
In a world where work can start without you, what happens to your role, your career, and your sense of control? This episode explores the rise of Agentic AI: a new generation of artificial intelligence that doesn’t just respond to prompts, but anticipates goals, plans multi-step actions, and executes them before you even ask. We’ll look at how this technology is already reshaping workplaces, the opportunities it opens for career growth, the risks we need to manage, and a real-world case study that shows the future in action.
Listen to the podcast at our podcast host, Blubrry.com, or find it on your platform of choice, including iTunes, Spotify, Amazon, Audible, and YouTube.
Transcript
So…you wake up on a Tuesday morning. You’re halfway through your first coffee when a notification pops up on your phone. Your AI assistant has already accepted a meeting request from a client, rescheduled a conflicting internal meeting to next week, compiled some recent performance metrics into a tidy one-page dashboard, and has generated a personalized follow-up email for the person you spoke with yesterday. All ready for your approval. But you didn’t ask it to do any of this. It just… did it.
Welcome to a world where work can start without you.
This can sound either wonderfully empowering or mildly terrifying, depending on how much control you like to have. But the truth is this is no longer a fictional scenario. This is what is known as Agentic AI. A place whereAI doesn’t simply write emails and summarize documents, but proactively and intelligently structures your day for you. Many busy people have been looking forward to this for decades, and now it’s here. Be careful what you wish for. You might just get it.
From Generative to Agentic
Over the last couple of years, the world has quickly become accustomed to the idea of generative AI. Applications like ChatGPT, Perplexity and Copilot that create text, images, or code in response to prompts, are in heavy use around the world. You give these apps instructions, they generate and deliver what you asked for. Pretty straightforward. Their products are generated from a knowledge base of millions of documents they have read, they do some superfast research, wrap it up using the nuances of our spoken language and try to sound somewhat natural. It’s far more than the sterile listing of websites that Google has been serving up for a quarter century.
But now we are looking ahead to the integration of Agentic AI, and this is quite different. Where generative AI is reactive, acting on your instructions, agentic AI is proactive. Whereas generative AI works like a skilled assistant who sits quietly, waiting for you to issue some instructions, agentic AI is more like a colleague or partner who already knows the company’s goals, understands its priorities, and, before you’ve even poured your coffee, has started pulling together the tasks needed to get you there.
Agentic AI, it seems, can understand goals, can plan multi-step actions to achieve them, can work across different tools and systems to execute those actions, and can adjust its approach based on new late-breaking information, without being told to.
And that’s where the shift gets big and scary for a lot of people. We’re moving from technology as a tool, to technology as an active participant.
What’s Driving the Shift
Why is this even happening? Why now? There are a few converging factors:
First there is the constant pressure in organizations to boost productivity without adding headcount. Organizations continue to be squeezed by economic uncertainty, market competition, or talent shortages. So they are always looking for ways to do more with the same number of people, or fewer. It has always been this way. It is a constant within business.
But the new element in this mix is the advances in the abilities and the availabilities of large language models (LLMs) paired with APIs – the interfaces you see on your computer screen that you interact with. Agentic AI operates as a system that can not only understand complex instructions but can connect with dozens or even hundreds of other software tools to make things happen. The variations are already a little mind boggling: enterprise-scale orchestration platforms that chain together multiple AI actions into something that looks a lot like independent decision-making; customer service chatbots that can resolve issues without human intervention. Supply chain systems that predict shortages and order replacements. Finance tools that automatically rebalance portfolios in response to market shifts.
The pieces have been coming together for a while, but now they’re starting to click into place. It also opens up two parallel tracks of abject fear. First, once again, is this the end of our employment? And second what happens when we lose control of the process and hand it all over to a robot?
Opportunities for working people
I have spent much of my career teaching people who to make better use of their time, especially by prioritizing better, communicating better and handling mental blockers like procrastination. Now it may be possible for agentic AI to take care of much of the tedious tasks of the workday, those that are repetitive, time-consuming, and frankly, not the best use of anyone’s intelligence. That means tasks such as pulling together monthly reports, bouncing back and forth between people to set up a meeting, or rooting through different directories and files trying to find a single piece of information.
And perhaps most importantly, managing and completing certain workflows so that people do not need to bring it into the evening hours – their home life hours, or catching up on the weekend.
