Will AI Take Your Marketing Job? Here's What Two AI Experts Are Seeing

If you do it with a template today, a machine does it without you tomorrow. If you have 50 people in a role like social media marketer and you introduce generative AI tools that have very good capabilities, what happens is you just say, well, maybe we don't need 50 people home. We have 40 or 30 or 20 or 10 people who are now using machines capably, creating the same level of output at the same or higher level quality. Yes, AI is changing the way marketers work and the way they do their work. But at the same time, there are a lot of businesses that are using AI to, to do rapid growth Systems and process thinking is really, really valuable. AI is still changing so fast. There are a lot of people that use AI a lot, but they're using it the same way they did a year ago. There is critical thinking. Second, creative thinking, whoever has the most best ideas will win. And the third is contextual thinking. And this is Today I'm joined by two incredible guests who are on the front lines of AI and marketing. My first guest is an AI expert who helps businesses implement ai. He's the chief data scientist for Trust Insights, a company that provides AI consulting workshops and custom solutions. He's the author of the book, almost Timeless, the 48 Foundation Principles of Generative ai. Christopher Penn, welcome back to the show. How you doing Today? Thank You for having me. Awesome. My second guest is an AI strategist and founder of Divvy Up an agency that helps agencies develop AI operations. She's the founder of the AI Exchange, a membership for AI marketing ops professionals and consultants. Rachel Woods, welcome to this show. Thank you. I'm so excited for this conversation. Both of my guests have been speakers at our events trainers inside of our AI business Society, and we'll be speaking at AI business world in April, which is part of Social Media Marketing World 2026. Chris, let's start with you. Um, there's a lot of talk about AI taking marketing jobs. What's your opinion? Well, yes, of course it will. Um, we have, we already have good data about this. So, uh, there's a study from Stanford in Midsummer 2025 called Canaries in the Coal Mine that got updated when Anthropic released its work study, um, in November. So the current version of the Stanford studies now the November today it's called Canaries in the Coal Mine. Six Facts about Recent Employment Effects of Artificial Intelligence. What this shows unequivocally, especially in sales and marketing jobs, is that for certain groups of people, AI has a substantial negative effect on headcount for early career people in sales and marketing. 22 to 25 years old has a, it has caused a net loss of approximately 20% of head count in sales and marketing for that youngest cohort. And the higher up you go in seniority, the less of an effect there is. But there is an overall deflection effect of generative AI since, you know, uh, really January, 2023 when, when everyone figured out the chat GPT existed and the tap, it's not happening the way people think. People think, I'm gonna show up one day and there's be a terminator sitting at my desk typing, like, no, that's, that's not what's gonna happen. What happens is, particularly for larger companies, if you have 50 people in a role like social media marketer and you introduce generative AI tools that have very good capabilities, what happens is you just say, well, maybe we don't need 50 people. We have 40 or 30 or 20 or 10 people who are now using machines capably creating the same level of output at the same or higher level quality. And now you can say, I can do an 80% headcount reduction here, I'll give you one more anecdote. I was at an event, uh, last fall, two more notes, uh, of agency owners. One agency owner said, my client just came to me and said, I want an 80% fee reduction because you're using chat GPT, we're using chat GPT and we get a better result out of chat GPT than we get from you. So either you as a fee co or you're fired. The, uh, the agency owner was like, I don't know what to do. And it turns out not too long after that client fired that agency because they could not match what machines were capable. Second anecdote from, uh, a at a large industry conference, um, another agency owner came to me and said, I have my 2026 projections. I know where the market is going generally, I mean, as best as you can, I need, as he said, I need to cut 60% of my head cap. I need to have 60% for your employees all on the junior side by the end of the year or of 2026 or else I will not be profitable. Wow. Now, Rachel, you've been working with a lot of agencies and marketing teams as they're adopting ai. What are you seeing happen, uh, with the jobs front

and with people's roles? Yeah, I mean, I would echo a lot of what Chris is sharing around the narrative of cutting head count, looking at targets and saying, okay, great. AI is our way to hit those, you know, profit or margin goals. Um, I even have a similar anecdote of an, uh, actually on the brand side talking to the brand about, yeah, our biggest impact degenerative AI so far is that we fired our agency who was doing all of this work for us and now we're just using AI instead. So that's definitely happening. Um, but what we actually are seeing is there's also another side of the coin and, you know, we do work with agencies, but we typically try to really focus on agencies that are high growth agencies that are having actually a different problem where they can't find enough talent to meet the demand for their services, which those are great businesses for this time in ai, because guess what? AI can now take your team of 10. And instead of you need to add, you know, 50% or doubling your headcount this year, you can actually think