Bryan McAnulty [00:00:00]:

Would you hire a new employee, give them access to your entire business, and then let them work without really giving them clear instructions about what they're supposed to do or what you're trying to achieve?

Seth Earley [00:00:08]:

No. You don't ever let anybody loose on your business.

Bryan McAnulty [00:00:11]:

Many business owners are doing exactly that with their AI agents.

Seth Earley [00:00:14]:

And you would never let an AI agent loose on your business.

Bryan McAnulty [00:00:17]:

Today, we're interviewing Seth Earley, founder and CEO of Earley Information Science, a consulting firm that has helped Fortune 1000 companies organize and protect their business knowledge for more than 30 years. Seth is also the award-winning author of The AI-Powered Enterprise and the expert behind the philosophy, “There is no AI without IA.”

Seth Earley [00:00:34]:

The more you can describe it, the more you can structure it, the more you can componentize it into bite-sized chunks that answer specific questions, the better the AI will be able to retrieve it.

Bryan McAnulty [00:00:44]:

As AI agents become more capable of completing larger projects and taking real actions inside a business, Seth believes that they need clear knowledge, boundaries, and supervision.

Seth Earley [00:00:53]:

You're provisioning it with knowledge. You're uploading the documents that tell the system what to do.

Bryan McAnulty [00:00:58]:

So, what should you put in place before trusting an AI agent with your company? Let's find out.

AI is at the top of everyone's mind right now. But if you imagine a year or two from now where everybody's got access to the same AI agents, where we can just press a button and they do all the work for us, then there's an important question that we need to ask: what makes your business different or better than your competitors? I believe that it comes down to how you do things. I believe that in business, how we do things will be everything in this age of AI. So, I started asking founders this question, and something surprising happened. These conversations were so valuable that I knew I had to share them. And this is why I'm starting this new series inside The Creator's Adventure called Founder Stories.

Hey everyone, I'm Bryan McAnulty, the founder of Heights Platform. Let's get into it.

Bryan McAnulty [00:01:48]:

Hey Seth, welcome to the show.

Seth Earley [00:01:50]:

Hey Bryan, nice to be here. Thanks for having me.

Bryan McAnulty [00:01:53]:

Sure. My first question for you is, as we move into this world where AI agents are becoming more and more capable, if every business has access to the same agents, that we can all press this magic button and it does our work for us, then what makes your company different from the competitors out there? I think it's about how you do things. And so I'm curious what that looks like for you.

Seth Earley [00:02:15]:

Sure. We've been in business for three decades, so we have a lot of IP. We have a lot of work we've done for various types of clients — Fortune 500, Fortune 1000 organizations, Fortune 100 — across lots of different industries. We have methodologies that we've built. We have templates, approaches, and frameworks. And I've done a lot of writing.

One of the things I did was I ingested into a Claude project about a hundred articles that I've written over the years. And those are not just blog posts or trivial things. Those are things that went to publications, so they went through editing processes and vetting and approval and so on. And I use that to build a repository of my voice, tone, and style.

So that's one thing to think about if you're writing and creating content. The last thing you want to do is do a simple prompt and then have some AI spit out something that's obviously AI-generated, right? I have something called a “no AI slop” skill that I've built for my projects. Actually, I didn't build it. I downloaded it. There's probably a number of them out there. But what it does is it looks for the things that AI will typically generate, like the dichotomies or the reveals or the analogies, or the “it's not this, it's that,” all that kind of thing. “It's not this, it's that.” “Here's the load-bearing” or “the weight-bearing.” There's all sorts of things that it does that are the tells.

Now, you obviously want to let people know when you've used AI to generate stuff or used it to help you. But the way to make it your own voice is to write it yourself to begin with, use the tool to do some research and do an outline, and then even if you record your thoughts — even if you just look at that outline and start spitballing it, right? You can just start sharing, doing a brain dump, and ramble on about whatever it is. That is enough to be able to make it more about your voice than it is about the AI.

