AI Tools, Content Systems, and Growth in Marketing with Jakub Grajcar

Speaker A: Hey everyone, and welcome back to another episode of Paris Talks Marketing. Today my guest is Jakub Greizar. He goes by Kuba and he is the head of marketing at Talki AI. So Kuba, welcome to the show.

Speaker B: Hi Paris. I feel like this is a long time coming. We've known each other for a few years now and I'm super happy to be on the show. Listen to a few episodes as well. And I'm honored to be invited.

Speaker A: Thank you. Great to have you. Kick us off with a brief overview. Tell us, uh, who you are and what you do.

Speaker B: So I've been a marketer for over a decade now. My primary background is from writing public speaking. I studied English. Then I got my first marketing job and I basically hit three different industries on my way to where I am at present. The first one was software development. I was head of marketing and marketing manager. Started off as a frontline marketer, but advanced to marketing manager at a software development company then. And this is how we met Paris. I moved more towards the side of US agencies. And by agencies, I mean marketing agencies. I was at an agency called Zenpilot helping marketing agencies to get success with ClickUp implementations. And now another pivot. I'm at talkie AI as you mentioned. And we're actually a, uh, healthcare AI company. We've got AI agents that can take the burden off the medical front desk, picking up the calls, making calls themselves. So within that time I advanced from basically a content marketer to being a little bit more across the whole breadth of marketing at this new organization. And within that time we had Covid happening, we had the pivot to, you know, streaming everything to online webinars. Now we've got the kind of the pendulum swinging the other way towards in person events. I feel. I started pre AI. Now we're very much in the thick of it. So yeah, it's been an interesting journey thus far and I think my favorite part of it was honestly working with people and having my own team. That's the part that I'm really passionate about, mentoring market. So I, uh, asked myself recently would I want to switch careers. A lot of people are thinking about that recently, especially with AI in the picture. I would prefer to stay a marketer, to be honest. I really like it.

Speaker A: Yeah, we're going to go a lot deeper in that area. But I would agree with you that there's no better place to be for marketers now than to stick it out. And this phase that we're in right now is, I think separating the Exceptional marketers from average marketers. But if you believe and you're ready to strive for that excellence, you have tools now that can give you absolute superpowers. So I think it's just about how you frame that and think about it. Are these tools that give me the superpowers to be top 1% marketer, or are these tools that will replace the work I'm doing and therefore I should get out of this before it's too late? And that's entirely just mindset, I think.

Speaker B: Yeah, exactly. And you know, I think another really important skill right now is systems thinking. Right. You can be a great marketer, you can be able to produce great output content, for example, but you really need to be able to break it down into its core pieces if you want to scale it. Right. It's not enough to just prompt an AI with, okay, here's an article I wrote, write more like it, but you need to break it down into, okay, what makes it work? How do you see its success? Both, you know, you can do this thinking yourself, you can have the AI help you do it, but I do believe you have to self reflect too. Of like, what are the building blocks of, for example, great content or great PPC campaigns? Because if you can actually get to that, if you can actually set the criteria, then that scales really well with AI, right? If you prompt it right, if you set the right guidelines and guardrails, then, you know, you can see that impact basically tenfold.

Speaker A: Yeah, absolutely. Let's get right to it. We've already started on AI. I had planned to ramp up to that, but we're already right in it. How are you using AI? How is it helping to empower you, drive productivity and give us some inside tips of the trade, stuff where you feel like you've made some breakthroughs.

Speaker B: Okay, I've got some stuff that's more, you know, high level and then stuff that's more tactical. I feel like I want to start with the tactical. So I think, you know, if, if you're a marketer listening to this and you don't yet have a personal AI subscription, then I think you're missing out. That's something that I just recently did where I got my own ChatGPT subscription. I know that you know, in our previous discussions you mentioned Claude. Some people swear by Gemini. I tried, you know, the vanilla, the classic. And it's honestly helped me learn a lot about how it works and apply it to different contexts. It has shown me the value of like a longer context, conversation and through interacting with ChatGPT, I've learned a few things about how I work with AI and where it works and where it doesn't. So one tactical tip is I'm a talker, okay? That's why I love being on podcasts, both as host and guest. We have a podcast scaling practice management. Check it out. Shameless Plug. So one thing that I would encourage people to do is to check out the dictation feature or check out the kind of the conversation feature. These are two different icons. Microphone versus the, uh. I don't even know how to describe it, but it's in its own little bubble there.

Speaker A: Are you talking about the Mac app? The ChatGPT Mac app that has the record feature?

