same feelings. And so it was just like you got this qualitative feeling that things were happening. And what you and I had talked about is like, you know, what's the quantitative story here?
[00:12:54] Paul Roetzer: Like, is it really happening? And I think ChatGPT is such a tangible. Thing to look at because you can just look at pure numbers, like, oh, okay, yes, we know it was a million after five days. I don't know what it is today. I don't know that they've reached, released any data, but I would imagine it's probably in the five to 10 million user range for ChatGPT, if not more.
[00:13:13] Paul Roetzer: I, I don't know. But that's a quantitative piece. You're not going to see the follow on like investment data for a little while. We're not going to see. I mean, you can look at Google trend data and look at search volume. Like there's some things we can look at that tell us it's happening, but so much for me has really just been.
[00:13:32] Paul Roetzer: Again, these kind of conversations that are happening, the outreach on LinkedIn, I'm getting the amount of people that I follow in the marketing space who are now talking about AI for the first time in their careers, which I think, by the way, is, is wonderful. Like this is the moment I think we've been waiting for is for everyone else.
[00:13:51] Paul Roetzer: To, to kind of join the party. So like, say, yeah, like this is going to be huge because on our own, we weren't going to get there. Like just us talking about this was never going to be enough, even after five plus years of doing it. And that's why, you know, I think in what late 2021, we introduced that 1 million Marketer Challenge and the whole premise was like, we need at least a million marketers to understand AI before we can have a major movement created and really transform the industry.
[00:14:20] Paul Roetzer: And that was the premise back then is like, we can't do this on our own. And we needed more companies talked about it, we needed more influencers talking about it. And so that's what we're starting to see is so many of my author friends, speaker friends who never done anything on ai, are having points of view on it, which I, I think is wonderful.
[00:14:38] Paul Roetzer: And I, I hope that they keep going on on that. And then I think the same with marketing leaders. Like, you know, I saw a C M O yesterday, someone tagged me on LinkedIn on a C M O of a, a big company. And, he was saying like, all right, I'm in. Like I want to put my 10,000 hours and I want to learn about this stuff.
[00:14:57] Paul Roetzer: And that's what we've been missing. Like, I think we. We looked around years ago, I guess, probably about a year, year and a half ago when we were doing the AI for CMOs report. We couldn't find the CMOs talking about it. There was nobody, it was like four out of the top 50 CMOs in the world had ever mentioned AI online that we could find, like just Google searching for it.
[00:15:15] Paul Roetzer: So I, I think that's going to change. I, I don't know, that anyone is going to remain the CMO of a major company if they're not talking about AI in 2023 . I think there's just been a lot of that, a lot of conversations. And then even just like again, over the weekend, so December 31st, I'm looking at my Twitter feed.
[00:15:34] Paul Roetzer: Three, three retweets I did. The first was Greg Brockman, who is a co-founder of Open ai. Greater DALL-E and G P T three Prediction 2023 will make 2022 look like a sleepy year for AI advancement and adoption, followed by Andrew, who is the co-founder of Coursera, a Stanford profess. Head was head of Badu AI group and co-founder or co-lead of Google Brain, and one of like the foremost AI experts in the world.
[00:16:05] Paul Roetzer: His, his machine learning course, which I took on Coursera, I think that millions of people take it. He said 2022 has been a great year for AI progress. Excitingly, this has laid a foundation for even more progress in 2023. Also same day, December 31st. Mustafa Solomon, who's one of the co-founders of DeepMind, if I'm not mistaken.
[00:16:25] Paul Roetzer: Yeah. I'm one of the co-founders of DeepMind. 2022 has been the most productive year for applied AI ever. 2023 is going to make it look like it was a no big deal. . So again, it, this isn't just us, like the people who are actually building this stuff, who know what's coming. Like that to me is like the big prelude.
[00:16:42] Paul Roetzer: Oh, there was one other one I put on LinkedIn. , which was Sam Altman. I don't think I have that one like right in front of me. But Sam, said basically like the same thing. Like this is going to be massive. It's coming. So everyone who knows what they're talking about, who are in the research labs who are building this stuff and seeing what's coming in 3, 6, 12 months, they are tipping their hand on Twitter saying like, you've seen nothing yet.
[00:17:08] Paul Roetzer: And that's, that's jiving with everything I'm hearing, from people that we talk to who are building these things, is that no one's ready for what's actually going to happen. And I think like, we'll get into like the impact on marketers and agencies and leaders and stuff. But the thing I started thinking a lot about over, holiday breaks, I actually took a little time, time away and like, you know, try to process what was going on.
[00:17:30] Paul Roetzer: I was really thinking deeply about. What, what do Google and Microsoft and Amazon and Nvidia and Apple and meta like, what are they going to do? Hmm. And I, I know we're going to kind of get into that as like the, the really the meat of today's conversation, but that to me is the great unknown. Like maybe the trillion dollar question for 2023 is where do we go from here?
[00:17:54] Paul Roetzer: But, you know, I think first. Maybe we, we spend a few moments talking about like what does this inflection point actually mean to different people, different professions?
[00:18:05] Mike Kaput: Yeah, exactly. And I wanted to, there's a few groups I specifically wanted to ask you about. I mean, assuming the inflection point exists, I mean, and I think it's a pretty solid, it exists
[00:18:17] Paul Roetzer: projection, , I mean, there's very few things in my career where I.
