Creating AI Agents: The Future of Work is Here

Happy 2026. Hey, how's this for a New Year's resolution? How about you Master AI marketing? The AI Business Society gives you live training with experts who are getting real results. No hype, no fluff, just proven workflows that work. And you've got access to an incredible library of training from many of the incredible guests that you've heard on this show. So secure your AI transformation by visiting socialmediaexaminer.com AI.

Welcome to the AI Explored podcast, helping you put AI to work. And now, here's your host, Michael Stelzner.

Hello, hello, hello. Thank you so much for joining me for the AI Explored podcast brought to you by Social Media Examiner. I'm your host, Michael Stelzer, and this is the podcast for marketers, creators and business owners who want to know how to use AI. Today, I'm going to be joined by Sean Kanungo, and we're going to explore how AI will impact the future of work. If you're new to this podcast, be sure to follow this show so you don't miss any of our future content. We've got some great stuff coming your way. Let's transition over to this week's interview with Sean Kanungo.

Helping you simplify your AI journey. Here is this week week's expert guide.

Today, I'm very excited to be joined by Sean Kanungo. If you don't know who Sean is, he's the author of the bold ones, Innovate and Disrupt to become truly indispensable. Sean is an innovation strategist and formerly worked for Deloitte. He's also host of the Dangerous Ideas podcast. Sean, welcome to the show. How you doing today?

Hey, I'm doing fantastic. Listen, I've been waiting my entire life to be on this podcast. And by the way, if you've gotten this far, like, rate review, subscribe to this pod because we need to make this pod the number one AI pod on the planet. Because you're doing a great job interviewing some incredible people, getting really tactical. It's been phenomenal. I'm just excited to be on pod.

Thank you for hyping me up there, Sean. Today, Sean and I are going to explore how AI is going to impact the future of work. And I'm really excited to go there. But before we go there, I want to hear your backstory.

Yeah.

How did you get into AI? Talk to us about what you did before AI, what's when your wake up moment was and then kind of where you're at today.

Yeah, absolutely. So I think A big part of my journey is I spent 12 years at Deloitte leading a lot of their strategy and innovation initiatives in Canada and around the world. And that means helping organizations set up innovation teams, thinking about what the future is going to look like for organizations implementing disruptive technologies. And my introduction to AI was around this novel technology that came out that we started really implementing with our clients in 2016, 2017, which was called robotic process automation. RPA.

Okay.

It was essentially like an Excel macro on steroids. It was like a dumb bot based on rules within certain frameworks that could automate humans, that could mimic humans. And I have never seen in my life something that could, would create so much ROI than rpa. I saw finance team scale up, I saw operations team scale up. Like, it was incredible to see. At the same time we were experimenting with IBM Watson to see how we could apply that to client problems. And so that was sort of my starting point, seeing the fact that this automation world was just at the starting point. Flash forward to where I am today, whether it's advising organizations on generative AI, diving deep and building my own agents, just seeing what's happening in the world. I mean, it's incredible. I'm excited. I believe that this revolution is a thousand x what the Internet is. You know, that's sort of been my journey.

That's so cool. Yeah. Sandy, oh my gosh, I'm drawing a blank on her last name. She was on the IBM Watson team. She's been on my other Web3 show. Sandy Carter was like the head of the Watson team. So just out of curiosity, back when you're dealing with this robotic stuff, you were talking about like, what was it doing? I mean, like we know what AI can do today, but was it analyzing spreadsheets? Is that essentially what it was doing back in the day?

Rpa, Robotic Process Automation, what it was able to do is essentially automate repetitives or manual tasks in a very rules based way. And so what it would do, it could credentialize into various systems, it could copy and paste information from one software to another software. It could automate some of the different workflows. It's essentially what you would think that a bot is today. It was just very, very rules based. So there were a couple different platforms that we used that are still relevant today. You know, Blue Prism, UiPath, who are all, you know, starting to get into this generative AI space. But yeah, that's basically what it could do.

So when did you decide that you wanted to go out on your own? And what was kind of the moment that, that sparked that?

