Hello, and welcome to another episode of the Working Conversations podcast, where we talk all things leadership, business, communication, and the future of work. I'm your host, Dr. Janel Anderson.

What if your technology tools didn't just wait for you to tell them what you wanted them to do, but they looked ahead, spotted what you'd need, and started working on it before you'd even asked? Imagine this. Your calendar manages itself, your client outreach happens on autopilot, your travel plans adapt to flight delays in real time, and you never have to lift a finger. Well, on the podcast today, we're diving into the fascinating and still largely misunderstood world of agentic AI. Now, I know a lot of you are already intrigued by artificial intelligence, and you've experimented with ChatGPT, Perplexity, Microsoft Copilot, and so on.

But agentic AI, where the AI is an autonomous agent acting on your behalf, that's a different beast altogether, and one that is absolutely poised to transform how we work, how we run our businesses, and even how we manage our daily lives. So let's break it down. First of all, what exactly is meant by agentic AI? Well, in simple terms, agentic AI refers to systems that don't just wait for your commands. Instead, they take initiative, setting and pursuing goals on their own, within boundaries that of course, you have defined for them. Now, here's the key distinction I want you to understand. When you chat with a website bot, say the one on your airline's website, you're interacting with a reactive system. You ask, what's the baggage fee? And it searches its scripts or its database and it returns the answer. The same with Siri or Alexa.

You ask them to play a song or tell you the weather, or set a reminder, and they respond to your prompt. That's called assistive AI or scripted automation. It's powerful, but it's ultimately waiting for you to act and tell it what to do. Now, when you ask chatgpt or Perplexity to take notes for you in a meeting, or help you summarize a complex article or draft or edit a document, you're using generative AI, which is creating or generating something based on the prompt that you gave it, again, it is reacting to the instructions that you give it. Agentic AI is different. It's proactive. It doesn't need you to nudge it every time. Instead, it identifies what needs to happen.

It plans multiple step actions, and it carries them out autonomously. Think of it like this. It's the difference between a helpful assistant that you have to give lots of instructions to for them to be useful versus one who is a superstar and knows your goals, notices when something's about to go off track and just handles it before you even realize there's an issue. For example, they see that someone's declining a meeting invitation because the time doesn't work for them. Your superstar assistant finds a time that's mutually available for you both and adeptly handles the rescheduling. And you didn't even know that any of that happened. It well, that's agentic AI. Now, we're not talking about science fiction here, but we are talking about tools that are already entering the market.

And for those of you who want to start experimenting, I'll share a list later on of the key players in this space and what they can already do for you and when you can expect them to do even more. Okay, so why does this matter? And how would I use this? Or am I using this in my business? So let me make this personal because when I started imagining how agentic AI could help me, the possibilities get really exciting. Now, as you know, I run a keynote speaking and podcasting business and I wear a lot of hats. I'm a researcher, a writer, a marketer, a scheduler, a traveler, a content creator, a relationship builder, a speaker from the stage. There's a lot of invisible work between the fun parts that you see here on the podcast or if you see me speak on stage that you see on stage. There's a lot of back end business that goes into making this whole operation work. So let me give you a couple of use cases where I could benefit tremendously from having a superstar agentic AI partner.

The first example I want to share is finding and pitching speaking engagements. So right now, finding the great next stage for me to speak on takes sometimes hours of research. Digging through conference websites, identifying the right fit, finding out who they had last year, learning where their next conference is and then checking my available availability for those dates and locations, figuring out who the conference planner is and what their contact information is, drafting an outreach email, and then when I get a hold of them, negotiating some basic details about the speaking engagement. Now imagine an agentic AI system that's doing that work for me. It's constantly monitoring conferences that I'm interested in, as well as events, calls for proposals and other speaking opportunities. It's checking my schedule when it finds one to see if I'm already booked or to see if I'm available to be in whatever city it is on the date of the upcoming event. Then it's drafting a personalized outreach message on my behalf. Then it's negotiating within my present parameters, like my available dates, my fee, the audience type that I speak best in front of. And then it is surfacing only the best fit hot leads for me to actually review and approve.