[00:18:21] Paul Roetzer: I, I got the inbound marketing movement, right, like back in 2008. I bet my entire agency's future on Brian Halligan, Dharma, Shaw, HubSpot, and the inbound marketing movement. That was a good bet. I was, I was very confident. I was very young. I mean I was 29 when we made that bet. , I guess like I started 2005.
[00:18:38] Paul Roetzer: Yeah, 29. When we kind of bet the future on like inbound marketing and I would say my confidence level at that time. was probably in like the 80 to 90% range. I was making the right call. Hmm. I, I would say that right now I'm, I'm, I mean, I'm like 99.9% confident that, that we have reached the inflection point we've been waiting on.
[00:19:00] Paul Roetzer: We can talk a little bit later on, like that doesn't mean it's going to be perfect. It doesn't mean there aren't going to be setbacks. It doesn't mean there won't be unrealistic expectations applied and therefore the dissenters, were going to say, well, it didn't do what it was supposed to. That's the history of AI is people doubting the long-term potential because they're only looking at the short term.
[00:19:20] Paul Roetzer: And that's actually one of the things we can touch on as we think about marketers and agents Don't get caught looking at the shiny object, Chad, g p t, and do, they're the shiny objects. That is not the future though. It's just a tool. And so we need to do as marketers is we have to, we have to think about the bigger picture and that's why we're going to talk, you know, in a minute about these bigger companies and what's going.
[00:19:44] Mike Kaput: And before we get there, so talk to me a little bit, and you've, you've actually published quite a bit about this last few days, few weeks. What's the inflection points impact on both? Talk to me about like a brand marketer, individual marketers, and marketing agencies. I think there's going to be a pre profound shift here.
[00:20:01] Mike Kaput: I mean, we've been talking about this for a long time, but now it's here. If I'm a marketer, what do I need to be thinking about?
[00:20:07] Paul Roetzer: For marketers, it's going to start obviously to me in the creative realm, which I didn't think, like if you had to asked me two years ago, where would it start? I thought it was going to be analytics, personalization, things like that, you know, would be where, where most of the adoption happen early.
[00:20:22] Paul Roetzer: But those, those use cases and technologies aren't as accessible. They're not as like, go get the tool right now, and by the afternoon you're creating outputs with it. You're doing something of value. That's why I think the creative stuff, the generative AI movement changed everything is any marketer can now go get an image generation tool.
[00:20:44] Paul Roetzer: Mm-hmm. They can go get a writing tool and they can use it. So I think if you're a marketer and you specifically work in any creative realm, could be design, could be content creation, could be video creation, anywhere in there script writing, like anything. Your impact is immediate. Like you should be thinking right now about how to use AI in your existing workflows, cuz the tools are cheap.
[00:21:12] Paul Roetzer: You can go get anything for under a hundred bucks. Some of 'em aren't 30 bucks. Like you can get a lot of these tools, you can experiment with them, you can create outputs, you can improve the outputs like you're not replacing. So I think for marketers, the immediate thing in 2023 is find ways to infuse these generative AI tools into your existing workflows and to train your team how to use them so that they can become more efficient and better at their jobs.
[00:21:40] Paul Roetzer: So like I'm not talking about like let's transform entire departments and teams and like have to like spend all this time investing in AI education training. Like that should be happening in the back. But don't wait for that to do it. If you look at your team and say, wow, we're spending 400 hours a month creating content, whether it's demand gen content, or top of the funnel content, or podcast or webinar, whatever, how can we use AI right now to do that more efficiently and to do it better?
[00:22:08] Paul Roetzer: So for marketers, I would say it's a use case based thing. Find the use cases where you're already doing a task and find a way to do it better, smarter, cheap. . Yeah. And I
[00:22:20] Mike Kaput: think what's really interesting with, we've talked about our use case model for a long time, of doing just that, is really listing out kind of everything you do in a day and figuring out where could AI make your workflow more.
[00:22:34] Mike Kaput: More intelligent and something like ChatGPT as one example allows you to test that right now. I mean, anyone listening to this could stop the podcast, go write down a quick list of stuff they did today and then start playing around with Ave a free immediately accessible tool like ChatGPT to see, hey, where are areas that this actually could maybe help out or augment what I'm already doing.
[00:22:57] Mike Kaput: Now, how about marketing agencies? I know we just published something today about, with your comments on what agencies need to think about. Because when I look at some of this stuff and having worked with you at your agency, I start to wonder if we don't need to start rethinking agencies from the ground up based on the power of some of this technology.
[00:23:21] Paul Roetzer: Okay, so yes to context here for people who don't know me, I ran an agency for 16 years. We were HubSpot's first partner back in 2007, and I wrote my first book, the Marketing Agency Blueprint. I've spent my professional life thinking about and building agencies , so it is. It's never far from my mind of what does this mean for agencies.
[00:23:43] Paul Roetzer: I have a lot of friends who still run agencies. I still, you know, advise my agency. I'm not involved day to day anymore. But I think deeply about the impact on agencies and I, I, I, I think they're in trouble like and this is probably why there's a few professions and a few. Areas that I think are going to get impacted really fast.