Well, you know, to me, the reason why I left Deloitte was because I saw this power shift from institutions to individuals. I believe that individ now because of social media, because of the fact that, you know, one person can make an extraordinary impact. I just started to build my own power when I was in the firm. And part of the reason why I want to go off on my own is I believe that I could make a bigger impact in the world than actually working at the firm. So whether it's creating content or working with incredible companies or helping companies out, I, I've just felt that I could make a bigger impact. And so I left the firm about seven years ago and, and it's been an incredible journey, whether it's doing keynotes around the world, writing books, putting out content, advising organizations, when it comes to this new disruptive technologies. You know, that's what I've spent my last seven years doing. And I, I don't want to do anything for the rest of my life. This is just creating impact is what I wanted to do. And, and I believe that I'm, I'm, I'm living and, and doing that today. Like, I, I, I just wake up every single day just loving what I do.

Tell us a little bit about the, the premise of the book, the Bold Ones.

Yeah, so the Bold Ones is all about, it's, it's a playbook for individuals to go off and disrupt their own industries, their own careers. It's not necessarily built for the entrepreneur or the founder, but it's people that might be working for an organization. They just want to, they want to do more, they want to be more innovative, they want to be bolder. And, you know, I spent my entire career helping organizations disrupt themselves or disrupt their industries, but I really wanted individuals to do this and disrupt their own careers. And I felt like I was an example of somebody that did that. And so I want to give a playbook of how you can go off and do that. And the book highlights, you know, different people throughout history and also people in the modern culture that are really just being bold and innovative and really disruptive.

I love it. So bringing this now to today and artificial intelligence and really how AI will disrupt work and AI agents and all this kind of stuff. Just give us your kind of take on the why side of AI, why it's so important for marketers and entrepreneurs at this stage as we're entering into 2025.

I mean, this is the point of the podcast is that I believe that this revolution is a thousand x what the Internet is. This revolution is democratizing technology for everyone around the world. I believe that the most dangerous skill in the world today is just coming up with an idea. Just the ability to come up with an idea has become the most dangerous skill because what AI is doing it is democratizing technology. I just want to flash back to my history a little bit. While I was at the firm, I also had a mobile app development firm. So we created apps, mobile apps, a whole bunch of mobile apps on the app store. Some of those apps got tens of thousands of users. And when I went through that process, I'd had to do the ui, the ux, the marketing, the qa, testing, all the development, you know, either, you know, myself or outsourcing people, you know, using upwork and all these different sites. And today the ability to prototype, the ability to put something together has been radically reduced. Like it would have taken us a month or a year to produce things. And today you can produce things at a radically cheaper cost in days, in hours. And so it's really democratizing everything. And when it comes to marketing, you know, it's interesting. The role of the marketer is actually to be creative, it's to be innovative, it's to be authentic and to share authentic stories. It's a create trust at the end of the day. And so I believe that in the AI world actually marketing becomes incredibly more important. Like creative becomes the variable in a marketing world. Because in an AI world, when anybody can create content, where anybody can create a image or video or written content, some experts believe that the world will be inundated by AI content, that 99% of the content on the Internet by the end of 2030 will be just AI content because it's so easy to produce. And in that world, creativity becomes the variable. So for marketers, it's almost like a double edged sword, right? We need to be able to be proficient in the tools in order to make our lives more easier and seamless and remove those redundant, repetitive tasks. But at the same time, how do we stand out? How do we be unique? How do we be bold? And that's why I believe that. And not to, not to plug my book, but you know, that's the reason why being bold is going to be incredibly valuable just tactically today when it comes to AI, whether it's creating synthetic data to understand our customers in better ways, or if it's creating ads at scale so that we can test and optimize some of those Ads or whether it's optimizing what we're doing on social media or translation or production, there's so many ways that as marketers we can use this to simplify and optimize what we are doing.

I love what you're saying. I agree completely with what you're saying. And I know that there's still a large part of the marketing community that is resisting AI because for whatever reason it's complex or they don't understand how to use it, or they see it as a threat. And I want to double down on what you're saying because I believe that the two big unlocks for marketers are, number one, increasing their productivity, which makes them more valuable to their clients and or their employees. Lawyer. And number two, it gives them time to do deep work, time to actually think about things that they've never considered before, and it even enables deeper work. Right, because you can use these AI tools to essentially kind of noodle on your ideas where you might not have a sparring partner who's intellectually competent enough for you to to have these ideas. You can do this with AI and all of a sudden a mediocre marketer becomes a good marketer, and a good marketer becomes a great marketer, and a great marketer becomes an indispensable, ridiculously valuable marketer. Would you agree?