Speaker B: No, the ChatGPT app itself. In the same input box where you can put in text, you can also click a microphone icon. Yeah, you're saying that just voice dictation. But it's the best voice dictation I've ever used, specifically because you can switch between languages and it's seamlessly picks it up. Even dedicated voice dictation software doesn't do that right now. You know, I've got like Dragon Dictation, I think it's called in my SwiftKey app. Whichever. Anyway, so ChatGPT's dictation is the best right now. And I think that allows you to give a lot more input, input that's more human and reflects your voice better. So that's something that I would encourage marketers to do and then think about, you know, another tactical tip. But sometimes I'm pressed for time when interacting with chat, so I put it in what I call config mode. So I just prompted to reply, you know, Reply with like 5 options where I can just reply with A, B or C to pick from each option. You know, it's a small efficiency at the end of the day, but it allows me to just, you know, to just help it serve me in a way that I don't have to do so much typing.

Speaker A: Can I ask you a little bit more about that? So you're asking ChatGPT to just summarize its answers into three simple choices, or A, B, C. Or am I getting that right? Can you maybe give it an example?

Speaker B: Just this morning I was using ChatGPT to design my next tattoo. I gave it a couple images for inspiration, and then I replied with something like config mode. Make some suggestions on how to improve it. And for each suggestion, uh, let me pick from three options, and then I'll respond with just 1, 2, or 3 so what that means is, you know, it helped me take into account aspects of this that I wouldn't normally have taken into account. But instead of making it a long conversation, I just pick from the options that Chat pre selects. Another way that I tried in the past was, you know, because the more precise you can be with your prompting, the better. Right. So I was writing content, and I basically asked Chat to break it down into sub features, like, list all the features like formality or directness, or, you know, being insightful versus being salesy. And then I will respond to you on a scale of 1 to 10, how I want to configure you to be in each of those categories. So, like, you know, you can be 9 on the educational scale. So that means. Means no hard sell. Just focus on providing, uh, the insights. Right. And it's a nice way, I think, you know, for people who have a little bit of an analytical bone. And I'm kind of a mix between analytical and linguistic that can be useful to try and get more consistent outputs and just to play around with how it affects the output.

Speaker A: That's interesting. So is it an actual feature called config mode, or is it something that you type into?

Speaker B: It's not an actual feature.

Speaker A: Okay. It's just something that you include in your prompt somehow to get it to. Yeah, okay.

Speaker B: It just affects the flow of the conversation moving forward. It's a little trick that sometimes helps with the conversation. I like to keep it concise when I talk with chat. I mean, I either keep it concise like I just described, or I just talk for five minutes and I let Chad fish out the good stuff from that. And on a more high level, I was just talking with Chris Dubois recently, and what he showed me was something that I think all marketers should pay attention to is, uh, I'm kind of relaying his idea here, but basically the way it worked is he showed me a, uh, make.com workflow. And again, coming back to. Some of us are more talkers than writers. He uses a tool called Super Whisper. So he talks to Super Whisper about his business, and he helps agencies primarily with their offers. He's an agency consultant, and he has a whole flow where this input is taken into a variety of nodes within make. So the first node would, for example, outline the content. The next would write an intro and outro. The next would do, I don't know, SEO analysis. He has a whole separate node, a whole separate ChatGPT agent trained on his voice, which just takes as input, you know, a pre written article and outputs the same article, but paying special attention to writing it in his voice and coming back to systems thinking. I feel like that's how marketers should be thinking about this right now. How can I create like an assembly line? An assembly line has a negative connotation. I do recognize that. Right. But that's the trick and I think that's the challenge these days. You need to build an assembly line, otherwise you'll fall behind. But the assembly line still needs to hold your standard of quality. And that's going to be kind of the target to hit for a lot of marketers moving forward. And with automation like make and, you know, by following some smart people, I think we can get there.

Speaker A: Yeah, we're really deep into make automations ourselves right now, and it's fascinating what is possible. I think that AI has reawakened interest in automation because automation has existed for many years. Make.com, which was, I think used to be called Integromat, I believe. But that product existed for many years before AI came on the scene. But now the fact that you can inject specific AI outputs with prompting as part of that workflow, it really just changes the whole game. So we've built some, um, really interesting make automations mostly around content generation. But I totally agree with you about voice. I'm a big user of Super Whisper. I use it every day. And even a majority of my prompts, I use a mixture of ChatGPT, Gemini and Claude, but predominantly ChatGPT. But I'd say 80% of my prompts are voice prompts. And I know that the native dictation feature is very good. But I do have Super Whisper and since I have that and I'm paying for it, I feel obligated to use that.

Speaker B: The question is, are they paying us to talk about them right now?

Speaker A: Yeah, Right.

Speaker B: Which they're not, right? Reach out to us.