[00:24:09] Paul Roetzer: Writing is another one. Like writers, we, we can talk about that. Mm-hmm. . But agencies, they're, so, a lot of agencies still bill hourly rates, like they're still billing some form of. Work per hour. Many haven't shifted like a value-based model. And so if you just think about like, let's say your agency creates content or you pay an agency to create content, or you're a freelance writer who contracts with agencies, whatever, it's a big ecosystem and.
[00:24:39] Paul Roetzer: You're getting paid to create content today. And let's say like the example I used in, in the, the blog post was, let's say you're writing a thousand word blog post for a client and you're, you know, on average to write it, to edit it, to publish it, to put images in whatever it takes about seven hours, just roughly speaking.
[00:24:57] Paul Roetzer: So it's about $1,050 in revenue you're going to generate. If you're charging by the hour and your client pays you by the hour, you can make the thousand dollars. So if you infuse ChatGPT and any or, and or any of the dozens of AI writing tools that exist, like writer, Jasper, hyper write, whatever they may be, you infuse these writing tools.
[00:25:18] Paul Roetzer: There's no way you can't create that same piece of content out the door with AI generated images in under an. And I, I am not saying the r the AI replaces the right. I'm saying like I've, I've written posts with it. Like I know how quickly it doesn't and how good it is and how much time the human in the loop part takes.
[00:25:39] Paul Roetzer: So I'm not saying take the AI generate content and go start publishing a bunch of stuff, but you and I have both done this. We've done it for outlines of eBooks. I did it. Formats for our conference, like a sample agendas for the conference, business plans. I like, I've created all kinds of content using these G P T tools and they're really good.
[00:26:00] Paul Roetzer: And so if you're an agency and you are reliant on billable hours for any kind of generative content, again, whether it's visuals or video or whatever, the time it's going to take you to do your work is going to be sliced dramat. You know, maybe, maybe 10 times less than what you would usually spend, like a massive cut.
[00:26:20] Paul Roetzer: So if you're still charging by the hour, what are you going to do? Like you, you're going to charge one hour for what used to take you seven. Well, now you don't need your staff anymore cuz now you're able to . So I think there's a, there's a fundamental problem if, if you are an agency that is charging an hourly rate to create anything, any kind of output, you cannot do that anymore.
[00:26:41] Paul Roetzer: Like it's. And all it takes is brands to realize that, because let's say I'm paying an agency to do that. Now I'm, I'm going to put my brand hat on. Like I run a, the institute, let's say we were paying an outside agency to create content for us, and I knew what Jet ChatGPT and G P T tools were able to do, and they were charging me $500, a thousand dollars, whatever, to create blog posts for us.
[00:27:05] Paul Roetzer: There's no way in hell I would continue to pay that, like mm-hmm. , I'd be like, I'll just hire my own market. Like I can have someone on our team create content five to 10 times the output you're creating as an agency for less money. So if you're an agency that creates content for a living, your financial model is obsolete.
[00:27:27] Paul Roetzer: Done. Like, and even if it's a value-based model, it's still probably obsoleted because the, the, the value of content. So even on a value-based model, you have to rethink it so that the, the three things I call was a billing model, which I think I've hammered enough. Creativity. So the things you can now create, so like the, this is the opportunity realm.
[00:27:48] Paul Roetzer: What can you create now for clients that you couldn't before? What kind of imagery, videos, like language, what can you generate now that you couldn't? And then the last thing I touched on was innovation. Like what are the products and services you should be offer? That either doesn't exist right now in your firm or maybe doesn't exist, period, because you're at the front lines like you're working with these clients, you know the pain points better than most, you know, the opportunities, like what can you build now that you couldn't have built before?
[00:28:19] Paul Roetzer: And so that to me is like the, again, I could talk for agencies for the entire episode and maybe we're going to do that in a future episode, . But I think those are the key three key things. If I was running an h I would be, I would, looking at it right now. Mm.
[00:28:32] Mike Kaput: As a, yeah, starting today, you need to solve this.
[00:28:37] Mike Kaput: So zooming out a bit from marketers and agencies and, let's talk about maybe the people that would be hiring marketers and agencies. So business leaders, let's say non either marketing leaders or non-marketing business leaders, or people that own a business or help run it. What does this
[00:28:56] Paul Roetzer: mean for them?
[00:28:58] Paul Roetzer: You have to educate immediate. So like, again, the way I would look at it within the industry, we have five people and all of us are aware of ai and all of us are researching different tools based on what our role in the, the firm is. I would, I would look at the major areas where your team invest time and I would find someone who can own looking for smarter solutions within that space.
[00:29:22] Paul Roetzer: So one, I would prioritize education as we've, if anybody listened to our like state of ai. You'll know that, there isn't much, there is not much AI education happening within organizations, so you have to find content. Now again, you can go get the piloting AI for Marketers course and just infuse that in, like, just have everybody take that and the whole team will be up to speed in a day.
[00:29:44] Paul Roetzer: But you have to find a way to teach people what AI is, how to use it, how to find use cases, how to look at problems differently. It's kinda like the basic things. And then over time you may actually start evolving your hiring practices. But the immediate step is you cannot wait till the end of 2023 to take action.