So the first thing to think about is, how do you make it your own? How do you make it based on what you know about the industry or about your customers or about your methods or about your tools? How you solve the problems? Because that's the big thing that you're going to be differentiating on. If everybody else is using the same tools that you are, what's different? What's different is how you solve the problem, how you go about your marketing, how you go about your messaging, how you interact with customers.

We've built a lot of frameworks around this. I have an AI readiness maturity evaluation. Well, that's built on many, many years of IP, which really talks about data readiness. It talks about knowledge readiness. It talks about operational readiness, technical readiness, governance readiness. Those are all things we've done over the years, and they're applied to AI. I've been in AI for a very long time. But the point here is that the agents that I build that evaluate these surveys and these questionnaires and these assessments are doing it based on repositories of knowledge that I provision into the AI.

So every Claude project I build has a provisioning — I have a provisioning folder on Dropbox, and I use a connector to Dropbox. And I basically say, if you're doing a taxonomy or information architecture initiative — and that may not mean as much to people, but it's structuring the data and the information for use in AI — here are the methodologies you need to use. Here are the templates. Here are the guidelines. Here are the best practices. And so what you're doing is you're provisioning it with knowledge. You're uploading the documents that tell the system what to do.

The other is building custom instructions. So, what do you want that project to do? What do you want that agent to do? What are its goals? What are its tasks? How do you want it to think about the problem and process the problem? And you can use the tools to help you build instructions. But the more information you're giving it, the more context you're giving it, the more of your thought process you're giving it, the better you will be able to differentiate.

So I think that's the biggest thing, is to say, where does your IP live? Where are your past projects? Where are your approaches? Where are you writing or speaking or doing podcasts or doing webinars, writing white papers? All of those things are to build your expertise and to leverage your expertise in these tools.

We're lucky that we have 30 years of IP, right? We have hundreds of webinars and podcasts and papers and approaches, huge repositories on SharePoint. So I'm able to go in and harvest all of that and then build best practices that I can then apply to projects. And what it does is it's allowed me to vastly accelerate the work that we've done.

The work that we would typically do for an organization — let's say it was a 12-month project in the past. Well, now it's a 12-week project. And so we don't charge hourly time and materials. We try to do a fixed price. And of course, we're giving the customer the benefit of all of that expertise and all of that efficiency, but they're not the projects they used to be. You don't need five people on a project or 10 people on a project. I did a project myself that in the past would have taken a team of three or four people about three months, and I did it myself in three weeks.

So that's where you're really getting the value. It's taking your ideas, your concepts, your approaches, and applying it to the problems, and giving that as context to these AI tools and agents and projects. Excuse me.

Bryan McAnulty [00:08:01]:

Yeah, that's a great point. And so I wonder, you help a lot of larger organizations and companies organizing their information. And with those companies, they have this information, but getting it in the right places and figuring out what's needed is a challenge in itself. For the smaller entrepreneurs and teams, maybe they haven't been at this quite as long as you have, and so the stuff is in their head at some level, but it might not be out there published yet. Do you have any suggestions on what are the best next steps for them to get this prepared for an agent or an AI process?

Bryan McAnulty [00:08:38]:

Real quick, Bryan here. You know me as the host for this podcast, but what you might not know is I'm also the founder of Heights Platform. It's an all-in-one platform that over 10,000 creators have used to build their online courses, communities, and digital product businesses. We recently added some awesome updates to Heights AI to help you turn your idea into a viable business. Heights AI can build entire product offers, review your content, and even coach you on how to grow. You can try it for free for 30 days. Links in the description. Now, back to the podcast.

Seth Earley [00:09:08]:

The important thing is to brainstorm on the process. And one of the things that you can do is you could have somebody interview you. If you have someone who can ask you questions: “Well, tell me, Bryan, how do you do your business? How do you onboard a customer? How do you begin the process? How do you gather requirements?” That's a human being doing that. But you can also have an agent or a large language model do that. You can tell it to interview you.

You say you want to capture your best practices. You want to capture approaches. What you'll do is you'll give it as much knowledge and information as you can to set up the context, but say, “Interview me so that I can write a course,” or “Interview me so that I can provision another system with the knowledge that is needed to solve these problems.” So it really comes down to getting it out of your head and getting it onto paper or getting it recorded so that the system can process that.