Speaker A: Well, maybe he'll be my next guest then. But in any case, I think that that's another big aspect of this AI revolution is that it's allowing such a more natural input, which is the voice input. And there's all this talk about everybody needs to become great at writing prompts and prompt engineering. I really don't agree with that and I'd like to hear your opinion. I don't think that there's going to be a subset of people that can exceptionally master AI because they're so good at prompt writing. I think it's all trending towards voice and your ability to just bring in supporting Documents, transcripts from meetings or whatever else, along with a, ah, really extensive prompt that could just be you rambling on for five minutes and just let the AI pick out the pieces that it needs from your prompt, your voice prompt, and from the documents that you upload. It will give you a great output. So I don't believe, and I think that the notion of prompt narrowing actually prevents some people from really diving in headfirst into AI because they think there are greater writing prompts. And I am not comfortable. I'm confident and so I better, I better hold. But what do you think?

Speaker B: That's really interesting. So first of all I will say, you know, what you're putting in my head right now, which I felt but never named, is prompt engineering is a pretty bad brand for what you're really doing there. So, you know, that's where I would start off because it sounds like software engineering. Software engineering requires like deep domain knowledge. You need to really train yourself to get anything done right within traditional programming. I think with prompt engineering I would, you know, if I were to rebrand it M, I would think it would be more prompt tinkering, prompt experimenting, which I think is gonna put a better image in people's heads. Because the second thing I wanted to say is I don't think, I don't know if I wouldn't take the opposite stance here, Paris, that prompting is actually quite important. I mean just recently I was trying to put together my own podcast just to pick the clip that I'll use for like the initial teaser. Right. And it was honestly a little bit frustrating. The output kept coming out not exactly as I wanted it. So I do need to do more prompt tinkering, if you will, for this do agree on. You do need good context and I think that's going to be a skill in itself. How to kind of gather that context, how to present it in a way that the LLM can work with it. Uh, in a. But then the last thing I would say here is, you know, when you're working on something like this apt, you really need to be thinking about how to structure the prompt. And the good news is here though, you can use AI to help you with this, right? It's not over whatever skill you want to acquire a good, uh, even if it's AI related, a good first step is actually using AI to explore it. I don't know, um, if on our recent call when we were talking, just recently tested out the deep research feature on ChatGPT and it's amazing if you want to like dive deep into some sort of topic and you want to do it from your perspective. Like you mentioned, give it like a comprehensive voice input in terms of what you want, what you already know, what you've been thinking about a given subject, but then let deep research run with it and it can be a really full primer. So I found it funny, I gave it a prompt, honestly, uh, some stuff that I was uh, dealing with on the personal side of my life, but because it was part of a longer conversation where we were already discussing that all of the output was like an ebook that was super relevant to, to everything that I was dealing with. And yeah, I can really recommend that that's another thing to tinker with.

Speaker A: So when you did deep research, it was able to draw on its memory of your past conversations as well as the, the research itself. Is that what you're saying?

Speaker B: Not conversations. But for this particular, you know, was like interpersonal communication. Whenever I had an interpersonal communication problem, I used this specific chat, you know, this specific chatgpt. So I've been doing that for a few weeks and only recently I uh, thought o, okay, let's check out deep research within that same conversation I used that feature and it did draw on everything that I, you know, that I mentioned. I asked it for, for example, whenever you can give me specific words to use. Right. So it was very relevant to my context. And that's why, you know, the first thing I said was I recommended people to get their own chatgpt. A lot of marketers, I heard them say, well, I've got the company ChatGPT that serves my needs right, but non work challenges. And when you really want to, you know, show everything that you might want to learn, it's better to have a personal account. And it's also really inexpensive given what it can do. That's the part that really blows my mind.

Speaker A: Yeah, I think the ROI on a 20 per month ChatGPT account is a no brainer. And uh, right now I'm grappling with this as an agency leader. I haven't yet given every person a separate ChatGPT account. We do share one company account and I know that probably giving everyone a personal account would unlock more usage. So secretly hope that a lot of them have their own personal accounts anyway that they would use. And I've been thinking about this aspect of a personal AI and a work AI and if those two things should be merged because the memory feature is getting so good, especially with ChatGPT, that it's drawing all the past conversations and using that as context for any answer that it gives you. I don't know if it's appropriate for a company to Give a company ChatGPT account to someone for them to develop their overall Persona, so to speak, and to build those memories with all sorts of probably personal questions and things that maybe we shouldn't ever have access to. And I just wonder, is it best for people to have one account that they personally own and pay for that's more or less for their own personal life and another one for work where those conversations are more restricted to the work? Because still haven't come up with a good answer for this yet.