I 100% agree with that. A big part of what I use AI for is for inspiration. I think for marketers this can be a great source of inspiration. You know, there's this concept that people have been talking about for decades. I always hear somebody on stage shouting this out. The idea of begin with the end in mind. It's the idea that you should think about where you want to go and then solve problems to get to that particular end. But today I really believe that you can build with the end in mind. So what does that mean? That means that as a marketer, you can come up with a hundred different. If I need to create a website today, I'm like, let me just get a kid and pay him with some M&Ms. To generate a hundred different websites, different variations of what will that will look like. That becomes the starting point, the generation becomes the starting point. Then you go through your creativity, your creative processes, and then you ultimately get a final output or deliverable. You can do that for websites, for designs, for ad campaigns, like get the AI to come up with some of the inspiration. So then, then you go through your creative process and get a final output. So it's like building with the end in mind. And I believe that we are actually fundamentally rewiring how we look at work because of this new process. And I'm excited, I'm incredibly excited for marketers to leverage this. You know, a great example is Sora, or, you know, Google just came out with their VO version. Their competitor Sora. So if you don't know what Sora is. Sora is a text to video generator by OpenAI. I've been using a text to video generator for a while now known as Runway ML. And, you know, it's fine. Like, it's, it's, it's a great starting point. Are you going to use that in your creatives? Probably not, but at least it gives you a great storyboard to start with. It can give you some inspiration of how do I actually show people what I'm thinking? And I think what everyone's missing when it comes to this text to video or even some of these other platforms that are coming out. Like, we're actually creating a new communication layer because of some of these things that AI is producing. Like Sora, like Veo, like Runway. Somebody now in Bangladesh, in a third world country with very little resources, can now, now visualize an idea, an ad campaign, a movie for someone. And that would have taken. Nobody would have been able to do that. So I think it's incredibly underrated and under hyped around some of the advancements, you know, around what, what these platforms are doing. It's incredible. It's funny. Like, we're like, yeah, you know, it's not that great. But I, I feel like our expectations are so high that when something like Sora comes out that has figured out the science, the physics video is like one of the hardest things to replicate and they've been able to do it. Our expectations are too high. I don't know what you think, Michael.

Well, I'm going to say I think we're thinking too small. I'm going to tell you, like, and I like the concept of a title, the Bold Ones. I think Bold people think much bigger than regular people. Right. And I think that it's time for us to now have a much bigger vision than what we thought. Because I think most creatives have been kind of unable to be creative because they've been limited by budget or access to resources. And now what we're learning, right, is that that's kind of going away, right? Because we don't have to spend a lot of money to go out and hire people when we can do this stuff ourselves. So I believe that marketers and entrepreneurs need to think much bigger than what they're thinking because they don't yet understand how easy it's going to be to go from A to B. Would you agree with that?

You're totally right. I think with every technological revolution, we always think about the use cases.

We're the old tech, Right?

Exactly. I'll give an example of this. Like when films came out, the first thing that we shot were plays.

Right?

Like that was our context of what we should shoot. When the Internet really came out, newspapers, what they did is they copied and pasted the front page of the newspaper on a website. That was their mindset of what they could do. And I think that we're doing that right now. When it comes to AI, we haven't even unlocked what the possibilities, what the industries are going to look like. And I think it takes incredible entrepreneurs, it takes people that are bold to say, hey, now we have a. You know, just like we had the iPhone come out, we had all these mobile apps. Instead of having a calculator and a maps app and, and weather, you know, you needed bold entrepreneurs to say, yeah, you can create Airbnb, you can create Uber. This could be a entire business that's created on this platform. And I think that's, that's exciting. I, I would say that when it comes to AI, that this is the next, from a front facing perspective, a front customer facing perspective, this is the next paradigm shift. Like when we had websites, we had companies that come that were thinking, you know what we're hearing about this Internet, we should probably have a website. And when then mobile apps came out, we said, you know what, people talking about apps, maybe we should have an app. When social media really sort of came into existence, people said, you know what, we should probably get a Twitter account or a Facebook account or a LinkedIn account. And now I think with AI agents, it's going to be the new sort of platform. Companies are going to say, hey, I think we probably need an AI agent, a conversational agent that customers can interact with, because it's going to be a common way for people to interact with our brand or a company.