And interviewing customers, right? You can talk to people about how the project went. You can talk to them about the value that they got. You can talk to them about the approach that was used. If you're using any type of information-gathering instrument, whether it's a survey or whether it's an interview, whatever that might be, using those pieces of information and those data-gathering and information-gathering instruments is critical to what you're doing. So you want to ingest all of that.

But the way to do this is talk about the problems, talk about the solutions, talk about how you've done it before, talk about how you think you might want to do it. Talk about ways of potentially doing it. You don't have to have the answers. You really just have to have the questions. And then you can provision the AI with enough background, and of course have it do research. You can have it get references. You can have it find best practices. Those are all the things that are so valuable.

Think of it as a collaboration, right? You're collaborating with the AI. You're not just asking it to solve a problem or asking it for the answer. You're going back and forth with it. When I go through stuff with some of my tools, I might go through 14 or 15 iterations. I was working on a plan the other day, and I'm on version 16 because I'm going in and I'm making these subtle changes, and I'm giving it feedback and saying, “Don't say it this way,” or “Don't use that phrasing,” or “Don't use that terminology,” or you're pushing back.

The other thing you need to do is you need to constantly challenge the tool. You need to ask it questions. You need to ask it for justification. How did you arrive at that? Why do you think that? Why did you come to that conclusion? Can you give me some evidence? Can you give me an example? Can you find a reference? Can you find a citation? And those are the things that you would ask a consultant. Those are the things you might ask a customer. Those are things you might ask a partner, right? Think of it as your partner.

And thinking of this as a collaboration or a co-creation — it's a cognitive amplifier, right? You don't want to just use it to create something from whole cloth. You want to give it as much foundation and basis as possible so that it can really reflect your thinking, and then it amplifies your thinking. It's reflecting back to you your thinking.

That's what I find is so fascinating, is I can give it questions and problems, have it solve them, have it harvest stuff, have it research for me, and then I can go on to something else, and I can come back to it and say, “Oh wow, that's pretty cool, but what about this?” or “Here's some other evidence that you might want to consider,” or “Integrate this other white paper,” or “Here's an article I came across.” It's always a matter of capturing anything that you can that is relevant to your business, or relevant to the business that you want to be in and the ways that you want to solve problems.

And what's interesting is we built a tool methodology for assessing whether your content and your website is visible to AI agents and to AI tools for generative engine optimization. SEO is still important, but more and more research is being done by these agentic approaches and by large language models. So you need to think about your content and your website in a certain way to structure that content so it's visible.

We came up with this 10-dimension model. It's called the AI Retrieval Readiness Assessment, and it's based on all of my principles around information management, information architecture. All of our pillar content — some of our content runs to hundreds of pages — that's ingested into these tools. But what it does is it now will capture that information and it will structure content in such a way that it will be more visible to the AI engines.

Now, I'm always pushing back on that, and it's built a lot of tools for me. It's built the instructions. It's given me new knowledge to provision. I'm not a coder, but this tool allows me to build code so that I can achieve these objectives and I can solve these problems in ways that I never would have been able to before.

Bryan McAnulty [00:14:25]:

Yeah. So I want to get into a couple more details there, because you shared so much about how you're actually getting started with these things. I think way too many people make the mistake of wanting to get right to, “Hey, do my thing for me,” instead of this interview process, this information-gathering process.

And so I will tend to do also a combination of what you described myself. I will tell the agent, “Hey, you're going to be doing this kind of thing. Go out and find some research about this,” whether it has to be on the web or maybe it's in my own documents and my own stuff that I know it can already access. And then now, “Interview me and figure out about this.”

And the way I try to make it easy for myself is I have this nice professional mic here, but I also have these little magnet clip-on things, and I'll just wear this and walk around all day just talking to the thing. And then it transcribes that, and I'll use that to just answer it instead of having to sit there typing all day. And now, even if it's a stream-of-consciousness kind of thing and me blabbing about it, that's going to be a lot more detail than I would have otherwise maybe even thought to share with it. And now it's got these better ideas.