Now suddenly I'm not in the business development department anymore. I'm the strategic overseer. I'm truly in a leadership position here instead of doing the back end work myself. Then it gets even better. Agentic AI can manage the post contract touch points and make sure that no balls get dropped. Again, this is peeling back the covers a little bit and showing you what's under the hood. Once a speaking contract is signed, the real work begins. Sending marketing materials to the event planner, scheduling customization calls, shooting promo videos, drafting social media announcements, tracking deadlines, sending invoices, and so much more.

Now, if I was using agentic AI, I could have a system that automatically sends the right files at the right time. The conference planner needs my headshot. Boom. They get my headshot. Or maybe they want a full length body shot. They get that. Whichever one they want or need. They also send my bio.

Now, I was talking to a client the other day who asked for my bio and I said, well, what length would you like it to be? And she said, you know, and I said, do you want a long one, a short one? One that's 75 words, one that's 150 words. I said, because I've got it drafted already to meet your exact specifications. And if it's not meeting your exact specifications, I will redraft it to meet expectations and specifications. Because sometimes in a conference program, there is only a certain number of characters or words available for something like a speaker's bio. So again, agentic AI would be asking those questions on my behalf and then sending the exact right bio. Likewise with my introduction, because I do a customized introduction for every speech. So agentic AI could be handling all of those tasks and sending them at the exact right time.

Agentic AI could also schedule time on my calendar to shoot that promotional video for the conference, and then do the edits and send the final video to the conference planner. Like, how cool would that be?

Agentic AI could also coordinate the customization calls with the client, set those up on my calendar and the client's calendar so that we could have those calls without me having to juggle the calendar or my assistant juggle the calendar to figure those things out. Agentic AI would also be proactively researching the industry and the organization that I'm speaking for and prep talking points or questions that I should be asking during the prep call with my conference planner. Agentic AI would also do some of those basic backend things like ensuring invoices get sent to the right person at the right time. And agentic AI could even manage a timeline to make sure that nothing slips through the cracks. Because what I've talked about here are certainly a lot of the touch points, but not all of the touch points. Between signing the contract and taking the stage, there are so many little minutia. And for agentic AI to be handling all of those and scheduling all those would give me tremendous peace of mind.

And then of course, after the event is over, whether I have just left the stage and I'm still on site at the event, or if I have left the premises and I'm traveling back home, there are still so many follow up events that need to take place after the event in order to make sure everything comes to a close. So that just is the tip of the iceberg as it relates to a speaking event. So for a solo business owner like me, this isn't just a convenience, this is a total game changer to be using agentic AI.

Now let's step outside of my business world for a moment and look at another example that could be for any one of us. Travel agentic AI in our personal lives could be managing travel for our vacation or our business travel. Or even for, let's say if you're managing somebody else's travel, maybe you work in an administrative capacity and you're managing your boss's travel, or you've got other people in your family or your household who are traveling. So it could be, you know, again, personal or business. As we look at travel and anybody who travels with even irregularly, much less regularly, knows the stress of flight delays, misconnections and last minute rebookings.

So imagine this. Your agentic AI monitors your flights, your hotel booking, your car rental reservation, all in real time. So if a flight delay or cancellation pops up, it immediately searches for the best alternative flight, it rebooks you and then it adjusts your hotel arrival time or your car rental and anything else, any other touch points, and notifies all of the people who might be expecting you at a certain time. Now, tools like TripIt Pro do already give us a tiny taste of this. But true agentic AI will go far beyond alerts. It would solve the problem without having you to intervene at all. So I use Tripit. Tripit is a fantastic tool.

Every email confirmation that I get my flight, my hotel, car rental, whatever, all gets forwarded into TripIt. And TripIt pulls it all into this nice little itinerary wherever. Everything for a specific trip, including my parking at the Minneapolis St. Paul airport, all just gets coordinated into one place. So when I need the QR code to scan in for my prepaid parking place at the Minneapolis St. Paul Airport, I pull that up on TripIt. I put my phone up to the scanning machine at the airport as I'm pulling into the parking garage and it knows that I've already paid for my spot and the gate opens and all of those things. So TripIt already does much of this consolidation.