Speaker B: You know, for a company like ours at ah Taki, we're in the healthcare space, right? So the answer becomes much more clear for us when we're dealing with, you know, potentially patient data. It gets a lot more sensitive in that respect. And for any other company, really, you've got data that you do not want your workers to be putting into a personal GPT, personal AI. So in that respect, I do think it should be separate. And I think this also influences the culture of companies because, you know, we do have a ChatGPT account. When we go into that account, it's as if you could see, you know, previously what everybody is googling and you're not always Googling stuff that, you know, you think you should be googling. I don't even mean, you know, anything risque, but I mean, okay, this is a marketer looking up like how to write a successful blog intro or something like, shouldn't you know this? You know, it can affect your perception within the company. Still, I think I would rather the culture moved in a different direction where it's okay to seek knowledge or even update the knowledge that, uh, you think you have, but you want to improve it so that part I think can be shared. In the end though, I think when we're looking at the perspective of a few years out from now, probably companies are going to have their own brains, right? There's going to be an agent or more likely a set of agents. I think this is where this is all going. We're just kind of in the nascent phase of it where you will meet, you know, you join a new company, you meet the human team and then you meet the AI team, right? This is the agent that reviews our content. This is the agent that can answer sales questions. This is the agent that can answer product questions and the companies that can figure out kind of the pipeline of data so that the product agent knows everything. That the product can do and is constantly updated. I think that's kind of a key factor towards winning here and towards getting ahead of the, uh, competition.

Speaker A: Yeah, I do agree it's probably heading in that direction. And I'm anxiously awaiting ChatGPT 5, which is supposedly dropping sometime this summer. And initially the rumors were July, but it's getting close now. But a lot of rumors are that that ChatGPT5 will be an entire operating system for AI that companies can move away even from, say, Google Workspace or Microsoft. But I think we're still in such early innings of this that there's a lot more innovation ahead, for sure.

Speaker B: You know, there's another interesting aspect of this is how, you know, because I'm leading a team, right? It's interesting to be a, uh, leader of a team or ahead of or in any sort of managerial position right now when you've got team members and you kind of want to calibrate them as well on how enthusiastic they should be about AI, Right? And I think different leaders are going in different directions right now. And there might not be one right answer. History will show. But I feel a little bit strange. But I also feel like it's the right move where whenever I'm presented with a problem or an issue to be solved by one of my team members these days, I do ask them, have you tried ChatGPT? Because, and you know, it can feel, it can feel really dismissive. But at the same time, I think it's a part of training and it's a part of affecting the culture in the sense of, okay, apply AI to the problem, to any problem first. Not just kind of, okay, I drafted this blog. Could you review it? But also, can we have an AI that does the review, or at least 70% of the review? We do want to have, as they say, human in the loop, or there needs to be, I think, a final human qa. But how far can we get it without the human input? Because that's the way we scale and that's the way we stay ahead. It's unfortunate. I feel like all marketers have been put on this kind of hamster wheel right now, and it's just the hamster wheel keeps speeding up and that can be stressful. At the same time, I feel like we're getting towards, towards a point where we're just going to kind of change our thinking here. And instead of thinking, how can I automate this? I think a lot of people are going to be thinking, okay, how can I summon create an agent for this. And how can I expand the team to handle more of these tasks? Because I might be a manager of people, but to be honest, my frontline team members, my individual contributors should be thinking that they're also managers in their own respect and they have AI hands to, to do the work with. And I do wonder. Last thing to this point would be, you know, think about kind of the generational differences here. Right. So, uh, I would say my generation is one that, uh, was not like Internet native, but almost. We were quite young when the Internet first.

Speaker A: Are you a millennial or a Gen Z?

Speaker B: I'm a millennial. I look like a Gen Z because I've got vampire genes.

Speaker A: Could be kind of on the border. I could have believed either one, but so, yeah, millennials. So you actually have a, you have a memory of pre Internet, don't you? You actually can remember. Okay, yeah, for me it was high school, so I definitely do.

Speaker B: Yeah, exactly. But think about it like that. You know, there's going to be people born that will have, or like even right now they have vague memories of pre AI. Right. And an AI native generation is coming and I really wonder what they're going to, to do with the world, basically. I know I'm getting a little bit philosophical here, but you change your thinking when you've got AI from the word go. You can use AI in school even, for example. I know that's a big debate. And I wonder what that generation is going to show us with what really can be done using AI.