Yeah, that's a perfect transition into my next question, which is, why don't we define what an agent is? Because people might be thinking bots, which have been around for a very long time, and people in the industry that are technical use the word agentic a lot. Why don't you define what is an AI agent and what it can do today from your perspective, and then let's imagine what it might be able to do tomorrow.

Yeah. If you want to sound smart, you say agentic revolution. You say something agentic and people will be like, oh my God, is this person from France? Like, oh my goodness. Gentic revolution. This is nice.

Yeah. What is it? Define it to me.

An agent is mimicking humans. It is enabling your customers or employees to get answers, to solve problems, to take action through a natural conversational experience and doing it autonomously. That, to me, is what an AI agent is all about. It's simply just like how robotic process automation was mimicking humans. That's what an agent is essentially doing. An agent can be an external customer facing entity, or just like a human, it can be internal as well, automating existing processes. I actually think that the simplest way of describing it is an AI agent is mimicking humans. Within our organizations, we might have a sales development rep that can be now an AI agent. We also have an internal finance payroll agent that could be also an AI agent. So it's mimicking what humans can do. And that's always been what rpa. To me, the starting point of this entire revolution has always been part of. And everyone's going to be talking about AI agents in 2025. Every single massive platform is talking about it. Salesforce is like doubling down on Agent Force. I just like listen to a podcast of with Marc Benioff. The only thing he's talking about is Agent Force and how everything is about AI agents.

Yeah, he just hired 2,000 employees to sell it.

Right, There you go.

Just so people understand some of the stuff that's going on with AI agents, I just saw a website today where you can use AI actors to create videos for your ads. So instead of hiring an influencer, you can use AI actors, quote unquote, that are not real, that are showing off your product. That's an example. We at Social Media examiner have a customer service bot that lives on our social media marketing world website that's working 24 hours a day, gets trained every day on the latest stuff that's going on, and answers all of our prospect and customer service questions. You mentioned the finance side of things. Does that mean it's interacting with other customers, like in voice or in text? Or how does that work on that side of things? Because those are the only two examples I can think of that are here today. But how does that financial one you were talking about work? Is this like a someone sending an email and an agent responding to it internally inside the company or what is that financial one you were talking about.

So internally, it's not necessarily interacting with outside customers, but it might be taking an invoice, taking the documents of that invoice. Within most organizations, there's so much copying and pasting between different systems. An organization is basically built up of many different systems. And so what this AI agent or rpa, at the beginning point when I started, you know, this AI revolution, what it would do is it essentially would take a receipt, an invoice, it would be able to take that information, read the information from that invoice and take it into an Excel, take it actually into their various systems, whether they're using Oracle or some other system, and being able to put that information in. So that is essentially what that was able to do. And I see the next generation of that. There's an onslaught of companies that are helping organizations create things like that for almost every single role. You know, there are companies like 11X, they have, you know, two people that they call AI agents that will essentially, you know, be your sales development reps. There are, there are, there are companies that will be your vertical AI agent internally for internal sales or whatever it might be like. There's so many companies that are coming out that are creating their own digital workforce. And my belief is that the future of organizations is not only handling your human workforce, but it's also managing your AI workforce. And that governance and processes around that will get even more sophisticated over time.

Let's talk about filling the gap between AI potential and AI results. I'm sure you've tried ChatGPT. Are you seeing real ROI? Most marketers are barely scratching the surface of what's possible with AI. And it's because they're missing the training and the frameworks that will allow them to be very successful in their careers and in their future. AI Business World allows you to get that training. When you attend, you get two full days and 20 expert practitioners, many of the people that you've heard on this very show, that will teach you exactly what you need to know to build the workflows and to succeed as an AI enhanced marketer. It takes place April 29th and 30th in Anaheim, right alongside Social Media Marketing World. And here's the good news. You can save $300 when you register by January 16th. Head to AIbusinessWorld live right now. This discount ends soon. Yeah. And when we were prepping for this, you talked about what Claude was doing. Do you want to share what Claude recently did with the ability to take over a computer?