This puts a couple of movie images into my head. The first is a scene from a 1985 movie called Real Genius. You probably have not seen it, but it stars a young Val Kilmer as a student entering engineering college and struggling with the whole thing. But there’s a superb running joke in there concerning tedious lecture halls. The first of these shows a auditorium filled with students, trying to follow what their math professor is saying, and taking lots of diligent notes. As the classes wear on, one, student leaves a tape recorder in their seat in order to capture the lecture without having to be there. Next class, a couple more students do this as well, and the punch line is the day where Val Kilmer’s character comes to class and there is a tape recorder on every seat, and when he looks to the front of the class, the professor, too, has left a tape player on his desk. So the entire room consists of machines doing the work that the humans don’t want to do. Here’s a link to the scene on YouTube.
The second movie that comes to mind came out much earlier and that is Walt Disney’s Fantasia, specifically the Sorcerer’s Apprentice scene, in which Mickey Mouse, as the apprentice sneaks a peak at the Sorcerer’s spellbook and casts a spell on the broom, enabling it to sweep the floor by itself. Unfortunately, the plan goes sideways when Mickey tries to control the independent broom, and ends up simply making more brooms and losing all control. I’m sure you can see where I’m going with this.
There’s much to be said for technologies that can do the things that we don’t want to do, but that becomes a little more terrifying when we decide to hand over control.
The Risks and Challenges
Agentic AI looks like it will be doing more activities without your direct involvement. How do you feel about that? Do you feel that this will make you less connected to your own work? And that decisions will be made “for” you in ways that you can’t easily reverse? Do you feel there’s no control? I wonder if we have ever been down that road before. Does you car count? If you have a car that is 20 years old or less, it’s likely loaded with computers that will tell you when to do some sort of maintenance, and for those with assisted driving features like self-parking, does this enhance your ability as a driver, or does it take some control away from you?
To go down the fear road even further, what happens if an AI agent acts on its own and something goes wrong? Maybe it sends the wrong file to a client, violating some confidentiality rules. Who is going to take the blame for this? The human user? The developer? Your organization? What if hackers break in and start getting your agentic AI device to do things you don’t want it to – from insulting your boss, or transferring money out of your bank or credit card?
For agentic AI to work well, humans have to trust it. This means the AI must be transparent in what it’s doing and why. A black-box decision process is a recipe for suspicion and resistance. But is transparency enough? What if a threat actor told the agent to hide its tracks – do some bad stuff and then erase any evidence? What if the AI agent itself did this?
What about the skills gap? Workers who have built careers around doing tasks themselves will struggle to shift into an orchestration and oversight role. There’s a significant learning curve, and frankly, many people like doing these tasks themselves. It feels good to put something in order or to arrange a meeting or a trip. Why would you want to give that up?
And what about misuse by senior management? In the wrong hands, agentic AI could be a micromanagement tool at scale, tracking every keystroke, every pause, every delay. That’s not a future anyone should be excited about.
Shift your mindset
So, how do we prepare?
It’s best to start by recognizing that these images that I have are nightmare scenarios that spring from the same place in our minds that real nightmares come from. In anticipating this new world of agentic AI, we are trying to see into a future that hasn’t fully happened yet, and as always, fear becomes the dominant influencer. I has to be that way. That’s how the human body works. The best way to stay safe is to fear the unknown.
But there are a few realities we can factor in here, to balance out the fear. To start with, we have already been down this road so many times that it is simply an unacknowledged norm of our existence. Almost every week your phone, email application or operating system will undergo an automatic upgrade. Some of this is for urgent patches for bugs, but some, too are new features, or a rearrangement of buttons or menus, designed by some engineering team and then delivered, installed and activated regardless of how we may feel about that. Most recently, Windows users started to see little Copilot tags pretty much everywhere. A gentle nudge from the powers that be at Microsoft to start using their AI products.
The agentic experience has been around too, in the form of meeting arrangement techniques – searching calendars for a common time for a group of people to meet and then booking it. Most of the time a human is involved in this somewhere, but the fact is, things that affect your life – or maybe just your day – are happening around you.
We cope. We get used to it. It becomes the new normal.
Frankly I feel more concern for our experiences with algorithm-driven social media than I do for an AI agent that can help coordinate a person’s day. Streaming media like Tik Tok and Instagram offer small amounts of a pleasurable experience for free and then develop an ongoing dependence on it for more of the same. This to me is most profoundly symbolized in a crowd situation, whether it be a concert, or a fistfight in the street, in which everyone watching is holding their phone up at arm’s length to capture the scene – showing, in that single gesture – their belief that the memory of that handheld computer is more important than their own.