And TripIt Pro, the pro version also will notify you if there is a delay in your flight or a cancellation. And it gives you some options, but still you are the person who is selecting from the options that it's providing and you have a certain element of control as opposed to, again, if you go back to that like superstar administrative assistant that maybe you have, that person's rebooking you and then just letting you know that your flight time has changed, not asking for your permission. So TripIt Pro still asking for your permission, which again, some people absolutely want. Now you get a sense from that what agentic AI is from those couple of examples. Now we are much closer to having agentic AI working for us than most people really do realize. In fact, early versions are already here and we are in the experimental stages. But the pace of development is absolutely astonishing. Companies like OpenAI with its chat GPT, four agents and plugins, other companies are again already breaking out some of these technologies and allowing users to experiment with them.

Adept AI, hyperrite and Auto GPT are all working on agentic systems that can carry out these multiple step tasks, from managing research projects to coordinating calendar actions, booking haircuts for you and drafting and sending emails, doing all the things that are sometimes the little minutiae of administrative work that we get caught up in our daily lives that can take so much time. Now big players like Google and Microsoft are racing into this space too. And they are layering agentic capabilities onto products like Google Workspace and Microsoft 365 Copilot. And they're aiming to create AI that doesn't just answer your prompts, but proactively shapes your workflow and anticipates your needs and carries out things in advance. Now we are probably about a year to two, so 12 to 24 months away from seeing polished, enterprise ready agentic AI embedded in everyday workplace tools that large companies would be using now. Smaller companies like say mine can be early adopters in this space, particularly startups, tech forward industries and really small businesses like mine. Again, we are already experimenting with some of these agentic features already. Now let me give you the lowdown on the market leaders in this space so you can be watching for them.

And for those of you who are ready to experiment with them, you can start experimenting with them when the timing is right for you. Okay, the first one I want to mention is Auto GPT. This is an open source experimental project that builds autonomous GPT powered agents capable of planning and executing multiple step tasks. It's rough around the edges, but it has sparked huge excitement and a little bit of anxiety if we're being completely transparent here across tech circles, because it shows what's possible when you let AI start to act semi independently and autonomously.

All right, the next one I want to mention is Baby AGI. This is another open source project similar to Auto GPT. It's designed to let an AI autonomously manage tasks, create sub goals and work towards larger objectives with minimal human oversight. It's mostly used by developers and AI enthusiasts at this stage, but it is a peek into the near future landscape.

The next one I'll mention is hyperrite and hyperrite is the product name. It comes to us from the company Otherside AI. Now, Hyperrite's new personal assistant is in beta and it is designed as an agentic AI that doesn't just generate text, but helps you manage tasks across tools. It also helps you anticipate needs and take action. It takes actions on your behalf, such as drafting emails, filling out forms, scheduling meetings. Again some of that very administrative work that a superstar administrative assistant could do for you. Another tool is Adept AI and that comes to us. Adept AI is the company and the product is called Act One.

So Adept is building a general purpose. They think of it as an action transformer called Act One and that's AC that can interact with web browsers, enterprise software and digital tools on your behalf, effectively working as a digital co worker who can click, type and navigate apps just like a human would. Another tool that's out there is called Cognition AI and Cognition AI is the company and they recently announced a tool called Devin, an autonomous software engineer AI that can take high level engineering goals, write test code and write and test code like software and can even submit pull requests, essentially acting as an independent developer agent. And this is one of the most advanced examples of agentic AI that's happening right now in the professional domain.

Now, OpenAI, the company that brings us ChatGPT, also has custom GPTs and plugins and they're building out a whole ecosystem for this and they are soon to be expanding into and agents model. They are bringing agentic behaviors directly into the chat GPT experience, allowing people to create their own AI agents tailored to their own workflows, from automating research to completing tasks across connected tools. Now if you've ever used Zapier and written zaps, you could even have OpenAI's tools write their own zaps to connect via Zapier to other systems. So it's getting to be like, I mean again, it's the stuff of science fiction, but it's really, really happening.