And then you talked about building all these versions. It's so cool that now sometimes I will know, okay, I don't have all the details here yet, but I will just say, “Okay, good. Just go do it, and I want to see what happens.” And that will sometimes even inform me. I know I'm going to throw it out, whatever it gave me. I know I'm going to throw it out, but that will help to inform me of where might I be missing something that I should have explained.

And then even when I know I have all the details and I have it working on it, it's such a different world now that before we would think, okay, well, I have something, I've got to continue with it. Now you can throw it out or just iterate on some of the versions and see where everything goes.

Seth Earley [00:16:14]:

Yeah, that's a great point, and it's a great way to do it because you want to tap into your creativity, your human knowledge, your perspective. The AI can appear to be creative, but it's not really creative. I mean, yes, it is inferring things. It is doing a lot of things that are very interesting and that emulate human creativity. But really, the creativity comes from you and your experience and your knowledge and your expertise and your approaches and what you've learned over the years — your tacit knowledge.

So the best way to communicate that tacit knowledge is to record your thoughts, right? Because you might be struck by something driving down the highway. Something might come to mind in the middle of the night. Something came to me the other day, and I thought, “Oh, this is great.” I was thinking that content chaos is a tax on large language models, on AI. It's a tax because when you have bad content, the model needs to iterate more on getting the answer, right? It takes more time. It takes more tokens.

It came to me in the evening one time, and I grabbed my phone and I just recorded my thoughts on that. And then I can process that and I can incorporate it into one of my white papers. So again, these are the ways that you can tap into your expertise. You know more than you can say, and you can say more than you can write, right? That's kind of the secret of tacit knowledge. You know more than you can say, and you can say more than you can write. So when you're thinking about these things and turning them over in your head, the best thing to do is just record, as you say, those thoughts and that brain dump.

Bryan McAnulty [00:17:59]:

Yeah. So before Earley Information Science existed, what frustrated you most about information architecture, and how does your company solve this problem? And maybe I should add that I saw that you often say, “There is no AI without IA.” So for those of our audience who are not as familiar with the term information architecture, maybe briefly get into that, and then about how your company is helping solve this.

Seth Earley [00:18:24]:

Sure. So think of it this way. If you were going to build a house, you would not start digging foundations and pouring concrete without a plan, right? You would have a plan. And you don't just have one plan that builds the whole house, that's for the whole house. You have a foundation plan. Then you have a framing plan. Then you have an HVAC plan. You have an electrical plan. You have a plumbing plan. You have an exterior plan. You have all sorts of different views on that house to build it, because there are many different components to it.

And you can think the same way of an information house. You need to build this structure to handle your information and manage your information, and it consists of lots of different things. People may have heard of taxonomies. Taxonomy is a way of categorizing and organizing information. Well, you can have navigational taxonomies that you might see on an e-commerce site. You can also have classification taxonomies that tell you the essence of a piece of content.

If I hand you a document, what is it? Well, maybe it's a statement of work or a proposal. Well, that's the “isness.” It is this thing that I can hand you. What is it about? Well, it's about this particular project. It's about this particular customer, about this industry. It's about this size project. What are the things — if I had a pile of a thousand of those proposals in a big pile, how do I tell them apart? What other piles could I put them in? That's called the “aboutness.”

So isness and aboutness are one of the core concepts of information architecture. How do you describe a thing? Then how do you tell it apart? What are the attributes? What are the hooks that you would — if you have this ocean of information, how do you grab onto a piece of information in that ocean? How do you tell it apart?

So if we're looking at products on a website, well, what is this thing? This is a product page. Well, what is it about? It's about a shirt. Well, what else is it about? Well, it's a large and a small and a medium, and these colors and these brands, and all those are aboutness. The isness is it's a product page. You could also say the isness is it's a shirt, but again, the aboutness are all the ways of distinguishing that. And what we want to do is we want to be able to structure all of the information in the organization in a way that describes it so you can retrieve it.