[00:30:04] Paul Roetzer: So I would say you want to prioritize who is going to own researching solutions within a space to find smarter tech. Cause again, AI's just smarter technology, to do what we're doing more efficiently and, and better. So I would start there, the use case side. And then on the problem-based side, this is what we teach in the piloting AI Marketers course series.
[00:30:24] Paul Roetzer: You have to look at all the different things, all these high value problems that you might be able to solve more intelligently. So the am example I use, I think in the course was, our conference I, I go into great, like inside the institute, kind of great detail around the challenges of running, in-person conference, right?
[00:30:40] Paul Roetzer: And how do we solve it smarter than maybe we've tried to solve it in the past, but there's all kinds of business challenges around demand generation, customer experience, churn, you know, profitability, whatever your challenges are as a leader, you're looking at what else can we be doing to solve these problems smarter than we've previously done?
[00:31:01] Mike Kaput: I think it also might be worth reiterating. Your unique solution to these problems will, will. It'll look different for everyone, but we're not talking about hiring a data scientist or a machine learning engineer or finding one within your organization, though that's helpful. In certain contexts, we're talking about taking the people that already have the domain expertise and having them learn and understand what is possible with artificial intelligence, which anyone can do regardless of technical or non-technical Back.
[00:31:33] Mike Kaput: What about investors like anyone from VCs to p to people with portfolio companies, to even angel investors that are looking to invest in the next generation of companies or have invested in them already? I mean, what did they need to be thinking about?
[00:31:51] Paul Roetzer: So again, context like I'm an investor in an angel fund that invests in startups, mostly software companies.
[00:31:58] Paul Roetzer: I have friends who run venture funds. I talk with Venture Studio people like pretty tapped into the space in general. And what I've been telling my friends who are angel investors or who run funds, whatever it is, is one, do not invest in a SaaS company that doesn't have an AI roadmap. Like if, if this, if you're investing in.
[00:32:20] Paul Roetzer: And they have no idea what they're doing with ai. Get someone on their team who does, or a board member who does whatever it is, like someone has to lead the immediate development of an AI roadmap for the product and platform itself. How we're going to make the software smarter and the company as a whole.
[00:32:38] Paul Roetzer: That I think is obvious. Like I, I hope that's obvious to a lot of people right now, or it's becoming quickly very obvious to people, but the, the blind spot I'm seeing. Like you're seeing all this VC funding dry up. You can look at any pitch book, you know, CB Insights? Mm-hmm. , where wherever you're going to go, the, the funding is like, like you said, fell off the table in 2022.
[00:32:57] Paul Roetzer: So like the funding stopped. However, what I don't know is happening enough is the audit of the existing portfolio. So if you're an angel investor or if you're a fund manager, whatever it is, and you have 2, 5 50, a hundred SaaS companies in your portfolio, Same deal. I would do an audit right now that says, who on your team understands ai?
[00:33:20] Paul Roetzer: Hopefully it's the c e O. It's probably not going to be, but you want someone in every portfolio company that deeply understands the impact AI is about to have on their business, because that software can be obsoleted overnight by someone who's just playing around. On the weekend building stuff. I mean, we're seeing it.
[00:33:37] Paul Roetzer: We like you listed some of these like crazy tools that are popping up and so like if I was an individual like point solution software company, I would be terrified right now and I'd have to be moving really quick. If I was a platform software company, c r M, automation, marketing, sales, whatever it is. I would be like Google style code red.
[00:33:56] Paul Roetzer: Like what in our platform is going to be obsoleted by someone building a smarter version of that feature, that tool, or the entire platform. So if I'm an investor, I want answers right now as to who in each portfolio, company or future investment understands AI and can build a roadmap for how it's going to be infused into that company.
[00:34:16] Paul Roetzer: Or I'm not making that investment, or I'm certainly not doing a follow on investment. So, That's again, like I have pretty strong feelings about this one, kind of like I do with agencies. And I don't think enough investors are thinking that way. Cause I don't think enough investors understand AI enough yet.
[00:34:34] Paul Roetzer: Like it's like everybody, business leaders, marketers, agencies, investors. Very few people deeply understand AI and the impact it's going to have. And that leads to the shiny object thing. Mm-hmm. , they think ChatGPT, it's like all a sudden in the last four weeks, everyone's talking about ai. They're talking about ChatGPT.
[00:34:52] Paul Roetzer: They're not even talking about the underlying like structure architecture that enables chat. G P T, they're not talking about the other innovations that are happening in these major C. That are going to be bigger than ChatGPT because all they know is the shiny object. And that to me is like where people could get themselves in trouble.
[00:35:11] Paul Roetzer: If they, if they stop at just trying to solve for ChatGPT, they're going to miss the much bigger threat and opportunity.
[00:35:18] Mike Kaput: Yeah, and I think that's one of the motivations behind today's episode is with all this kind of blistering pace of innovation and releases, we wanted to sit back and say It's time to wake up.
[00:35:30] Mike Kaput: I don't think it's too strong of language to say this is an alarm bell for both threats and opportunities like today needs to be the day whenever you're listening to this that you start paying attention because it is here and it is going to be a crazy year. Now you referenced before a trillion dollar question that we probably won't solve just on this episode, but has to be on everyone's minds.