Okay, Google also has Gemini. So its Gemini agents are going to roadmap out things and deploy agents that proactively assist users across the Google workspace, across search and cloud services, moving from reactive assistance like Bard, if you've used Bard on through Google or Google Assistant to something that can really take initiative and handle tasks more holistically. Again, think end to end workflows where there are multiple tasks or multiples steps along the way. We're getting there. And then Microsoft Copilot has now been rolling out autonomous agents in the office apps and this is just the beginning. The company's been teasing out future releases where Copilot becomes much more agentic, capable of again managing that multiple step workflow, coordinating across multiple systems, and even negotiating or deciding when to escalate tasks to the human users. Okay, so I know that was a lot to take in. So whether it's startups like Hyperrite, Adept or Cognition AI, or the giants like OpenAI's, GPT, Google, Microsoft and so on, the race to bring agentic AI into our everyday tools is already underway.

And it is so exciting. This isn't something that is merely on the horizon and it certainly isn't science fiction. It is arriving right now. All right, so now you have a really good sense of the landscape.

Let's look at some of the upside and downside of using agentic AI before we close up. So side first, what is the promise here? Well, offloading repetitive tasks and administrative headaches, all of those little things that take up so much time, answering the very straightforward emails, scheduling meetings, rescheduling things when things go wrong, or there's a conflict in in schedule and again sometimes taking in an inquiry, adding that data to this system, then sending a follow up email through that system and so on. Again these multi step and sometimes multi tool approaches. This is going to free up time for creative, strategic, human centered work.

The actual relationship building that I talk so much about on this podcast that is so important to our work lives and of course, our lives outside of work as well. Just imagine if you could free up a couple of hours of that administrative minutiae every single day so that you could have the networking coffees with people or go to a longer lunch with somebody. Go to lunch with somebody at all. Okay, so this is also going to allow small businesses and solo entrepreneurs, much like me to scale our businesses without adding headcount, which is super exciting. And the headcount that we do have can get deployed to be doing much more strategic things instead of some of the more rote things that they might be doing. It's going to create systems that are responsive, adaptive, and shockingly efficient. And as you think about how we might scale that out over larger enterprise organizations, it just is like absolutely astounding. So it is no exaggeration to say that agentic AI could literally change the way we think about running a business, managing a team, or even living our day to day lives.

Now, if any of you are a little bit scared about agentic AI running your life for you or managing your business for you, then yes, we need to address the downside too, because there are certainly risks and challenges. It is not all magic and upside here. So let's look at some of these challenges that this new technology presents. First of all, what if AI's idea of success or a task done to completion doesn't match yours? You need clear, well defined goals and parameters or you could risk the system optimizing for the wrong things. So if my flight gets delayed and I'm out of Minneapolis and it rebooks me on, let's say, an airline that I'm not confident in that doesn't have a great track record. And you know, I don't mean to throw Spirit Airlines here under the bus, so to speak, the proverbial bus, but I did once miss a speaking engagement because I was on Spirit and my flight was grounded in Los Angeles, Las Vegas, and I was leaving from an event in Las Vegas to go to a different speaking event and that was the only flight that was available and it was grounded overnight and I missed my speaking engagement. Now, the client was very gracious and was able to juggle things at the last second and move me into a different slot at that same event.

But it might not have worked out that way. So I don't know if I want agentic AI rebooking my flights. Now, when you set up agentic AI, you need to be really clear and specific about the rules and parameters that you give it so that it can take autonomous action on your behalf and make the right choices. All right, let's look then at another thing. What? Who's accountable when something goes wrong? If your agentic AI negotiates a deal or sends a message that either creates confusion or harm, then where does the responsibility land? Now, for a small business owner like myself, I'm going to say if I create agentic AI and it, let's say, negotiates the wrong fee on my behalf, it's my fault. It is within the purview of my business that that is my mistake, and I need to own that. But between me and AI, I might be a little bit frustrated and might be wanting to place the blame on AI, but I have to remember that I'm the one who created the parameters. I think where this gets to be even more complicated is when we think about large organizations and rolling out agentic AI on a more enterprise scale.