about how do I enable my team of 10 that's already working? Great. And that is really interesting, um, for those roles. Then it becomes about how do you enable the people to think bigger picture, become managers of ai, think about how to systematize their work. And that is, I think, the flip side of, uh, what we're seeing with how it's impacting jobs. Interesting. Mike, do you remember back in social media marketing world 2015? Do you remember what I said on stage way back then? Uh, not exactly. I know you're on a panel. I think you were there with, uh, um, Sandy Carter from IBM Watson. Yep. Sandy and Gar and Guy Kawasaki and stuff. Um, said, if you do it with a template today, a machine does it without you tomorrow, and that was 2015, that was 11 years ago. When you look at so much social media marketing work, right? I have a Canva template for this. I have a, this an Instagram template, I have a TikTok template. Well, guess what? If you can use a template, so can generative ai, today's tools, uh, particularly the newest, uh, versions, and we're gonna talk about this at Social Media Marketing World 2026, um, like Claude Code and Claude Cowork, which are agents, you give them templates, you give them recipes, and you come back later and it's done. I was just getting ready for a workshop and I'm doing it a few weeks and I gave it my instructions and I said, I need you to generate synthetic data sets for this workshop, and it cranked out 135 synthetic data sets before lunch because it's all templated. I don't so explain what that is. People don't understand that, what you just said. So when I do a workshop, one of the, the, the issues people have is that their companies won't let them use their company data in the workshop for good reason, right? So we manufacture a, in this case I'm doing a fake hotel, um, in Alberta, Canada, and so they, you know, there's bookings, there's customer complaints, there's all this stuff, and general AI makes realistic data based on, you know, what it knows about the system so that workshop participants can actually use it and not have to use their own company data, which makes everybody happier. Okay. So, so far, here's what I've heard before I move on to the next question. Yes, AI is changing the way marketers work and the way they do their work. And at the junior level, if you don't have a lot of experience, those are some of the potential casualties in the bigger businesses. But at the same time also Rachel was saying that there are a lot of, um, businesses that are using a AI to, to do rapid growth, and part of the advantage to them is that they can do it at scale and economically if they understand how to do it. And what I wanna do is, um, I'll go to you, Rachel, can you give us an example either with a real client or, you know, you can, you don't have to say the real name, but how AI has transformed the day-to-day work of marketers. And then Chris, I'd love to hear also from you as well, just so people can understand at a practical level how in the world this stuff can really benefit us. Yeah, I mean I actually love the very clear analog of if you are working off a template today that a machine could do it tomorrow, because that's what we're seeing across the board with our clients. And when you realize that and lean into it, you could unlock a lot of advantages. Um, one of my favorite stories that comes to mind is we were working with an influencer marketing agency who, oh my gosh, I have not, I still to this day have not met a business that had so many documents in templates and SOPs or standard operating procedures already written down, but it was because they worked on pretty high profile influencer matching projects for big brands that had a lot of expectations and they had to have already gotten really detailed so that their team could run that service consistently. So going in pre ai, they already had all this, but that process whenever they're running it by hand, would take average two to three weeks for them to like put out an opportunity that brand wants to work on.

Um, they would get like thousands of applicants. They would've no way to review all those applicants. They had some basic filters they would try to do in Excel, but most of their vetting was they would open up their phone, go to the influencer that applies profile and swipe through the videos because they're also looking for like really detailed stuff. Like this brand does not, um, want to work with an influencer that has fast food on their, uh, profile 'cause isn't aligned with the ethos of our brand, all the stuff. So anyways, they had a really detailed process, but because they'd already had all the templates so to speak, or, uh, we call them playbooks, we were able to take that, teach it to ai, and then now they can run as many applicants as they want through this system. It takes about two hours to run huge batches of these applicants. So that means then they can go to their clients and say, great, we can get, uh, influencers for this campaign for you in two to three days. And so like they're scaling something that's already working, their team's now working on other pieces of the activation journey. Um, but it's just so interesting to see how if you have that template, you can really start to unlock a lot, which then the job becomes, okay, now what can we templatize next? Right. Oh, I absolutely love that. Chris, I know you've worked with some clients as well. I don't know if you have any examples or stories that you'd love to share, just so people can understand what could be unlocked when AI is applied to marketing actions and activities. I'll give you one from myself. So I write a newsletter, a weekly newsletter called the Almost Timely Newsletters. I've got like 300,000 subscribers. I have more subscribers that I like. I