[00:35:56] Mike Kaput: I mean, we've talked about companies like OpenAI, which are extremely well funded and have deep connections to the biggest companies in Silicon Valley, and they're pushing this innovation forward. But where are the big tech companies on this? I mean, in the book that we published this year, we talked about companies like Google, Amazon, Microsoft.
[00:36:17] Mike Kaput: Obviously Meta is another huge player. Maybe walk us through how you see big tech leaders responding to these types of
[00:36:27] Paul Roetzer: technologies. So I, I think it, it's really helpful to understand. At a high level what's happening or what happened And, and so what happened in 2022 is open AI stability, ai, ai, runway, ml, hugging face.
[00:36:43] Paul Roetzer: You had these, these kind of independent labs and companies that aren't bound by these massive barriers to release tech that the other big companies are. And they started releasing things the big tech companies would never have released, like couldn't, couldn't release. And so I'll, I'll actually read an excerpt from something cause I think.
[00:37:02] Paul Roetzer: Bri's home this point very clearly. So, DeepMind you, if you've listened to the podcast before you hear us talk about DeepMind quite a bit, dabba to me is like probably going to end up being the most important person of our generation, you know, co-founder of DeepMind and the current c e o. So this is from, I, I shouldn't even tell you the date because it'll sound like it's, it's coming from that.
[00:37:21] Paul Roetzer: But I'm going to tell you anyway, December 8th, 2021. So a little over a year ago. A year and two. The, the article was titled, language Modeling at Scale, gopher, ethical Considerations and Retrieval. So now this gets slightly dense, but bear with me cuz it's worth listening to an unpacking. This is, this is how the post starts language and it's, oh real re So, DeepMind built, AlphaGo built alpha fold, build all these insane advancements in deep learning like.
[00:37:52] Paul Roetzer: They're on par with open ai, probably well beyond open AI in terms of their impact, but they got acquired by Google in 2014 for like 670 million. So they are a research lab that basically functions independently within Google. Okay. All right. Language and its role in demonstrating and facilitating comprehension or intelligence is a fundamental part of being human.
[00:38:13] Paul Roetzer: It gives people the ability to communicate thoughts, concepts, express ideas, create memories, and build mutual under. These are foundational parts of social intelligence. It's why our teams at DeepMind study aspects of language processing and communication, both in artificial agents and in humans. As part of a broader portfolio of AI research, we believe the development and study of more powerful language models, systems that predict and generate text have tremendous potential for building advanced AI systems that can be used safely and efficiently.
[00:38:48] Paul Roetzer: Summarize information, provide expert advice and follow instructions. Bio language, developing beneficial models requires research in their potential impacts, including the risk they pro risks they pose. This includes collaboration between experts from varied backgrounds, thoughtful anticipate, address challenges that training algorithms on existing data sets can create again, December of 2020.
[00:39:12] Paul Roetzer: Today we are releasing three papers on language models. What I'm trying to get to here is Google is working on this stuff. Meta is working on this stuff. Apple is working on this stuff. A Amazon. Just because OpenAI is releasing stuff doesn't mean it doesn't also live within the big companies. So I'll come back in a minute.
[00:39:30] Paul Roetzer: So today we are releasing three papers on language models. They include a detailed study of a 280 billion parameter transformer language model called Go. I'll explain data in a second. A study of ethical and social risks associated with large language models and a paper investigating a new architecture with better training efficiency, real quick parameters, so it's a 280 billion parameter transformer language model.
[00:39:59] Paul Roetzer: Transformer is the underlying architecture, that allows these models to be built, these language models to be built. It's what G P T three is built on. Generative, pre-trained transformer is what G P T stands for. 280 billion is a lot. I think G P T three is 175 billion. The more parameters historically the belief, the more powerful the language model.
[00:40:21] Paul Roetzer: Second. Okay. Back to the. As well as quantitative evaluation of Gopher, we also explored the model through direct interaction. Now again, think about ChatGPT as I read this next line among our key findings was that when Gopher is prompted towards a dialogue interaction, parentheses like a chat, the model can sometimes provide surprising coherence.
[00:40:44] Paul Roetzer: Sound familiar? The shock people have at ChatGPT Go read the article. They, they show a chat occurring with Gopher. So goo deep. Research lab within Google in 2021, publishing something larger than CH than G B T three saying, Hey, we have this tech two, but hold on a second, we're not going to release it.
[00:41:05] Paul Roetzer: However, our research also detailed several failure models that persist across model sizes. Among them, a tendency for repetition. Now ChatGPT's gotten pretty good at. But it, it fails once it gets pasted, like 800 words. Like it, it still has limitations. So, persist across among them Tennessee of repetition, the reflection of stereotypical biases and the confident propagation of incorrect information.
[00:41:35] Paul Roetzer: What is every marketing influencer who's written about chat? G p d said, it's confident. It has an ego, basically, like it's, it believes everything it says. Open AI didn't invent this tech, the they ha. Google has it, DeepMind has it, Microsoft has it, meta has it, everybody has this stuff. Now, is it better than ChatGPT?