Then if I'm the employee and the agentic AI parameters were not necessarily set by me, but by somebody else in the organization, then exactly who is at fault? Is it a whole team? Is it a single person? And so on. So it does get murky, especially as the size of the organization grows. And third, and crucially, how much control are you willing to give up? Handing over autonomy means shifting from doing things to supervising. And not just supervising a human doing things, but supervising technology doing things for you. And not, not everyone is ready for that. I find it incredibly hard to hand over just simple airline booking to another human, to one of my assistants, because I am always navigating like, well, does my child have an event that day? And if so, can I make it to my child's event and then catch a flight? Was that too late of a flight for the night? Because I can't take the last flight of the night if my speaking engagement is first thing the next morning. So I have all these. This essentially an algorithm in my head that I'm going through.

So to translate that to an agentic AI system and have it take over right there requires some serious deep breaths. Okay, so this also all leads up to the idea of, will agentic AI really liberate us, or does it just create a new layer of labor for whoever is building out the agentic AI to do things for us? So, you know, where we're no longer doing the work, but we're managing and correcting machines, doing the work, which creates a whole other layer of work. Now, it might just be that that's an initial investment up front to create that layer of work, and then we can sit back and let agentic AI run itself and do those tasks for us.

All right, I know this episode's a lot to take in, my friends. So what should you be watching for? Well, I want you to start noticing terms like AI agents, autonomous AI or agentic AI or agentic systems. Watch for those in the workplace, watch for those in the tools that you're already using. So if you're using Microsoft 365 and you're using Copilot, watch for those things to start being introduced.

Maybe there's a beta system that you can try and just watch for these things to roll out. And ask yourself, where could proactive AI or agentic AI take some work off of your plate? And also ask yourself the flip side. What would you not feel comfortable handing over the control to agentic AI for? And remember, the key skill in this next era isn't just about knowing how to use the tech. It's learning how to set smart goals, learning how to provide ethical oversight, and learning how to collaborate effectively with these autonomous systems so that they can do the most for you without putting you at undue risk.

All right, the bottom line here, agentic AI marks a whole new phase in our relationship with technology. And with AI, it's not just a tool waiting for instructions. It's a collaborator, it's an actor.

It's even kind of a partner in us achieving our business goals or our personal goals. But with that comes responsibility. With great power, you know, it comes great responsibility. So as leaders, as entrepreneurs, as workers, we need to be ready to navigate this shift thoughtfully. We need to make sure that we're staying in the driver's seat even as the machines around us start to learn how to drive and start to learn how to take initiative. Again, a lot to take in on this episode. So if this episode sparked ideas or questions from you, I would love to hear them send me a message on LinkedIn or on any of the other social media channels that I'm on, or pop it over to my website and you can send me an email directly or a message directly through my website, but leave a comment wherever you're listening or if you're watching this on YouTube, by all means, let's have some dialogue there in the comments on YouTube.

And remember, the future of work is not waiting for us to tell it what to do. It's already moving. It's already moving and shaking and let's make sure that we are moving and shaking with it.

So my friends, as always, stay curious, stay informed and stay ahead of the curve and tune in next week for another insightful exploration of the trends shaping our professional world world.

Now if you learned something on this episode or you simply enjoy the content, please subscribe to my channel over on YouTube. Subscribe to the podcast on your podcast platform of choice. Follow me on social media. These are all excellent no cost ways for us to stay connected, stay engaged and for you to support me and my work. You find links to my social media over on the show notes page at janelanderson.com/215 for episode 215.

All right my friends go explore agentic AI and until next time keep thriving and keep working toward the future that we also desire. Take care my friends. Be well.