would be the 73rd largest city in the US if I was a city. I don't do a great job monetizing. So what I did was I sat down with Claude Code and I created four agents, a CEO agent, a CFO agent, a sales agent and a customer agent. And I said, here's my newsletter. It's about AI and analytics and stuff, and here's what I currently do to monetize it, which is like sell books and things like that and courses and things. I feel like I'm leaving money on the table. Let's review my strategy. And the four agents inside this environment all started arguing with each other about who was right. The voice of the customer is saying, no, if you do that, you're gonna p**s me off because that's not why I'm a subscriber. The CFO is saying, you can't, there's no market for this like that, that makes no money, no sense financially. And after 45 minutes of them arguing and me ref just refereeing, they came up with the unified strategy, tactics, execution and measurement plan for me for this year that says, you can go from this five figure amount of revenue to this seven figure amount in revenue because you are completely screwing up and just missing all these opportunities. And it, it gave me a, a laundry list. And now my next step is to say, okay, let's take which two agents. Um, I'm gonna have a, a sponsor agent. direct sales agent inside this environment. Say, start doing the, start drafting the pitches, integrate, here's my, here's the Gmail API, you can connect to my Gmail, start doing these things because I don't have, well, I'm unwilling to make the time to do all that grunt work, but I know I can make an agent that can go do it myself, do it for itself and I can supervise it. So that's a very concrete example of it's not just make me another blog post, right? We, that was 2023 today, it's, can I make a virtual workforce to what Rachel is saying, that makes 10 of me so that I don't have to do those things because it's not the best use of my time and I'm not very good at it anyway. Um, and, and be a force multiplier for me. Well, what I love about this is obviously all of us are running small businesses, right? We don't have unlimited resources. What I love about what you did there, Chris, is you developed AI agents that had, um, competency in areas where you didn't feel like you had competency. One in the financial side, one in the sales side, so on and so forth. And you had them interact with each other, which I think is really cool. Rachel, do you have any comments about what Chris was doing there? I actually have a very similar thing and I use it, um, in a, probably the most powerful way I use it is to coach me on skills that I'm trying to develop or that I feel like I don't have enough time to really do properly. A lot of things around, uh, management, leadership in my company strategy. Like I have agents set up that are basically there to be guardrails around how I am building that side of the company. Um, yeah, so like that entire way to think about AI as well of like, how do I build the workforce that's doing the things that I'm not even doing right now because I don't have time? That's a huge unlock as well. Okay. Um, I wanna start with skills. We'll go back to you Chris. Um, let's put ourselves, like for the listeners right now, which are mostly marketers, what are the skills that are more valuable for them, uh, in the workplace, uh, or the skills that will become more valuable that they ought to develop when it comes around ai? 'cause I just wanna like focus on like, hey, you could be more valued to your employer or your clients if you develop these kinds of skills.

So there's three of them, we call 'em the three Cs. There is critical thinking, AKA knowing when to call BS on the machines and say, I'm sorry, that is like, um, please use your fact check skill and fact check what you just said. Um, that's number one. But also at a higher level, being able to think, period. I have had several uncomfortable conversations today with, you know, uh, folks that I consult with who have relegated their thinking. So critical thinking is four pieces, right? Planning, organization, decision making, and problem solving. We, we call it pods. When you, when you are making decisions, your brain gets stronger. When you delegate those four skills to ai, your brain gets weaker. And so you have to be developing those four skills within the confines of critical thinking to be stronger. And the stronger you are, the harder you can push ai. Second creative thinking, whoever has the most best ideas will win, right? Because you AI takes skills and gives you minimum competency at the skills. Like, I have a musical bone in my body, but I can, if I have an idea, I can use AI to generate it with something like Suno and come up with an acceptable product. Does it go to win a Grammy? No. Uh, is it going to be better than what I would do? Yes. And the third is contextual thinking. And this is knowing what data you have, where it lives, and how to get it into a machine so that you can use it. Because the big unlock for a lot of companies is the proprietary data they have somewhere inside that nobody else has, that they can use with AI that will give them a better result. It's like having better ingredients. It's everybody has the same pancake mix, but you have better quality wheat, your pancakes will be better, all other things being equal. So critical thinking, creative thinking, contextual thinking. Yeah, I, I'm gonna get your thoughts on this in just a second, Rachel, but, um, what I love about contextual thinking is I, I kind of think of it as like institutional knowledge, right? There are people inside the company that know things that nobody else knows, and as a result of it, they understand the context of where to find the answers to these things right now. And that's