[00:41:59] Paul Roetzer: I don't know. I'm not in the labs, but my belief is yes, that open AI doesn't actually have anything unique. And that to me is the biggest challenge with where this all goes is what if, what OpenAI has isn't even state of the. What if ChatGPT is not the state of the art chat model, language model? I, I wouldn't say with the same confidence I do as an inflection point in 2023, but I would be fairly confident that what we're seeing right now, what people are playing with the shiny object right now is not even state of the art.
[00:42:36] Paul Roetzer: Now, OpenAI also has a more powerful version, like, again, my belief and what I've kind of been told from within the research labs. Historically, what we're experiencing as consumers, as professionals is usually like 12 months behind what lives in the research labs. So ChatGPT. Today, we're actually using tech they have had since 12 months prior.
[00:43:01] Paul Roetzer: Hmm. Now the other thing I'm hearing is that number is moving very quickly. So now it's like, because the release cycles are happening so quickly, it's like six months. So like we're just seeing ChatGPT today and they maybe have something that's like six months ahead of it, living at opening eyes, research labs.
[00:43:17] Paul Roetzer: But the trillion dollar question again is what does Microsoft have, because we know they did a deal within Nvidia, they announced, October of 20. Microsoft and Nvidia team up to train one of the world's largest language models, which they called Megatron Turn. I love that name. Anything with Megatron is quite anything I should just transformed in it.
[00:43:36] Paul Roetzer: So they had a, a 530 billion parameter language model. Again, October of 21, meta in two May of 2022. Democratizing access to large scale language models with O P T 175. So in line with Meta's AI commitment to open science, we are sharing our open pre-trained transformer 175 billion parameter Apple, who doesn't talk about anything they're doing completely secretive.
[00:44:05] Paul Roetzer: I look at it, it's like, well, shit, what if Siri's actually good? Like what if Apple all along has been building their own language model? It doesn't even have to be better than else. But what if Siri works? What if Google Assistant works? What if Alexa actually becomes something more? Asking the weather and the time, which, sorry, that's all I've ever used it for.
[00:44:25] Paul Roetzer: So that was like the thing I spent break thinking about. And, and again, like we don't, we don't have answers to this, I have no answers, but I found myself saying like, okay, what if Q1 of 2023, these big companies that historically wouldn't release things because of their concerns around misinformation, failure, repetition, whatever their concerns were, ethical bias what.
[00:44:47] Paul Roetzer: What if they have to? Like what if? What if open AI and stability, AI and hugging, what if they just forced their hand where all these big companies say, screw it. We gotta go. What is the safest version that we can possibly release of this? And I start to think like, well, well what if that then lives in the things we already use?
[00:45:06] Paul Roetzer: Like what if Google Docs and Google Sheet. Truly has this stuff infused in it. Not some, you know, cute machine learning stuff and some basic things like summarization of my doc and Google Docs is nice, like fine. But what if Microsoft Word and Office 360 and Google all of a sudden have the most powerful language model technology baked right into them?
[00:45:28] Paul Roetzer: What does that do to the third party AI writing tool marketplace? What does it do to the assistant marketplace? What does it do to all this like ecosystem? And that's where I'm like, oh man. I don't know, like, and I started really sitting back thinking like, what if meta breaks off the metaverse? Like what if, you know, and, and Callen, AIS, Jason Kakkis, who you and I both listened to, he and Molly Wood, is it, is that his co-host on Yeah.
[00:45:53] Paul Roetzer: This weekend. Startups? Yeah. Yeah. Awesome. And they were saying like, well, who, what c e o might leave this year? And they had theorized was like, well, what if Zuckerberg leaves and just takes the metaverse separately, you know, becomes its own, you know, subsidiary maybe of the larger company. , what if Meta, which has a massive research lab run by Jan Lacoon, like Meta does stuff in ai most people have no clue about.
[00:46:13] Paul Roetzer: Mm-hmm. , like you don't think of meta as a major AI company unless you're in the space, but they are like probably rival any other research lab in the industry. So, quick, quick context, Mike, and then I'll stop talking and let you, so one of the things we thought about, I, I came up with this concept like sometime in 2022, I think, and I don't think I've shared anything about it publicly, but it seemed like a good time.
[00:46:33] Paul Roetzer: So I started analyzing. SaaS companies, like tech companies based on how many AI ML staff they have, and then looking at the percentage of that staff to their total employee base, what I was calling their AI talent index. And so just for perspective on the investment these companies are making in AI ml, so ML being machine learning, if not familiar.
[00:46:54] Paul Roetzer: So I just took Sales navigator data, did a keyword search for title. And look for anyone that had artificial intelligence, ai, ml, or machine learning in their title. This is obviously not an all-inclusive thing. People may have data in their title instead, whatever it is, but it gives a pretty good representation.
[00:47:09] Paul Roetzer: Microsoft 4,000 plus employees with AI ML in the title. Hmm, Amazon 3,500, apple 3000, ibm, 3000. Google 2000 Meta 1500, Intel 1500, Nvidia 621, LinkedIn five eighty one, Oracle 400, Adobe three seventy six. Salesforce one eighty one. HubSpot 15, drift 13. Just for context, like two other companies we're really familiar with 4,000 at Microsoft and a billion dollar exclusive deal with open.