Yeah. So I Think what Claude has done. So, first of all, all these companies have come out with their versions of AI agents. So we talked about Salesforce, Microsoft, you know, OpenAI is going to be rolling out one Google. But what I think Claude has done with computer use is really remarkable. Where it is actually technology agnostic. For example, if you had an agent in Salesforce, that agent is specifically for that particular platform. But what computer use is is that it can actually take over your computer. Right now, the way to use it, you have to have a virtual environment. So I use it in Docker, which is where you can create a virtual.

Environment, a safe place, right?

Yeah, exactly. It's not great yet. It's not even almost usable yet. But you can see where they are going, where it will essentially take over your computer. It will start to mimic right now you can put in prompts and it can do stuff for you. Like, it can almost book a flight for you and it can go on to Safari and, you know, go to the United website and check out all the flights and it could almost do that. You can't like put in your credit card yet, but you can see a world where it's completely technology agnostic and it will take over your computer. It will start to understand what you do on your computer. And for example, if I need to edit this podcast, which by the way, you're going to be doing, but you know, maybe we'll go into final cut. It will understand how I edit my podcast, it will transcribe the podcast, it will upload this podcast onto Apple Music, you know, Spotify, YouTube. It will create the thumbnail because it will. It'll be trained on what we're doing and it'll be completely technology agnostic. That's where we're going.

That's really cool and scary at the same time, right? Absolutely scary, but cool. And.

And to be honest with you, I don't want to overhype where we are in this agentic revolution. See, I sound smart while I'm saying it. I don't want to overhype where we are right now. Like, I think it's going to be another five years until we get to that point.

Hey, but you never know. It's moving fast.

You know, it's moving fast. But right now, AI agents can do certain things that are not as complicated as going into different operating systems. And I think it's exciting.

Very cool. All right, so let's say that we want to start creating some sort of an AI agent. Where do we start? Like, what's the beginning point if we want to get into this world.

Yeah. So maybe I'll give you a personal use case and I'll give some guidelines on, you know, what would work. So to me, it's like I want to take the problems within my business, the challenges with my business that are the most annoying, the most repetitive, and the most frequent, identifying all those different challenges to see where I can automate some of those things. So one of the great AI agent platforms that I love is this site called Lindy.

Lindy AI L I N D Y AI.

Yeah. And so what you can do with Lindy is essentially create your own. I would say an AI agent is a stretch. They call it an agent, but you still have to like identify what the workflow is. So what are my workflows? Is I get an email. Once the email is in my inbox, it is enriched by an LLM, meaning that information is stored within a CRM. The LLM finds more information on the person that send me the email if it's someone that is not in the CRM and then it actually creates a draft email response. It doesn't send it out, but it creates a draft email response and it stores it in my sort of Gmail drafts. That's one way of automating, you know, one of my workflows or another one. It's, you know, it's a. It's a meeting note taker. So it's connected to my calendar. My meeting is transcribed into Zoom, it summarizes the conversation, and then it creates a draft of an email of the summaries of what that is. Now I'm just not comfortable yet with the AI sending things out for me. I just feel like I just, I'm still not comfortable I'm there yet. But what I'm doing in Lindy is I'm basically doing what I used to do at Deloitte, which is like I used to map out processes for organizations. Now I'm just taking my own process map and I'm saying, where can I add AI so that they can automate every part of this process? And my belief is that if you're an organization, an entrepreneur, you should take a look at all your processes and document it and see if where AI can be added in order to optimize what we're doing. And for me, like, we talk about all the exciting things about AI and the video and the, you know, the new industries. This is the most exciting part, which is how do I remove all the things that I don't want to do at work. I believe that not a single human should touch a CRM. It should be blasphemous. You should not touch a CRM in your entire life. And guess what? Not a single person that I've ever met enjoys entering information into a CRM or retrieving information from a CRM. Not a single person loves dealing with a ServiceNow or a Workday or an Oracle or an SAP. They hate it, but they have to do it because that's the source of information. My belief is that the AI should just be able to do that.