important because AI is only as good as, um, the access to the data that, that you have, right? I also love critical thinking because I believe at a higher order, um, we're all called, called to be strategic, right? That's one of the reasons why marketers are effective is typically they, they're, you know, they have this strategic mind, but they're stuck doing things and they don't get to unlock that strategy thinking. And, and I love that. And that's where the critical side of it. And I also agree with you that creativity, many of us have this latent creative desire to like actually think creatively. And we are very creative people, but we've been bogged down in tasks. So I love your three C model. Um, Rachel, to you, what skills will make marketers more valuable right now in this world where AI is becoming kind of integrated into all that we do? I would echo those. I love the three Cs. I would also add into that I think systems and process thinking is really, really valuable, especially in the examples we were just talking about around having your workforce of agents. It's really easy to make a messy workforce that does nothing or does things you don't expect. And so having the ability to kind of think in systems and create mini systems for yourself, and then a lot of what we do as well is think about how do we create a system that a team can engage in and where people can still collaborate in a very like AI centric way. So systems and process thinking, I think it's a continuous journey. You can always get better at that. Um, really valuable. The other one I would say that as becoming more and more in the forefront for us is adaptability because, and maybe in a more specific way than people say potentially, but, um, AI is still changing so fast. And a trend that we're seeing is there are a lot of people that use AI a lot, but they're using it the same way they did a year ago, and they feel good or comfortable in that way, and they're actually not as open to adapting into new ways. And there's a lot of new ways on the marketing, especially the last couple of months. Um, so being able to have that mindset of adaptability and also being willing to stay, stay in the clunky side of things when things aren't quite perfect yet, that's gonna help you be really on like the cutting edge of what AI can do for you. Um, and so having those two mind systems seeking and adaptability is what I would add. Chris, I want to throw over the adaptability. Like what, what's gonna happen to marketers who don't adapt? Like if you look forward two to three years, the people that are stuck in today or yesterday's methodology, like what's at risk here for them, Update a LinkedIn profile. Feel free to Explain, I'm serious unemployment. It is unemployment because one of the, the, the side

of Rachel is exactly right in that these tools give you the ability to punch well above your weight. But that also means for a company like Trust Insights, for example, we're four people, right? We're four people and we, we do very well. We have honest conversations routinely about whether there will ever be a fifth person because the systems are moving so fast that we don't need to. So we are not laying people off, we're not hiring either because we don't need to. And as the tools get more and more capable and cheaper, that becomes a bigger and bigger question. So as you see companies find margins and profits and revenue from these automations. The analogy I use is this, 150 years ago, it took to, to harvest 30 acres of corn, took a team of 25 people for two and a half weeks, dawn to dusk, right? Um, to, to clear 30 acres, uh, backbreaking work, not great work, but it employed a lot of people today, two dudes driving a John Deere X nine 1100 can clear the same 30 acres in an hour and it is harvested, it is cut, and it is shelled and it's re it's ready for market, right? That same process took the entire winter 'cause people would harvest in the fall and then over the winter, like, you know, basically process the crop. Today it's done like 30 acres is done in an hour. So that same farm can now produce either much more acreage or it can be done with their 30 acres in an hour instead of two weeks. They don't hire the team at 25 anymore. So if you do not have these skills, if you don't know how to drive a John Deere X nine 1100, your employability to a farm today is basically very low. That's what it looks like in knowledge work. Rachel, uh, first of all, Chris, I love that. Uh, Rachel, where do you see things going if people don't adapt in the next few years? I mean, I would agree that, I mean, I'm, uh, I love Chris how, like, pointly you say it, I think I tend to be a little bit more like, come on, like you gotta take this seriously and here's all the, you know, here's the door, let's open it for you. Um, but I do think you're very, uh, correct. I think something else that we like to talk to teams about is, um, you, there's still the value of the specialist. And so I, I'm a believer that everybody needs to have kind of their own personal way. They're using ai, you know, whether it's your personal life, your work life, like you have your own personal productivity system, very similar. But then inside a team, something that we spend a lot of time on is letting people like shape their own roles and specialties within ai where there could be some people, for example, that are more that systems thinker. Um, I know Mike and I have talked about this before, and, uh, but it's the AI operator's, what we call them. Uh, so that's like a way that you can go down the path of being more operational with ai. You can also be somebody who's just like, Hey, I'm not gonna be the systems builder, but man, am I gonna push the limits of what AI can do in my creative domain? And I'm gonna be almost like this