And by the way, this is also what retrieval-augmented generation does. When you think about getting a piece of content or generating a piece of content, you want to base it on a source of truth, right? Retrieval-augmented generation — generative AI will sometimes hallucinate. It'll make up answers. It's getting better, but it doesn't always use the ground truth. But it also doesn't know about your organization. It doesn't know about your products and services. Now, it may know about it if it's public, but many times the competitive value, competitive differentiation, is based on IP. That IP is internal to the organization.

So what you need to do is you need to say, here's the source of truth for this retrieval agent that's going to go in and search for that information. And the more you can describe it, the more you can structure it, the more you can componentize it into bite-sized chunks that answer specific questions, the better the AI will be able to retrieve it. So that is information architecture, and that's why it's so important.

Now, some people think, well, that's going to go away because the AI is getting better and better and better. It is getting better. You'll still need to tell it about what's important to you. You'll still need to tell it what's important to your customers. You still need to tell it how you do business. How do you solve problems? For highly technical organizations, it's about their tech docs and about their approaches to building these things, whatever they might be. But again, it's based on information and knowledge that's internal to the organization.

So when you are thinking about the architecture, it's structuring that information for retrieval, and a retrieval layer for any AI. Because at the end of the day, that's why it's called retrieval-augmented generation, right? It's retrieving. You're using the AI to understand the question that the user is asking. Then you're telling it, “Don't make up the answer. Go get the answer from this information source. And by the way, when you bring it back, now you can make it more conversational.”

So you're using the large language model to interpret the question. You're then saying, “Okay, let's go grab it from this information source.” And then we're making it more conversational and putting it into more understandable terms by presenting it back to the user, using the LLM for that.

Now again, there's a lot of complexity in there. There's a lot of nuance in there. There's a lot of issues around how these things perform, how many tokens they take. I mentioned that the tax on LLMs is chaos, right? Because they can do it, but it's going to take them a lot more turns. It's going to take a lot more queries. It's going to take more fine-tuning of those queries or re-ranking of those queries. So there are a lot of nuances to this. But at the end of the day, it's the structure and the quality of that data and that content and that knowledge that's going to determine how well your AI performs.

Bryan McAnulty [00:23:25]:

Yeah. So the way I like to think of that is, essentially, for everyone who has never heard these terms before and is less on the technical side of this, if you want your AI to help you with something or to help your customer with something, you can't just say, “Go do the thing.” It has to have access to that knowledge, or it's maybe going to make something up. Worst case, make something up. Maybe it will hopefully at least admit it doesn't know. Unfortunately, that doesn't happen all the time.

And so what you want to do is, as close to as possible, give the answer to it directly and then have it be able to repeat that. But of course, what the tools are able to do there is, if we're able to structure the data directly so that way it just has the answer there in front of it, that it's easy for it to find, it can now transform that answer into, as you said, the conversation of specifically what the person's asking for. And that's where the AI does really excel. If you say, “Hey, here's a couple things. The answer's in here. Here's what the person's asking.” Now transform that answer into the thing that they're looking for.

Seth Earley [00:24:34]:

Yeah. And you also have to tell it, if you don't have the answer from this source, say, “I don't know,” right? Because you don't want it to make it up. And so that's an important thing. It's called boundary adherence. And there's also groundedness and faithfulness and precision and recall, and adherence to escalations, hallucinations.

Many times you need a large language model to judge the results that another large language model is giving you, because many times it'll change the phrasing. So two answers may be equivalent, but they are phrased in different ways, or they may be phrased in different ways and they're not equivalent. So you actually need a judge, an LLM as a judge, to say whether those are the equivalent answers or not. But exactly as you're saying, you want to give it the knowledge grounding, the knowledge truth, the source of truth, so that it knows where to get that information. And that is that well-structured knowledge base.

Bryan McAnulty [00:25:29]:

Yeah. So when it comes to the founders out there, what do you think is valuable knowledge they already have, but they often seem to fail to capture it?

Seth Earley [00:25:41]:

Yeah. Well, one of the interesting things is that the things that are obvious to you are not obvious to other people. I find that very frequently when I think of something and I say something that is crystal clear to me, it may be completely opaque to somebody else, right? They may not understand it. So you have to realize that you have a lot more knowledge and context and experience and expertise than you realize. And the things that are obvious to you are not obvious to other people.