Fascinating. Okay, so we have had a number of people on the show that talk about AI automations utilizing things like make.com and Zapier. And the challenge with this is, despite the fact that make.com is very drag and drop and somewhat intuitive to use, it still requires a lot of technical prowess. Like, for example, we've got my director of tech innovation who's struggling, if you will, and even though he's highly technical, sometimes he just finds he's just going to go direct to the APIs, because these things are complicated and challenging, despite how people make it sound so simple. And what I like about what you're saying with Lindy AI is it sounds like either A, they've got out of the box integrations built in, or B, they make it really easy. Talk to me a little bit about that. Are you using out of the box kind of applications that Lindy does or are you building them yourself?

Yeah, it already has the APIs built in so that you can integrate. All you got to do is just, you know, connect the APIs within. I can't even remember what I did, but I. It's already integrated with all these different, you know, the common platforms that people are using. So it's just a matter of. And I'm not a super technical person. I'm. I'm relatively technical. I can code, but not like amazingly, yeah, it's just a matter of connecting those APIs, just like a Zapier. And it makes it simple. Now it's like a low form of AI agent. I'm talking about computer use and all these things that it can do. It's essentially taking a workflow, it's taking some rules, and it's automating against those particular rules. And the reason why you need those processes and rules now is that AI is actually not good without rules. And the reason why people don't implement AI is because it can hallucinate, it doesn't create, it doesn't provide the Most accurate information. And so that's what we learned with rpa is that actually RPA was very rule based and generative. AI unfortunately is not very good at rules because it can come up with almost anything. And so what I love about Lindy is that it actually defines the process for you so that you can get the most out of your LLMs in a defined process.

Yeah. And you know, you happen to have a special advantage because of your work you've done at Deloitte. You understand process and workflow and defining these things. That's already got it figured out is what you're saying.

Yeah, yeah, yeah. So it already has like a recipe for you. So if you don't have a process, it will give you the recipe for the process. I, I think they even call it recipes.

Okay. So they've got a little database of all sorts of recipes that you can modify. Is that what I'm hearing you say?

Exactly. By the way, that's the world. It's just like.

Yeah, I love that. But I want to you some, some questions because, you know, every business that's listening right now, they've got their own processes and give some tips on like, hey, you said like the thing that you want to do if you want to automate or create an AI agent, is to identify something that's really annoying or repetitive that you do regularly in your business that you would love to offload, if you will, to an AI agent. Once we've identified this, just give some tips to people who maybe aren't as good at breaking down the steps. Like how do we think, how do we need to reframe our mind to identify what would be the obvious steps? Because maybe there's things we do in our company that, that Lindy doesn't understand or any other thing doesn't understand. So do you have any tips on, like, here's some basic things you want to do, to document, to think about before you walk into some sort of an AI agent kind of thing.

Yeah. So number one, if you've already done the step in terms of identifying the smallest problem possible, that's the best point. Identify all your challenges. Identify one of the points that you want or the processes that you want to tackle. And then I would ask the AI to figure out the next point is ask the AI, where can I actually automate this exact process? Where do you think there are points where you can automate this process? It's like almost asking the AI for a source of inspiration of how you can go on tackle that particular process.

And any AI, probably ChatGPT or Claude could help you with this.

Right, exactly. And then what I would do is I would, using the smallest experiment possible, you don't even have to use your inbox. You don't have to, you can sign up for a, a separate CRM like a Salesforce, whatever, whatever it might be, so that it's the lowest risk possible. Because I, I, I really do believe that in order to experiment, you need to create an environment that's low risk so that you can convince other people that this will actually work. So what I like doing is creating a sandbox, a virtual environment where I can test this and start experimenting, get, get a couple people to say, how might we integrate some of these different sources of information together? Then we start tweaking, we start iterating, we automatically get into the testing piece. And as something like Lindy is like, it's so, so it's so intuitive, it's so easy that it's just a matter of starting to integrate those APIs, creating that sandbox and running, you know, a couple different tests. And once you have a test, then you can actually scale up. Not scale up, but now you can test it in a real environment within your business. And once you do that, you know, 10, 50, 100 times when you feel comfortable, then you can start operationalizing it, whether it's through your employees, your customers. And once you have something working there, then you can scale it up within your organization. So I think unfortunately it's a very experimental sort of model. It's, it's, you have to build small, but then, then you can scale up over time. That's how every automation process has ever been done, whether it's RPA to this AI agent revolution. You got to start in a sandbox.