inventor of new techniques and new approaches for let's say how we do content. And that subject matter expert person could work with an AI operator to operationalize whatever they figure out. And then there's also like the leadership level of knowing where to even go with ai. What are we pushing towards? Like, you know, not all projects are created equal. So what are the things that we're actually investing in? Um, so back to the adapt question. I feel like it's not a, uh, if we should adapt, but the way we look at it is, you know, what direction are you going to adapt and how do you double down on what you're good At? Love this. Okay, so by the way, this is the reason we decided to announce AI Business World Conference, which you both will be, uh, speaking at, which is part of Social Media Marketing world this year, because we at Social Media Examiner, see what you see, which is like, there needs to be people training the world in this kind of stuff because the change is here and it's not optional. Um, what I'd love to do at this point is talk a little bit tactics. Chris, I'll start with you. Um, for people that maybe they're listening, they're like, okay, you've convinced me. I need to like, evolve my thinking. Um, what's something they can do right now, um, to begin, um, kind of maybe with a tool or something along those lines to just begin doing something to get them down the path to ultimately, uh, adapting, if you will, into this world. One of the, the best general meta skills to have is to be good at requirements gathering. Okay? So you're, let's, let's say you're gonna, you're gonna do something. Maybe you're gonna write a blog post, maybe you're gonna create a piece of software. Maybe you're gonna make a music video with dancing pigs, who cares? Um, you need to sit down and think through what goes into requirements. Kevin, what are your functional requirements, your non-functional requirements, your hazards, your risks, your user stories, all the things that go into, say, a product requirements document

or a business requirements document. Because guess what that is, that is a perfect prompt, right? That is here is exactly what I am trying to do and bring to life. And by the way, if I had to give people exactly one thing, one sentence that would make AI doubly good for them, no matter what, it would be their sentence, ask me questions until you have enough information to successfully complete the task. But at the end of every single prompt you ever do, and it will go from trying to decode what's on your brain to being a useful, helpful tool that says, oh, you want me to write a blog post? Who's it for? What, what's it about? Right? You have a lot of people who treat these things kind of like, you know, magic lamps, and they're not, they are prediction engines. The more data you give them to predict that's relevant, the better they predict. So if you just take that one sentence, ask me questions until you have enough information to successfully complete the task, you automatically double your capabilities because you're no longer asking the machine to guess. You are saying, tell me what you need and I will give it to you. And by the way, that prevents, well, it reduces executive function decay because, you know, saying, oh, just do the thing for me. You're saying ask me questions and it asks you questions. You're like, huh, I didn't think about that. Yeah. What do you mean executive function? It means it gets us using our human brains a little bit more. That's Yeah. Yeah, Yeah. Absolutely. Any organization decision making, problem solving? Uh, Rachel, what's something that people can do right now to, you know, effectively, um, uh, you know, begin, uh, adapting, if you will? So I think it depends in my mind on where you're at in the spectrum. If you still are not using ai, let's, yeah. We'll start with, yeah, we'll get to the more advanced stuff next. Okay. I say, if you're not using AI at least daily, like a really good checkpoint is you have the tab open in your computer every single day you're in that tab doing stuff. Most people I know have 10 tabs of chat PT open or whatever tool of their choice every single day. And they're popping, they're using it a lot, right? So if you're not there yet, I do think that one of the things that can be really helpful is being immersed in communities that are there so that then you have a little bit of both, like the social pressure, but also the, um, inspiration of how people are using it. Um, I met an amazing woman in Hawaii this summer who, you know, was crushing in her business, but really hadn't gotten into AI just as she joined a Facebook group that was kind of local to Hawaii and talked to her a month later and she was like, I could not believe how much she had been experimenting and learning and trying. And it was all fairly passive, right? Of like, Hey, let's just join a community and pay attention and start to try stuff. So if you're in that beginner state, that's what I would recommend. And then, yeah, there's, there's a lot more things that you can be doing to keep upping, uh, your question. Well, okay, let's, let's move on to the more advanced, uh, those, let, if you want, since I know you're ready to talk about it, Rachel, let's, to those that are already kind of doing what you're talking about, right? They're pretty active in ai, uh, and they're kind of using it every day. What's something that they could try to maybe like next level? So, uh, in our work, we call them AI playbooks, or basically, instead of giving AI one task at a time, you give it a series of tasks or series of steps. Um, but this has gotten so easy to do and start thinking about if you have used Claude. Um, there's Claude skills is a really easy way to basically give AI a step of ta a set of tasks that you want it to do claim Skills. 