[00:47:43] Paul Roetzer: I started thinking like I'm ready to move some of my stock . This is not stock advice. I'm not providing any kind of like public guidance, but what I've always tried to make the bets on who's going to win an ai and to me that was like the future of my investments. Like that's, that's where I put my money, which is like betting on the future of AI and the obvious companies are the players, but like you start to wonder like what happens to Google, like their whole businesses ad revenue.
[00:48:08] Paul Roetzer: Amazon is dominated by their cloud business. A w s drives massive profits and they own a huge part of the market share per cloud. But Microsoft's kind of the dark horse, just like quietly moving along. And then Apple's the great unknown, like what do they do? They got 3000 AI ML employees and you never hear anything about it.
[00:48:26] Paul Roetzer: It's, it's infused into your iPhone. So yeah, I mean, like you kind of let off, like this episode isn't about, here's the answers. This episode is about stretching your thinking to. This is way bigger than ChatGPT. Mm-hmm. . It is way bigger than AI writing tools or image generation tools that is the tip of the iceberg.
[00:48:45] Paul Roetzer: It's like, again, the shiny object that got everyone thinking about AI and talking about it. But most people, and this is a cautionary tale, most people right now talking about AI who haven't talked about it before, have no idea about this stuff. So what I, what I'm encourage people to do, If you are going to publicly have a point of view, and if people listen to you, like if you have a platform to talk about this, talk about it, do it, but also invest the time to understand what's going below the surface, because that's where the real innovation is going to happen.
[00:49:18] Paul Roetzer: And you have to understand who these players are and what they're building to understand why these shiny objects exist and what might come. And that's again, kind of what we're hoping to do with the institute is connect these dots for people. Or if nothing else, ask the questions that get you thinking about this at a deeper level.
[00:49:38] Mike Kaput: I'm going to throw this out here as you're talking through this. I feel like maybe a person could have felt like right before you found out about the Manhattan Project, right? Like it, like the project that created. Like nuclear power and weapons. It feels like what we have seen to date is just scratching the surface and there's this vast development that is happening behind the scenes that we know nothing about that could change honestly, things
[00:50:15] Paul Roetzer: overnight.
[00:50:15] Paul Roetzer: We didn't even get into the government. , . It's so, God, I, I really don't want to go down this one , but I would say that, I, I don't think it's happening. I, I hope it is, but I, I don't think it is. We need an Apollo level move on from the Manhattan Project's going to be top of mind with Oppenheimer coming up, but like Yeah.
[00:50:37] Paul Roetzer: We'll go into a positive framework, let's say the Apollo mission, . Yes. And the money, the government and the resource government put into, you know, landing, humankind on the, on the moon. The governments, and again, you may be listening to this all around the world, governments should be doing that. Like there should be a massive investment being made.
[00:50:57] Paul Roetzer: And I know that there's billions if not trillions being spent on this. I know in, in the US in China, we've talked about AI superpowers by Kaili. Phenomenal book on this topic. If you want to understand what's going on in the race for AI supremacy, it is a race. There is trillions being spent on it. You don't hear about it much.
[00:51:16] Paul Roetzer: And that's a wild card because the stuff that's being developed by darpa. So if you want to learn about darpa, read the Pentagon's brain going back to the eighties and earlier where they were trying to emulate, the human brain, you know, trying to, to rebuild how the brain works to create super soldiers and stuff.
[00:51:32] Paul Roetzer: The tech from there is actually what gets commercialized. So some of these companies that exist, like example, Sury was government technology like that that was built for, you know, the battlefield. Yes. Like the government is a whole nother realm of like what's being built with Black Ops money. And I have zero insights into that other than
[00:51:56] Paul Roetzer: Yeah. So I'm not even going to like, oh, and they're, I love it. There's Sury talking to me. I just, oh, there you go. Talking about sur she's listening. My
[00:52:04] watch
[00:52:04] Paul Roetzer: is like always listening.
[00:52:07] Mike Kaput: Yeah. No. Yeah. And that's, and that's a whole, we don't want to get into conspiracy theories here. It just feels like this vast.
[00:52:15] Mike Kaput: Collection of talent. When you mentioned the kind of AI talent index that you're working on, yeah. You have to ask yourself, these are the, the most in demand, highest paid, most highly trained professionals, probably on the planet at, in this field at this point, regardless of how many people are in school right this second, for those types of roles in that type of work, how on earth do you bridge that gap as a company if you haven't hired that talent already?
[00:52:40] Mike Kaput: If you're going, if you're looking at a hundred verse 4,000, I
[00:52:43] Paul Roetzer: mean, Well, and they're all going to leave. Like, so this, this, the other thing, again, I, this episode could literally go for three hours. , we could just keep pulling at threads here. But what's going to happen is already started to happen. You're going to see this explosion of generative AI tools like image generation, video generation, things that generate stuff within sheets, things that generate like PowerPoints we talked about.
[00:53:04] Paul Roetzer: Anything that you. AI is going to assist you in building it through text prompts or through some prompt. You're going to tell it what you want and it's going to do it. So if you are a leading language model researcher, for example, at Google, cuz this is what happened and you're working on the state of the art, you're watching, opening eye release stuff and you're like, we've.