We need to shift our mindset from hapless victim to owner. Here’s another example. How many times have you seen commercials for cars – on TV or on social media – that always show that idealized driving situation – either racing along an empty coastal highway, or happily driving through empty city streets usually on the way to an exotic party? That’s nice, but it’s not how most of us will be using those cars. The reality is much different. They never show people navigating a crowded Walmart parking lot on a dreary February day.
The idealized image sells product. The reality of its usage is much different.
The tools inside an agentic AI program are designed help continue the tradition of the digital workplace, to help get things done more efficiently as and when you want. Think for example of a meeting or an activity that has a number of steps or tasks that need to be done before, during and after. A meeting is a good example – whether online or in real life, finding that common time, sending out calendar invites, sending an agenda, booking the room or online space, setting the agenda, preparing handout or support materials, then maybe scheduling preparation of a summary, transcript and action items based on discussions in the meeting and then maybe following up on these – scheduling later appointments and reviews. That’s a lot of stuff! And people who are good at this have been doing it with a checklist – manually entering these items into calendars and emails – work that takes up more time than the meeting itself, but which is never quite seen as the work that your job truly entails.
It would certainly be a great help to automate this process wouldn’t it? In fact some of it has already been automated – or agentified, if I may use that term. Zoom, Otter, and other meeting-related tools will provide a transcript, summary and a list of action items in a matter of minutes.
Regaining a sense of control
One of the healthiest states of mind that we humans can enjoy is when we feel we are in control of a situation. Not necessarily over other people but over the circumstances in our lives. Control creates a comfort that is the opposite of fear, and it is the manifestations of recognizing that the uncertainties that cause our fears have been tamed. That’s why vacations are so attractive: you go somewhere and pay money for the privilege of feeling completely in control of your time, at least for a while.
And so it will be with agentic AI. It is a tool that can do a great deal of the grunt work for you, but it does not have to become your overlord. You have the capacity to control how much of this behind-the-scenes organizing it does.
As has always the case, it is up to you to learn how to say “no.” (which, by the way, I will expand on in the next CoolTimeLife episode.)
Take, for example the hell that is back-to-back meetings, or, simply overloaded days. Our calendars get crammed to the point that there is no more space because our culture tends to believe that a crammed calendar means you are busy, and busy means you are productive. That has always been an incorrect assumption, on par with thinking you will travel farther and faster in your car if you do not waste time stopping off for refueling or recharging.
Even before the advent of agentic AI, people have been able to book meetings with you, either by scanning for spaces on your online calendar, or by simply bullying their way onto your diary. It has been up to us as individuals to push back – to add “unavailable” time blocks, and to negotiate rather than merely accept our fate.
So it will be with agentic AI tools: they can do many useful things for us, but we can and must remain in control.
Learn to set goals and parameters for AI clearly. Ambiguity can lead to unexpected, and sometimes unwanted, results. But when you understand the tasks and calculations it ca do for you, give the tool the leeway to perform this task.
Double down on human strengths like critical thinking, empathy, negotiation, and ethical judgment. These are the things that make you valuable in your organization. The agentic AI tool can help clear the way for more of that, again once you set its parameters
Get comfortable with AI oversight tools—dashboards, logs, and alerts that let you see what your AI is doing in real time. Understand what it is, how it works, and how it can work for you,.
Stay involved in AI pilot programs in your organization. Early adopters often become the go-to experts when a tool is rolled out company-wide.
Remember, it’s the feeling of control that is of most value here. The feeling of control floods your mind and body with positive chemistry, keeping stress and nervousness at bay and allowing for your creativity and excellence to push through. I have been teaching this concept for decades as part of my time management program, and always with the same bottom line: when you feel in control over your work and your time, you become the best professional person you can be.
So, let’s go back to that phrase: A world where work can start without you.
Agentic AI doesn’t make you irrelevant. Nor does it make you a servant to your computer. Instead, it enhances the way you do things.
Once you determine that certain repetitive, routine, and predictable work can be done before you even start your day, then your value shifts toward creativity, judgment, and leadership. But you make that decision in exactly the same way you decide whether or not to use the “Do Not Disturb” feature on your phone overnight. That’s an agentic act right there. Your phone silences calls and text messages during the hours you proscribe, except for the special people that you wish to allow through. That’s using an agent while still retaining control.
Thank you for visiting. Do you have comments or thoughts about this episode> Feel free to get in touch through our Contact page.
Leave a comment