'cause my audience isn't gonna be familiar with that. Yeah. So, um, in Claude, there's basically a way that you can write a file, uh, like a document that gives an overview of what, uh, some task or process you want it to run is. And then you store it in the settings. So then whenever you are in any of your chat conversations, you can either directly reference it, like, I want you to use my, um, what would be good one, oh, I have a good one, which is a steelman skill, uh, which is basically arguing against me. Um, and then whenever you mention it, it will use that from its settings to apply that to your chat. Oh. So it's a way to make something, you know, make a process or a task you do frequently reusable, and then you can use it just all the time. Once you start doing that, I will warn you, becomes addictive and you're like, oh my gosh, I can make a skill for everything. And that's really where you can start to see the power of, like I mentioning earlier, the systems thinking and the process thinking to start to get AI to help you with a lot more things. Chris, any thoughts on what Rachel said and any other advanced things that, I mean, I know you can get really advanced, but keep it, keep it reasonable for this audience. I was gonna say, if you wanna go super advanced, um

if you are, if you consider yourself already pretty advanced, um, the, the place that I think people should be spending some time is in the coding tools. Not to write code, but because the coding tools have those skill in agent capabilities built in, so that's Claude code, that's GP open AI's Codex, that is Google Gemini code, uh, or Google's neutral anti-gravity, which might Rachel by the way, support skills. So you can forward 'em right from claw into anti-gravity, which is super awesome. And what these tools do is they allow you to do two things. One, they work with your file system, right? There is something incredible about a tool that has access to a folder on your computer. You tell it like, you can only use this folder, but it can then take notes. It can read and write to that folder, and it can have persistent memory. So a very generic example would be you turn on QU code, for example, and you say, here's this folder, let's have me start journaling, right? And you just dictate out what you're talking about and it writes your journal entries every single day. And then at the end of the week you can say, Hey Quinn, how was my week? And it can review the journal entries that it wrote for you and say, your week sucked. So, but that gives it also the ability to do, as Rachel was saying, very complex tasks because you write out a recipe, you say, Hey Quinn, I want you to do these 22 things, put it in plan mode. It writes a plan from your recipe, and then it just goes and does it and use, and you come back 45 minutes later and it's done. So you could say, I need to write the captions for this pile of Instagram videos for, for my company for this week. Here's my template, here's my brand voice, here's the restrictions, here's the recipe for how I want you to do this. Go do it and come back 45 minutes later and it's done. Chris. And if I were to, yeah, go ahead. If I just jump in giving another example. 'cause uh, I think this is so powerful. So if you're here, like go try these things. So in cloud code, one of the skills I use very regularly is I call it my today skill. And the first thing I do in the morning I open my computer is I just go into cloud code and I type today and it goes and looks at my calendar and pulls my, everything I'm doing for the day, it goes and checks clickup, which is what we use for task management internally pulls that it compares against yesterday's today in file. And then it helps me prioritize what I'm gonna work on in what order. And that's just one part of like a whole system I have for my personal productivity. But what's so powerful is how when you start creating these little things, it can build into a whole system helping you in your tools, like everything. So definitely go try that. Yeah. It's how assistant, Yeah. Uh, Claude came out with co-work, uh, within the last, since we're recording this in late January, uh, have either of you tried cowork and is it like more Chris, give us a little skinny on that. 'cause code scares the heck out of marketers, you know what I mean? 'cause it's code Co-work is basically clawed code without the coding portions and stuff. And it is a rapidly involving product. They have released up three releases in the last week alone. It is built, it was actually built by Claude Code, which is kind of funny. Um, and essentially it allows very similar to what, uh, Rachel and I are talking about. It allows you to use Claude as a co-working assistant. So you give it a folder that you want to work with on a desktop, you give it the skills that you want it to have. Um, I believe subagent actually, it has some subagents already built in. And, and you'll be able to add more. And you could do things like, you know what, here's an outline for talking. It'll be giving a social media marketing world turn to a PowerPoint. Uh, and it will pick up the, the template from your templates folder for your brand and stuff like that. It will pick up your brand style guidance, your brand guidelines folder and say, okay, here's your, here's the first draft of your talk for social media marketing world. Rachel, have you messed around with it at all? I have, and I do have a little bit of a warning story, uh, in a good way. But, so one of the, um, tools it has access to is it can open up your browser and actually go and scroll on websites and it like, takes screenshots and can navigate and all that stuff. So I was demoing it for a group and I was like, oh, let me go. I have a bunch of LinkedIn notifications, let