I think you mentioned a sandbox tech. I don't know, I'm not sure. I'm sure a lot of my listeners have no idea how to do a sandbox, but is there some tech stack that makes that possible? Or is it just a matter of opening up a personal Gmail account and opening up a free account with some software? Or what are we talking about here?

When I say sandbox, to me, I'm just saying a safe space where you're not going to impact your core operations. Now that could be using different accounts, different CRMs, a different platforms that are not connected to your core operations. It could be a virtual environment like in Docker, where you're running simulations. It could be anything where it does not impact your core ops within every environment. That could be different.

Got it. Okay. So you mentioned a little bit of the risks on this front, but I'm sure people are curious about what are the risks of leveraging the AI agent. Presumably we've gotten to the point where we ran our experiments. We think it's intriguing. But you mentioned hallucinations, and I'd love you to expand a little bit more on what are some of the challenges that people might be facing at least today. What are your thoughts on that?

AI is not good without rules. And the problem with generative AI today is that there is still, I mean, we're getting a lot better at it, you know, over time, but right now it can still hallucinate. Some organizations have. There's a new phenomenon, new sort of invention that has been built post this generative AI revolution called RAG retrieval, Augmented generation, which is a technique that makes AI systems more reliable. It essentially retrieves data from your internal documents and sources, and then the LLM creates its prompts or creates its output based on that, those documents. And so very sophisticated organizations are implementing RAG on top of their LLMs to make sure that it hallucinates less, it breaks less, it's actually creating more accuracy because the problem with LLMs, it's not deterministic in a way. And so, so we're getting to a point where it's getting more accurate, it's getting more deterministic. Now that's for like very sophisticated organizations. I think some like, really interesting problems that I've seen is around just the governance, right? You don't want people across the organization, you know, having separate, I've seen this like separate ChatGPT accounts or even in the RPA world, like separate UI path path accounts or Blue PRISM accounts, it just gets very expensive. So having a good governance and policy around how you orchestrate all these agents across your organization, you're going to get great value out of doing like a good governance model around AI. Having one account or having a person sort of run those accounts is really important. I think what we should also be understanding when it comes to AI is that that with great power also comes, you know, bad actors, right? There's going to be people that will use AI for nefarious reasons. They, you know, the ability to create really sophisticated deep fakes, whether it's through audio, through video, the ability to create scams, the ability to, I, I can't even think about all the, the downsides. But I believe that 51% of technology is great and then 49% is terrible. And so that means that, that you know, there's going to be people that will use AI to penetrate your systems. And I think having a great cyber plan is like cyber is probably the most important, you know, scale outside of, you know, building a muscle around AI that in this particular space. So those are some things that I would sort of watch out for. But it should not prevent us from building a muscle around this. And we are definitely early in the AI agent revolution. But the best thing that we can do is to build a muscle around it, play with it. Try building an AI agent, whether through Lindy or there's platforms out there like rlit where you can create your own sort of personalized software. It's good to build a muscle around it because when you start building a muscle you actually understand the downside. You'll start to understand maybe even some of the, the downside implications of what people can do. And so that, that's what I would say into that.

Okay, imagine it's a year from now, it's January 2026 full years gone by since this interview. Where do you see AI agents in one year? What do you think they'll be able to do that they, they likely cannot do today?

I love this, I love this crystal ball question. So first of all, I think AI agents is going to be 2025. It's going to be the word of 2025. Every organization is going to be starting to talk about AI agents and at a very, I would say it's not going to do what computer use can do. It can take a look at, it can go between all these different systems, it can mimic everything that we can do. But I think every organization is going to hire some level of an AI agent, whether it is an AI agent that can optimize a very small problem within our organization or organization that will have an AI agent. From a front facing perspective, every organization is going to hear about an AI agent and they're going to hire at least one AI agent internally. And it's probably going to be for the smallest problem possible. And the reason why I'm not as like bullish to say, you know, there's going to be organizations, there's going to be one. I, I believe that the potential eventually is going to be you're going to have a one person, $1 billion organization because you're going to have an entire AI digital workforce at your fingertips. The reason why I'm not there yet is because I just think that the problem is still very hard to solve and I think we will need some Time to figure out the kinks to get to the other side. And I know I can see it by playing with Lindy, by building stuff in replit, by speaking to organizations that are building their own AI agents. It's like. Like it's almost 70% there, but the last 30%, there's still a lot of human intervention. There's still a lot of human oversight. And it's interesting because that last 30% is actually the hardest mile. Right. But I'm excited. I'm excited about the space.