So many times it's having that beginner mind, right? To say, or to tell the AI to explain it like a person is five years old. “Here's my expertise. Here's my knowledge. Here's the thing I know. Let me explain it to you. Now explain it to somebody who has no experience in that field.” Explain it to a 10-year-old. Explain it to somebody who is completely non-technical, a complete businessperson. Maybe they don't even have an understanding of these basic concepts. And you'll find that it starts explaining things in ways that you really were not appreciating or not thinking about.

Many times I'll see people present content, and I'll be like, “Wow, that's a really interesting way of putting it.” I know those things, but I didn't really simplify it in that way, or I didn't take that person's perspective as a beginner. So it's having that beginner mind and saying, “How do I look at this with fresh eyes?” And that's very difficult to do when you've been in something very deeply for many years. So finding that way of explaining your knowledge to other people and telling stories about it, giving examples, and walking through exactly how the problem is solved, building analogies — those are all important ways of packaging your knowledge.

Bryan McAnulty [00:27:29]:

Yeah. And as you said earlier, one part is about your knowledge and information, but the other part is, what are the things that your potential customers are actually asking? Because often, due to the fact that we are in our own heads about all this knowledge that we have, and we're at such a different level sometimes than where our customer may be approaching the same kind of thing, especially if we're selling some kind of information product or service, we have to understand what is the way that they're communicating it in order to be able to have the right information to not only communicate back to them, but be able to digest it the right way with AI.

Seth Earley [00:28:11]:

Yeah. And I think that one of the things that is important is many times a customer will ask for something that is not really the thing they should be asking for, right? We used to have people have problems with their collaboration or workflow or how a system was designed, and they'd say it's a taxonomy problem. Well, actually, it was a search problem, or maybe it was a content problem. Taxonomy is one aspect of it, but they'll put a label on their problem that may not be the right label. They don't necessarily have an understanding of what this problem is.

Organizations are dealing with failed AI projects, right? Well, they're thinking that it's the model. Well, no, it's not the model. It's actually the information underneath the model, or it's the configuration of the search engine, or it's something along those lines. Many times people will say, “I have a problem with search.” No, you have a problem with content. No, you have a problem with architecture.

So customers don't necessarily know how to frame their problems because they don't necessarily know what that solution is or what that root cause is. So when you're explaining your solution, you have to translate it into language that the customer used that may not be accurate. It may not be appropriate. It may not be exactly what the thing is, but you have to connect those dots. As I say, the things that seem obvious are not obvious. And especially when you're looking at areas that are either new or emerging or changing very quickly, again, the customer isn't necessarily going to use the right language to describe that, but they'll describe what they perceive as their pain point. And their pain point may not actually be that pain point. There's a deeper root cause. So that's why it's important to take those perspectives.

Bryan McAnulty [00:30:01]:

Yeah. The phrase I like to use around this that I feel is kind of related is, never underestimate a customer's ability to misunderstand, which can always be another level. And what I want to get into now is kind of—

Seth Earley [00:30:15]:

I never underestimate my ability to confuse people. So I think I've been completely clear, right? I think I'm crystal clear because I understand it and I'm going on, and people are like, “What the heck is he talking about?”

Bryan McAnulty [00:30:28]:

It's an incredible skill to be able to communicate something in a very, very simple way.

Seth Earley [00:30:34]:

Yep.

Bryan McAnulty [00:30:35]:

What I want to get into now is a lightning round. So I'm looking for a quick take or 10-second answer on these next few. What is the biggest mistake businesses make when adopting AI agents?

Seth Earley [00:30:48]:

Not describing what they want, not being clear what the outcome is, and not measuring the baseline and then measuring what the outcome is. So not knowing, not being able to describe what good looks like.

Bryan McAnulty [00:31:01]:

Should you trust an AI agent with access to your entire business?