me go and have it somewhere for those notifications and figure out what to respond to. And so it did that and it gave me a really clear like, Hey, this would be a really good one for you to respond to this person, you know, says a nice thing about your program goes to the page. I'm like, great. Draft a response, it drafts a response and then clicks send. And I didn't realize that if for like a couple minutes until somebody on my, uh, and the group was like, Hey, did you need for it to actually post the comment response? So just make sure you're watching your agent. That is my, uh, learning from that. And um, I wouldn't deleted it, but I hope the person who saw the response and because it's clearly not my, well, And we're going back to what you guys talked about earlier. It's really important to have critical thinking, especially when you are using the gentech ai. 'cause it's gonna go do things and it's just gonna do what it thinks is best. And this is gonna be where you're gonna get yourself in trouble, right? Like

Because it, like, so the, the genies, right, which comes from the gin, uh, the, the Arabic cultural concept. They give you your grant, you your wishes, right? So you have to get be real clear about what you wish for because they will take it quite literally and go and do the thing. If you say, you know, draft a response for my LinkedIn thing, the, these tools have reasoning capabilities, their reasoning chain will go, I've made a response for LinkedIn. LinkedIn is a place people go to post things. I shall post the thing, right? Unless you say draft it and do not send. Got it. Okay. So, uh, Rachel, um, do you have any like, um, some people, uh, there, well, a lot of us frankly feel this, like, I can't keep up. It's overwhelming because there's just so much happening. And, um, do you have any like rubric or concept that's, that we could apply to like, Hey, am I falling behind or not? Like, do you have, what do you tell people, because I'm sure you have people talk to you all the time that are freaking out that they can't keep up. What are some signs, legitimate signs that they might be falling behind or, so people wanna ask themselves? Yeah, go ahead. Yeah, I think it, it's probably unfortunately an ever-changing, you know, rule stick, right? Where the, as the market changes and evolves, if you felt like you were on top of it a year ago, I promise you aren't today if you haven't been keeping up with it, because there's so much that has changed again, even in just the last three to four months. One thing that I think is really allowed for, uh, our work right now is, uh, when we, you know, when I meet people that have a whole list of assumptions of what AI can't do, but then they haven't kind of created the time to dig in to understand why it can't do those things, then oftentimes they're missing like the next level of understanding, which is, oh, here's why I can't do those things. Well, if I adjust how I'm asking it to do that, it actually could do those things. And so then whenever you get through that breakthrough, then now you have a new set of capabilities at your disposal. A really good example of this I think is, um, math or really advanced reasoning and logic. If you feel like some of those tasks that you gave AI in the past or even now are too complicated, spending a little bit of time understanding why struggles with those and reshaping how you're structuring the task, it actually can do a lot of math and really complex reasoning and logic. So that'd be something I would pay attention to right now is if you feel like you, uh, see capability limits with AI and you don't really understand the next level, spend a little bit of time trying to dig into that. 'cause there's probably a, a workaround or a breakthrough, Chris, any warning signs or like rubrics that can help people know whether they legitimately are behind or not when it comes to all this innovation, Those tasks. So the, the specific tasks to look for, like, can it make a slide deck six months ago, the answer was kind of what was half-assed, right? Um, when Claude deployed the PowerPoint skill, suddenly it can now can make a slide deck, right? And Gemini can now make a slide deck. Uh, so that's a good indicator that of capability. So what are the things that it can't do right now? How is it progressing? So can it generate coherent audio and video together as a single video longer than eight seconds? Right now, the answer for most models is no, but you'll get to 30 seconds, 45, 60, et cetera, looking at major evals like GDP valve, for example, which is a economic task index that OpenAI came up with, which, uh, no surprise their models do great on. Um, but also having a set of benchmarks and tasks for yourself, right? To say like, here's the tasks that I do, how much of this can AI do? And the answer every few months should be more and more of it. Once like, Gentech stuff came out, for example, there's even more things that it can do than previously. Um, the big thing is having a system for where you get information that's specific to your needs, right? So what is, what are the, the needs that you need AI to, to work on? And do you have searches, do you have influencers? Do you have sources that you look at specifically say, okay, I care about AI that makes slide decks. Do you then go and see where those things are and how often you check? So you, the, to your, your question, your rubric for if you're falling behind is how often you're checking in, right? So for someone like me, I'm checking pretty much daily to see what's going on. Plus when there's a major conference like ICLR, uh, or NPS grabbing all five or 6,000 papers at a time and say, okay, which of these papers do I need to pay attention to? Because that's where AI's going to go in the next six months.