Speaker A: Yeah, that will be fascinating. I want to come back to your point about trying to lead a team responsibly with AI and to find the right ways to encourage, I think, daily AI use. Now it really needs to be daily. And this is another thing that I think, uh, about a whole lot leading an agency. And I'm thinking about introducing a framework. And, uh, I'll bounce this off of you where there's three phases of it. And the end goal is that I want people to think that ultimately they will all become orchestrators of their own AI teams of maybe there'll be agents, AI agents or who knows, but that in the future everyone will actually be a manager, so to speak. But it's more of an orchestrator or a conductor of AI systems and agents that are doing the work that they do today. That doesn't mean that they're going to be replaced. It just means that they're moving up to a higher level of orchestration. And to get from where we are today, where people are doing the Work to where they'll be orchestrating the work. I'm, um, thinking about three different phases. And phase one is validation, which I would ask everyone. Before you ship any piece of work, before you launch a campaign, push out an email campaign or post a piece of content, just pause for a second and give it to AI and let's see what AI says is good or bad. See if AI can give you any extra added value before you pull the trigger. And there we can assess what AI is good at and helpful at and where, where are the gaps. And then we would move eventually into the second phase where it's collaboration, where you start to give AI some of your work, the work that it's already demonstrated that is good at in the first phase. And there it's a little bit of. This is probably the most dynamic phase because the marketer is doing some of the work, it's giving some of the work to AI and it's really at the discretion of the marketer. And then eventually you move to the orchestrator or the orchestration phase, where the marketer really pulls out and builds AI automated workflows that now are largely doing all the tasks, the daily work that they used to do, whether it's writing content, doing keyword research, producing a creative, or uh, running email campaigns or other type of things. And the orchestrator is really just checking the work of the AI the way that we asked AI to check our work at the beginning in the validation stage. And I don't want to rush that, but I'm thinking that it's maybe like a 12 month. These three phases is probably a 12 month entire cycle. But that's uh, right now I think one of the things that I struggle most with is I really want to know how often and how people are using AI, and I can't see that too well with just one shared ChatGPT account.

Speaker B: So it's really interesting because the metaphor in my mind for this is a little bit different Due to kind of the nature of my company and also due to some recent and not so recent conversations I've had, I'm really leaning into the agent side of it all. And I think, you know, one of the best steps that you could take as a team, I mean this as a general one could take as a team, but it might also apply to your example here, Paris, is to really huddle together and workshop the first AI agent together so you can set standards for what you want the agents to be able to do. Like what's something that if we could create Hire another person. What would that person be doing day in and day out? That's like a real need and that AI can do, right. So I would start there and see, okay, what kind of data does this agent need? How can we make it the best it can possibly be? Because I think it's less valuable to solve like a dozen problems partially. I think it's much more valuable to solve even one problem completely. Right. Something like this one handles PPC campaigns. But we've trained it on enough data, we've kind of pruned the outputs, we've made it consistent enough. Or maybe it's like a make workflow already where, okay, if we're dealing with a PPC campaign, there's a guy for that. The guy is not a real human person, but still there's somebody that you can just offload it to. And I think, you know, if you can hit a bar of quality with that and if you can show, first of all, like, imagine the additional motivation for the team. Right? Okay, this really works. We've got like a genuine solution for this one problem. What's the next problem we can solve? So I think that's, you know, quality over quantity here would be important because, you know, to your point about, for example, that first validation phase, I feel like there are some types of work where that would be useful and then others where it might not be. Like if I draft a blog article and I do it the old fashioned way, let's say content craftsmanship, right. I don't know if I would necessarily trust an AI to be my editor and try to improve it. Maybe just poke holes in it. Right. It could be a proofread, it could be making suggestions. Like you mentioned, I would still be making the final call here. What's interesting is we've often been using AI to actually create the first draft. And then you know how it is, the blank page is super scary, but you've got a, uh, pretty crappy draft in front of you. Oh, I know how to improve this and it makes everything happen so much faster. Right. So I think there's a bunch of ways to approach this. And again, experimentation is key and depending on the use case, it'll be different. But yeah, I would ask. And uh, this question I've been asking myself recently too is like, okay, if I really lean into the agent side of things, if I think I can create a very specialized new hire, they're not a person, but they're still real in a sense of their impact, what kind of person would I hire? What kind of agent would I bring to life? Play Frankenstein for a moment?

Speaker A: Yeah. For me, that person would be the agent builder. And maybe it's like an AIOps engineer or some kind of role like that. To me, that's what we need most right now is someone who can take the rough concepts of agentic workflows and make them reality.

Speaker B: Okay, but you're talking about a human person here, or are you talking about an agent that builds agents?

Speaker A: Oh, I'm talking about a human person who can accelerate our building of agents.

Speaker B: Yeah. Yeah. Well, that might be an interesting trait to get into. Yes. What I did mean, though, is, you know, on the side of agents, like, what's the first agent you're going to build yourselves?

Speaker A: But yeah, I'm going to agree with your point that you can easily get overwhelmed, but it's best to start small and think about, here's one thing that I always have a problem with, or something relatively small. Let me see if I can have a focused, narrow scope and have an agent tackle this, and then you just iterate that and you move to the next thing. And then eventually you can chain them together into more complex workflows. But I think starting with a modular kind of approach, one piece at a

Speaker B: time, that's kind of how I've been structuring even non AI teams, honestly, instinctively, over the years. For example, at Zenpilot, what we were doing is we were editing the podcast and we were also producing clips from the podcast and we used external services for both. And very often those are bundled. Right. We've got one guy who does the editing and the shorts. And what I found is, cost effectiveness aside, because it was honestly really similar, I prefer to have a separate person just for the main episode editing and a separate person just for editing the clips. There's no clash between the timeline for the clips versus the timeline for the main episode editing. I just knew if I had problem X, I go to person Y. And I think that's, you know, a frame of mind that's useful for. For putting together AI teams as well, where it's like, if it's more specialized and you just. It's much more clear who to go to or which agent to go to using. Who here is sounds a little bit dangerous, feels dangerous for each particular problem. So that's kind of. To add to your point.