Yeah. And I think that before we know it, it's going to be like C3PO, you know, we're going to have robots wandering around basically doing all sorts of fascinating things. Definitely in the next few years. I see it coming, by the way.

What do you think? I want to know. What do you.

Okay, well, here's my thoughts. 2026, I believe that customer service agents will be predominantly non human and they'll answer the phone instantly. I also believe that we're going to have doctors and nurses that are going to use Teladoc to begin to do initial diagnosis instantly. And it's going to allow us to reduce the cost of health care because we're going to have video and audio agents, if you will. That will essentially be the front line for, like, Teladoc, where we'll be able to call up, they'll be able to tell us what our symptoms are. This will cost almost nothing. And then doctors will only deal with patients that really need patients. I think that's all coming for sure in 2026.

Y. Wow.

And I think it's because the voice thing is so darn good. I mean, right now with OpenAI, their voice agent is human, like, and it sounds like a human. It's got almost no latency. It's blowing my mind. I'm sure it's blowing your mind as well. And, you know, the video stuff is coming for sure next year. I mean, that's been. That's the big frontier that they're solving for. Right.

Can I just talk about one of the things that we just talked about internally here, which is the AI revolution right now is remarkable. And I think one of the craziest things about the AI revolution is OpenAI had this thing, you know, their 12 Days of Christmas where they drop something every single day. And one of the most coolest things that I just saw was OpenAI introduced the video piece where you can open up the video and it can describe.

What it's looking like while you're talking to it. Right. Yeah, exactly.

Hey, what are you seeing right now? By me, it looks like you're in a well lit room with a camera set up behind you. There's also a bright light and what seems to be a plant in the corner. Does that sound right? Like the fact that you can turn on your video and it can describe everything around you, it can give you. What are we talking?

I'm like, this is why it's coming faster than you think, right? Because think about it, it's got audio and it's got visual, right? We're just one step away from injecting this into robotics, Right? Like, there's no question that if this thing can see and it can hear and it can speak, this is coming a lot faster than we think, my friend.

I agree. I think it's going to be fast and furious.

Is it's early days. It's early days. Sean, this has been amazing. Listen, if people want to connect with you, what's your preferred social. And then if they want to learn more about your speaking and all the other stuff you got going on, where do you want to send them?

Well, first of all, before you do that, I want you to rate, review, subscribe to this pod, whatever you got to do. We got to make this pod the number one AI podcast because Michael's putting in the work. So do that first. And if you have any time after that, I would just say you, you can follow me on LinkedIn, on YouTube. Just put in Sean Kanungo. I'm posting my latest thoughts and ideas and keynotes on the idea of AI and how it's going to transform the future. The punchline is I actually don't know what the future is going to look like. So hit me up and you can let me know what the futures look like. I'm always here to learn from others.

Yeah. And you can check him out also on his website by his same name, SeanCanungo.com, sean, thank you so much for, for coming on the show today.

Thank you. It was a pleasure. Appreciate it.

Hey, if you missed anything, we took all the notes for you over@social mediaexaminer.com A36. Be sure to follow this show on your favorite podcasting app. And if you've been a long time listener, would you give us a review on whatever platform you're listening on. Also, let your friends know about this show. I'm Stelzner on Facebook, Stelzner on LinkedIn and Ike Stelzner on X. Do check out our other shows. The social media Marketing Podcast and the Social Media Marketing talk show. This brings us to the end of the AI Explored podcast. I'm your host, Michael Stelzner. I'll be back with you next week. I hope you make the best out of your day and may AI help you become more successful.

The AI Explored podcast is a production of Social Media Examiner.

AI Business World, early bird pricing ends on January 16th. You can save $300 on your tickets. Register at AIBusinessWorld live today.