Wow. Okay. Well, um, social media marketing world slash AI business rule will be at the end of April. We're recording this January, so we're talking three months from now. But Rachel, what do you think everybody's gonna be talking about in three months from now? Because, you know, you talked about how the last three months have completely transformed. Where do you see everything going? What's gonna be kind of the, the buzz? What's your take on that? It's a great question. I think hard to predict, but I am really paying attention to right now the systematization of work. So all the stuff we're talking about around whatever agent tool, file system, skill system, et cetera, you're using, those capabilities are just going to continue to grow. And right now it's kind of, it's a lot easier if you're doing it as an individual. It's kind of a one player game. But I think that pulling that into the team and working as a team with a set of AI agent teammates that are genuinely helping coordinate across different functions, helping to manage teams, helping to be a mentor or coach to teams, apply consistency across work deliverables. I think all of that stuff is coming very, very fast. Um, and one of the thing that I would say around like keeping up, like with all of this is it is really easy to look at all of the tools and be like, oh my gosh, I need to spend all of my time looking at these tools, understanding them, testing them, chasing them. I subscribe to all the updates, et cetera. I do find in our work at least where we're really focused on getting AI to do real work, it should be like a 90 10, if not 95, 5% in terms of the time you're spending on your own problems, your own processes, how you're getting AI to do the work versus like trying to keep up with exactly where everything is going because there's just so much work to be done in structuring your systems and processes to even get to the place where AI could help. Like that's where businesses are way under investing right now is all that structuring. So, um, we have a, a saying that we use, which is, you know, own the work, own the playbooks, rent the tech, like the tech is gonna get better and better. You can just assume that, but if you don't have a good foundation of a business process operations to rely on, then uh, you're not able to use that capability. So that 90 10, 95 5 thing, you mean like five to 10% of your time should be keeping up with the tech, but the majority of your time should be working the tech? Is that what I'm hearing you say? Yeah, Because there's so much to do and actually just applying it to your use cases. Love it. Uh, Chris, three months from now, where do you see, what do you think the word on the street is gonna be? I mean, it's almost like we're taking years and we're compressing in the 90 days, but what's we are? Yeah. Talk to us about where you see it going in the next 90 days. Uh, it's all about agents. It is all about agents, uh, and the agent harnesses around them, uh, in particular the one. So there are five companies that I think are worth paying attention to, right? OpenAI, obviously Google, roic, Alibaba and Deep Seek are the five big companies that, uh, we're paying. You'll notice, by the way, two of them are from China. Uh, the biggest blind spot in AI today is that people are not paying attention to what's happening, particularly in China. I was at a conference room, in fact, you and I were at the same conference last fall and there were a lot of people talking about, oh, they open AI is doing this. I'm like, you guys are missing the boat. All of the major advances in AI are coming out of China. Um, last December, deep seek released a frightening, interesting model called Deep Seek Math. What it does is it solves math problems, but what it does, what they've trained it do is to solve math problems, then retrain itself on the math problems it solved and solve progressively harder math problems, is the first model of its kind. That is self-learning, like autonomously self-learning, because the secret to AI is it's all made of math. So whoever has the best math is going to win, and it is China. Um, Alibaba's Quinn models, Quin, the Quinn family is so incredibly powerful and compact that you can run them on your laptop. If you have a nice laptop and you can unplug the internet, turn off the wifi, and you have Gen AI capabilities that rival the best just from your machine, you pay nothing. It's completely private. Um, and it has all the things you would expect of a first class AI experience. So for the average user, pay attention to agents, particularly from the Big Five models, pay for the, you know, one of open ai, Claude or Gemini. I would personally recommend, uh, op, uh, Gemini and Claude as the two that I would pay for. Um, I don't think open AI is doing a particularly good job of keeping up, um, and pay very careful attention to what's happening, uh, out in, on the frontier out from Chinese labs. Yeah, and just so everybody knows, Claude is part of Anthropic, just in case you heard him say Anthropic and then Claude, Gemini

and Google, same, same company. Uh, and they've got a whole bunch of great stuff that they're doing. And then OpenAI is, uh, chat, GBT and all the Regulat related services. And then what are the actual products of the Chinese companies? If people wanna go check out those products, like at the, maybe at the more marketer level, not the technical level. Deep Seeq is obviously the na the company and their model name and they and their tool and Alibaba's is called Quinn, QWEN. Awesome. Okay, well first of all, if you want to meet Chris and Rachel, they're gonna be at AI Business World and Social Media Marketing World. They're very approachable. They hang out, they talk to people all the time. Um, Chris, let's start with you. If people wanna connect with you before, um, where do you, where do you want 'em to connect with you? What's your preferred social? And if they wanna work with you, where do you wanna send 'em? And then Rachel will ask you that question as well. Social media is in existential danger at the moment. Adam Moer was talking about how Instagram has basically become unusable because of AI slop. So the easiest way to find me is go to my website, trust insights. ai, and we'll let social media work itself out. Rachel, Um, love that. I would say LinkedIn. I'm still, uh, on the train of social media at the moment, but, um, yeah, I'd love to see be on LinkedIn. We have really a, a bunch of people who are passionate about this, you know, operationalizing AI that hang out and, um, talk on there. And then if you're interested in working with us, then uh, check out our website@divvyupagency. com. DIV up. DIVI, right up, yes. agency. com. Yeah. Rachel Woods. Chris Penn, thank you so much for joining us today and sharing your wisdom with us. Thanks for having us.