Speaker A: Yeah. I think what's also interesting about what you're mentioning is that because AI agents are. Think of them as specialized workers, but there's no limit to how many we can hire or bring in. So because of that you can get extremely specialized with an agent and you wouldn't really take that same approach with a human team. It's maybe you would think about this person is more valuable because they have this broad set of skills. They can do a lot of different things. But with agents I think you're just not limited with the number that you can build and deploy. So why not make each one ultra specific and ultra focused. And everything that I'm hearing and experiencing is that they work better when they have an extremely narrow focus and they do one job really, really well and then they can hand it off to the next one. But then on the human side, actually what's interesting is that if marketers are becoming these orchestrators of agents, then the human marketers are becoming more of generalists really. I think it's helpful for them to have a pretty broad spectrum of knowledge across all the disciplines in marketing and not to be specialists. I think if you're going to hang on to being a social media marketing specialist and that's all I do, that's what I'm good at. I think it's just going to be harder for these types of marketing specialists in the future.

Speaker B: It's harder and harder to look at some part of the marketing stock and say I don't do that. Where you can learn it much faster using AI. And also you can off source some of the work to the AI. I know some marketers who are quite frustrated with that, to be honest. You know, people saying, I signed up just to write. I just want to be passionate about creating great content. And it's unfortunate. I still don't have a good response to that, whether that's still going to be possible. I think there's going to be different kinds of new types of marketing jobs that we haven't even thought about yet. Or it could be marketing orchestrator like you mentioned earlier in this episode. Right. It's a season of a lot of change. But at the same time, we've been here before. I feel like saying as a civilization, you know, it's like the industrial revolution. It's like the computer revolution all over again. I just feel like people need to accept that there's some upskilling to be done on their end. I hear often, I don't know if you do, but I'm sure you do as well. People having various takes like, are we even going to have jobs anymore? And I think people are still going to want to work. We're just at the stage where we haven't come up with the new jobs yet. The old jobs, turns out we can pass on to AI in a lot of respects. So what are the new things that us, uh, humans can be doing or that we can still be doing? So that's kind of part of what creates anxiety within marketers and I think professionals in general, myself included, being perfectly honest with you. Right. But at the same time, at the same time it's at least a little bit exciting. Right. It can be a different kind of work. It's almost as if you need to come up with your career all over again. You know, you thought you picked for life and, you know, it used to be that you could do that. Now you kind of have to rethink yourself.

Speaker A: Yeah. It's just fascinating and really inspiring to think of this utopian vision of abundance that AI could create. AI could create such abundance that work becomes optional and more people will be able to follow their passions. And some people with a naturally stronger work ethic might get agitated and think, I've got to do something, I've got to stay active.

Speaker B: But then think about it though, Paris, we're already there. Think about all the people who are making a living by just being content creators, just being lifestyle content creators. Content creators, they live their life, record it, share it. They have a Patreon or something like that. I know of YouTube channels where you've got, uh, a guy usually kind of my age talking about their favorite video games or obscure video games, retelling the story, doing in depth reviews. They've got a Patreon and they make their living that way. Streamers can make a living just being an Internet personality, playing video games and talking about it and becoming entertainers. So I feel like we're already making that shift where stuff that you would think would be pure fun in the past now becomes honestly not even a, uh, valid, but sometimes a lucrative way to make a living. So I feel like, you know, there might be more jobs like that or more professions like that on the horizon.

Speaker A: I think so, yeah. As long as there's enough audience to watch. Or maybe that doesn't even matter anymore if people are able to express themselves. Maybe it doesn't even matter. In a world of future world of AI abundance, it may not be necessary for these types of creators to monetize and support, you know, hit a certain level of income to support themselves, that they just do it purely out of their own passion.

Speaker B: Well, the issue of income and, um, how money is made is, I think, separate and much more complicated. And I Couldn't comment on that in any smart way, I don't think. But let's. Let's see how that unfolds.

Speaker A: Well, yeah, I think there's still three bases. Basic needs that we have to pay for, which is food, clothing and shelter. And food could become a lot more efficient. I mean, there's some companies now that are working on food production, where you can order custom boxes for delivery at a much cheaper cost. And I think housing could be transformed as well. But I think a lot of these big costs that people have could come way, way down, and then the level of income needed to support themselves would also come way down. And that would give them a lot of creative options.