Seth Earley [00:31:04]:

No. No, because just like you wouldn't let an employee — would you let a brand-new employee have access to everything? Would you let them write checks and issue POs and purchase products? No. You would train them. You would be clear on what their duties are. You would describe their job. You would bound it. You would have points of escalation. And then when you trust them, you give them a little more. And then when you trust them with that, you give them a little more. You don't ever let anybody loose on your business, and you would never let an AI agent loose on your business.

Bryan McAnulty [00:31:34]:

Yeah. People are out here ready to make their AI agent the CEO right away. And arguably you could say it's even worse than that, because with an employee, at least there's the incentive that they don't want to get fired. But the AI, if you were ambiguous to the thing, it thinks, “I guess I'm doing what you asked,” and then it deletes all your stuff. Great.

Who owns the knowledge an AI agent learns about your company?

Seth Earley [00:31:59]:

Well, you would own it. Or in some organizations, they'll have people who are responsible for methodologies or approaches. They might be a director of delivery. There might be someone who is a project manager, program manager. But you really need to own your knowledge and be able to manage that knowledge — capture, manage, and evolve that knowledge.

Bryan McAnulty [00:32:28]:

Yeah, I agree. Would you say a company's knowledge is more valuable than its AI tools?

Seth Earley [00:32:35]:

Oh yeah, absolutely. That's what you're differentiating on. The AI tools and how you're using those AI tools, and how you're applying your knowledge to those AI tools or your AI tools to the knowledge.

Bryan McAnulty [00:32:47]:

And are most businesses ready for autonomous AI agents?

Seth Earley [00:32:51]:

No, because they don't understand their processes. You can't automate what you don't understand. And many times organizations don't necessarily understand the key processes. Once you map those out in detail and you understand exactly what needs to be done — many times people are running by the seat of their pants. They're like, “Oh, I just figure it out.” Well, you have to define the process in order for your agent to figure it out.

Bryan McAnulty [00:33:17]:

Yeah, completely agree. All right. And Seth, on this show, I'd like to have every guest ask a question to the audience. So if you could ask our audience anything, whether it's something you're curious about or just want to get people thinking about, what would that be?

Seth Earley [00:33:29]:

So I would say, ask yourself what your value proposition is, right? Your value may not be writing content, or it may not be the thing you think it is. You may be spending a lot of time on stuff that you think is valuable. Are your customers seeing it as value? Because you can cut out a lot of that busy work, or your customers can cut out a lot of the things they used to do.

I'm using Claude Design. Well, I don't necessarily need the graphic design person to use Claude Design. I can do it, right? But I still have to come up with the core ideas and the things that I want to communicate. So ask yourself what your core value is and how you're differentiating and how you do things differently than somebody else who might be in your exact industry, working with your exact customers, doing the exact same thing.

My wife is an EOS implementer, right? EOS — Entrepreneurial Operating System. There's a ton of implementers out there, but she's very unique in what she does. And the way she learns about that is by talking to her customers and saying, “Why did you choose me?” So I would say ask yourself, why should they choose you? And ask them, “Why did you choose me?” And how do you differentiate? What is your core value proposition?

Bryan McAnulty [00:34:41]:

Yeah, that's great. Seth, thanks so much for coming on the show. Before we get going, where else can people find you online?

Seth Earley [00:34:47]:

Well, you can connect with me on LinkedIn. I'm Seth Earley on LinkedIn. And you can also go to earley.com. That's Earley — don't forget the E before the Y — dot com. There's a podcast. We have the Earley AI Podcast. We have webinars. There's a lot of white papers and a lot of good stuff on our website. So you can go there and take a look. And again, follow me on LinkedIn and connect, and join the podcast as well. So I'll look forward to seeing people, and feel free to reach out. I'm happy to talk shop.

Bryan McAnulty [00:35:20]:

Awesome. Thanks.

Seth Earley [00:35:22]:

Thank you.

Bryan McAnulty [00:35:23]:

I'd like to take a moment to invite you to join our free community of over 5,000 creators at creatorclimb.com. If you enjoyed this episode and want to hear more, check out the Heights Platform YouTube channel every Tuesday at 9:00 a.m. US Central. To get notified when new episodes release, join our newsletter at thecreatorsadventure.com. Until then, keep learning, and I'll see you in the next episode.