[00:53:27] Paul Roetzer: We had that 18 months ago, like we had it 12 months ago. We had say we had that tact and look at them like they're getting all this glory and all this money, and they're doing the thing we invented, like Google invented the transformer. They're, they're built and you're sitting within the research labs and you can't release that stuff.
[00:53:43] Paul Roetzer: Like you just leave. You just go like, I'm out. Like I'm just going to start my own thing. I'm just going to build a generative AI tool. I don't know, like we asked. But Don Misha was, you know, from, works in AI research, from Google, and he was at our conference, and I asked him point blank, at the keynote said, why are you at Google?
[00:54:03] Paul Roetzer: Like you could do anything, be anywhere. And he said that to him, his ability to work on agi. I, so general intelligence was greatest at Google. That, that he felt that his ability to be at the forefront of the, the true holy grail, the end. Was, was greatest within a research lab like Google's research lab.
[00:54:27] Paul Roetzer: And so I don't, again, I haven't asked that question of a bunch of AI researchers, but I think it's a general consensus that if you, if you want to be at the forefront of agi, like which again is the mission of DeepMind, it's the mission of open ai. It's the unspoken mission of research at probably Meta and Google and a w s, like it's why they're doing what they're doing is to get to a g I.
[00:54:49] Paul Roetzer: then you're, you're going to stay at these big companies and do it if you think that's the best path to it. But if you at some point realize like, I can just go build my own stuff, you may have this like crazy exodus of all these like top AI researchers who just go build insane stuff like character. Dot AI that we've talked about.
[00:55:11] Paul Roetzer: A conversational agent built by a guy who was one of the leading researchers of language models at Google. Hmm. And so I think that's part of the reason we're going to see a massive explosion of, of AI tech and generative AI is because, one, it's, it's relatively easy to build because stuff's getting open sourced.
[00:55:25] Paul Roetzer: The models to do this are getting open sourced. Two, there's a lot of talent that can build these things and they can build them quick. Now going back to business and marketing, the question becomes, well, what kind of talent you hire? And my guidance there would be they don't exist. Like if you want to go find a bunch of marketers to do this, like it's part of what we're doing with our online academy at the institute.
[00:55:46] Paul Roetzer: The piloting AI for marketers course is like, let's create the next generation of, of professionals who understand this stuff. And eventually we can almost play matchmakers. Like, oh, you want to hire AI savvy marketers? Like, we got thousands of 'em. Like, that's part of the play. But in the meantime, you just have to train your existing team like.
[00:56:05] Paul Roetzer: You can't just like stop and go find these people, right? Cause they're not coming outta university. Just go like, build an internal training program, get 'em the foundational stuff about AI and get 'em using the tools as like the best path forward.
[00:56:19] Mike Kaput: That could not be a better way to pull all these threads together.
[00:56:23] Mike Kaput: Wow. I we, I love that we came into 2023. Really hot here. Coming in hot. This is great. Yeah. . Well, Paul, thank you so much for the insights here, for the work you're doing to understand this stuff. I mean, I think the audience is going to get a ton of value out of this.
[00:56:40] Paul Roetzer: Yeah, I mean this is the stuff that's fun to talk about and again, I think, you know, for people who listen, thank you, for people to reach out to us and let us know.
[00:56:46] Paul Roetzer: Like it means a lot to us. Cuz this is the stuff Mike and I have like, sat around the coffee maker for the last like six years talking about by ourselves. And so it's, it's like wonderful to have a growing community of other people who also are interested in the topic. Cuz we could talk about it all day long.
[00:57:03] Paul Roetzer: There just hasn't been enough people listening, I don't think for, for a few years now. Granted, like we have 35,000 subscribers themes, so it's not like it's, but, but this is just, feels different, like mm-hmm. , it feels like, like progress is happening and that's very encouraging to us and it's encouraging, when we look out to the future of the industry and business because I, I think this stuff is essential.
[00:57:29] Paul Roetzer: In your career and in your company, like you, you truly have to solve for it. So I, I think it's great that we're hearing from more and more people who are trying to figure it out and, like I always say, just like, be curious and keep taking the next step to learn this stuff. You're not going to figure it all at once.
[00:57:45] Paul Roetzer: That's fine. You can figure it out in a day with the pilot AI series, but like, it's not all going to make sense yet. But I mean, this episode goes deeper than I've ever gone publicly. Like I've never talked about all this stuff publicly. Sprinkle it into keynotes every once in a while. Yeah, this is probably one of the most far reaching conversations we've ever had about it, certainly publicly.
[00:58:05] Mike Kaput: Awesome. Well, thanks again and Happy New Year.
[00:58:08] Paul Roetzer: Yeah, happy New Year everyone, and thanks for joining us. We'll be back again next week, so it's, we'll try and stick to the weekly cadence, throughout the year. So yeah, definitely subscribe and, reach out to us and let us know and, you know, give us a like and a rating and whatever you can do to, to help spread the word.
[00:58:23] Paul Roetzer: We appreciate it. Thanks everybody.
[00:58:26] Paul Roetzer: 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 marketing ai institute.com. Be sure to subscribe to our weekly newsletter, check out our free monthly webinars, and explore dozens of online courses and professional certifications.
[00:58:47] Paul Roetzer: Until next time, stay curious and explore ai.