Speaker B: There's definitely something to be said for customizability. Right. That's much more accessible right now. It used to be if you wanted like a custom design for something, custom logo or whatever, you had to go to a person, pay them to do it, or, you know, do a lot of learning of how to do it in terms of software composition, etc. I mentioned earlier I was using ChatGPT to design my next tattoo. Right. So I am feeling what you're saying in the sense of, of the future might be much more custom. Um, right. And then everything will be kind of adapted to your needs due to this increased production capacity, easier access to either design or, you know, code development. So it might be like that.

Speaker A: Yeah. Let's make one quick pivot because we have been talking a lot about content and earlier you mentioned that you are working on a new framework that is about a content flywheel. Tell us more about the content flywheel. And in particular, content generation is definitely changing, of course, because of AI and people are worried about AI slop. But I believe the reality is, in a lot of topics, AI now with its research capabilities that you just experienced with deep research, can produce even higher quality content than humans with deeper and better researched content. But I want to hear your opinion about the future of content marketing and tell me about your flywheel framework.

Speaker B: Right, so this is how we'll measure the conversion rate from this episode. I have a substack right now. It's not getting traffic from anywhere else, so look me up on substack. Jako Kreizar. You can also search for the name of my brand, Silk and spider, which I'm really proud of. No chatgpt involved. It's a legitimate organic shower thought. You know, there's a whole inbound versus outbound element there where the silk is the web that captures attention. The spider is the outbound that kind of goes out and grabs attention and that's a little bit related. So a piece I've got there is called the Content Conversation Flywheel, or how to turn networking, uh, calls into content. I'm passionate about both networking and content. And I think where this is going is the flywheel is very simple. You have a conversation which leads to a piece of content which leads to the next conversation. I think the important part of it, though, is how the content leads to additional conversations. Used to be you were waiting for the conversations to happen. I think think content marketers should have more of an outbound mindset about this and actually seek out the people in front of which they will put their content. So, for example, today we're talking about orchestrating AI teams. We were talking about, honestly, a lot of, uh, valid ideas. And I could use the transcript from this, right, to write a new piece for my substack, time permitting. But I'm actually kind of motivated to do that right now. The key part is though, once you have a new piece of content, you need to go out to the market and present it to them. And I don't mean just posting it on LinkedIn. Think about the five people in front of which you want to put this piece of content. Reach out to them directly, have them see it or read it, and if not, then, you know, use it as an excuse to strike up another conversation. I think this is where things are going because the one thing that AI can't replace is what we're having right now. So, like real live face time with each other. And I think that's where the human element will still be very important for the time being. And I think that's where marketers and content marketers should be going. So I used to spend my time drafting an article or, you know, reviewing it, putting it up on the website. That all is getting minimized with the additional time that's freed up. I can go out and have some coffee chats with other marketers. I can look for partnerships. You know, content is commoditized, right. But there's still platforms for the content. Right? And you need to be spending time seeking out the people who have the right platforms to put your content there, not just on your website, even though that can have its own impact. But I think there's going to be more of a, uh, you know, of a pivot towards, call it brand, call it pr, call it reputation building. Right. You can write a hundred articles, but can you get your article featured in? Okay, Forbes comes to Mind, even though they've got a bad rep for it now, but like an industry publication of your choice. Right. And I think that's going to become much more important. So I do sympathize with the content marketers who want to do just sit in their cave and write. I don't know if that's going to be, always be possible. Right. For much longer. I think there's going to be a stronger element of working on these partnerships and on this networking element. And you can do this solo too. I think, you know, that's how you bring momentum to your own content and your own ip, your own thought leadership is you need to be also having conversations with people who have seen your content and then, you know, they can poke holes in it, they can have counter arguments. It becomes discourse, you know, not just content for content's sake. And I feel like that's a lot more engaging. And also the people you'll be contacting are the ones that are going to remember you the most. That can drive word of mouth as well. And I feel like that's kind, uh, of a big part of where this is all going. Because AI can write, AI can edit AI with some automation, publish content, it can even run your social media, but it can't have the conversations. It can't jump on a call, at least not yet. Right. And it can't make an impression, an emotional impression on another human being that helps you be remembered. Remembered. So that would be my approach. And again, Jakub Greysar on substack, if I get a new subscriber from this, I'll be super happy.

Speaker A: I think the art of the conversation becomes all the more important because that's really what we're left with as humans. When AI can produce any kind of content at any amount of volume at scale, what's left is conversation. And I think also that's what we're trying to do in our content marketing right now is infuse use quotes, for example, quotes or even video highlights from real conversations. They could be podcasts or interviews with subject matter experts. But we're trying to integrate that into our content because these are words that have just been spoken from people. That this is not content that can be crawled and indexed or can be used for learning purposes. It's not in any knowledge base of any LLM. It's me and you having this conversation right now. Now is really original content. And if we want the content which is still heavily AI assisted, if we want that to perform well, we need to integrate this conversation that we're having and other, other conversations with that uh, content. And I think that sends the signals, if your goal is still SEO or GEO or aeo, whatever it's called now, I think that if that is your goal then LLMs and Google are still looking for those signals. Google has this acronym, EEAT key, which is experience, expertise, authoritativeness and trustworthiness. The conversations is what brings out the real experience and it provides that context that the uh, AI can't just go and curate from the web because it's not out there yet. Actually this conversation isn't out on the web yet. So I think that's always going to be there.

Speaker B: Again, I think it becomes about reputation, doesn't it? So you know the way you increase your authority is by being associated with the right other brands, the right other people who already have some sort of following or some sort of expertise piece. So all the more reason to try and reach out to the market much more directly than just through marketing broadcasts. But to actually have more of these calls, more of these conversations. And the piece itself can be based on a conversation. What I'm also saying is try to strike up conversations about the piece specifically with the people in your network these days. It's part of the nature of my job, but I think it's going to become more prominent in the future even for lower level role. I literally have a list of people that I want to be keeping in touch with all the time. Kurt Schmidt is great about this. He's got a good podcast too and great networker. I recommend him. So I feel like that's a big part of my job is there's a new piece of content, who do I distribute it to? 1 to 1 basically to try and get them to consume it. Because consumption is a challenge, right? The more there is of it, the more effort you need to put not into the production but to ensuring the consumption of the content. And I feel like doing this through one to one conversations is a great way to do it.

Speaker A: The thing that is ironic here is that this works because it's not scalable. Actually uh, one to one outreach for a piece of content is not at all scalable. But that is how you take a piece of content and use it to spark a conversation which can move that conversation forward, lead you into the next piece of content. And as you said, the snowball effect or the flywheel effect can start to happen. But I'll be much more likely to read something that if you ping me on LinkedIn and say I'd really love to get your opinion on this. This. I'm going to read that almost certainly. Yeah. And I probably should subscribe to your substack. What was the name of it again?

Speaker B: The URL is jakubgreizar.substack.com but something you can Google for it is Silken Spider, Jakub Grajsar or Silkenspider Substack.

Speaker A: Kuba, this has been fantastic. I think we could go on for hours here, but all good things must come to an end. And as we wrap up here, is there anything that I didn't ask you that you hoped I would have asked, or anything that you think could benefit our audience?

Speaker B: That's interesting. I feel like talking about YouTube for a second. I have a feeling that YouTube is in its golden age right now when it comes to content. That's not something I was planning on mentioning today, uh, but it's something that I'm feeling this morning as we're recording it. And if you're a marketer or if there's any kind of niche interest that you have, I want to inform you that there's YouTube content about it. There's been such a boom of creators there. So whether you want to dive deep into something you want to learn or, you know, the discourse around a particular discipline, like, you know, be it SEO, ppc, whichever, there's a ton about that. But also I found it really fulfilling to just look for obscure content on YouTube or at least content about like, obscure video games or stories or, you know, I'm even catching up on some series I've never seen through these summaries and reviews. And it's great. And I think it can also be part of our marketing strategy there too, where, you know, the algorithm there is great at surfacing other creators. So yeah, that's something that would also add to the mix. I honestly was not expecting our conversation to go this way today. Paris. We touched upon a lot of almost existential stuff, but I like that. I like the big talk.

Speaker A: Yeah, we had a, uh, we had a backup script that we looked at and I just have totally thrown that out the window. It's much better this way though. We just never got to it well. So, uh, Kuba, thank you again. And tell people listening. Where's the best place to find you and connect with you online?

Speaker B: The one best place to find me is definitely on LinkedIn. Jakub Greysar On LinkedIn, I post Semi regularly. I also do accept connection requests, DMs. If you want to have a conversation with me, which might lead to a piece of content. Who knows? Then definitely look me up. Right now, I'm the head of marketing at, uh, Talki AI, so particularly if you're in the healthcare space, I have a podcast of my own called Scaling Practice Management. You can go to Talki AI to find out more about that, about the podcast, and about what we do. So anybody from that kind of industry is especially encouraged to reach out. And I hope off the back of this conversation, I can make some new connections.

Speaker A: Go for it. Yeah, absolutely agree with you on YouTube. And overall, I think you've got some amazing original ideas for marketing. What you said at the beginning, I, uh, want to echo, which is that this is not the time to get out of marketing. This is the best time to be in marketing. And I hope that our listeners would agree with that and seize the market moment. So, Kuba, thank you for being with me today. I had a great conversation, and I'm looking forward to keeping in touch.

Speaker B: Yeah, thanks, Paris. And thank you. Washes and a snares all, uh, right, so long.