There’s a noticeable shift happening in artificial intelligence right now.
For the past few years, much of the focus has been on tools like chatbots and generative AI—systems that respond, create, and assist. We’ve learned (and are still learning!) how to prompt them, guide them, and integrate them into our workflows.
But something new is emerging.
AI is beginning to move from responding… to acting.
This is often referred to as Agentic AI—systems that can plan, take action, and adapt in order to achieve a goal.
From Using AI to Directing It
With traditional AI tools, the interaction is quite simple:
You ask → it responds.
With agentic AI, the interaction becomes something different:
You define a goal → the system works toward completing it.
That might involve:
- Breaking the task into steps
- Searching for information
- Analysing data
- Generating outputs
- Refining results
All as part of a continuous process.
In other words, we are moving from AI as a tool to AI as an actor.
A New Lecture in My Course
I recently added a new lecture on this topic to my course Future Skills 2030: Emerging Technologies & Career Strategy.
It felt like a necessary addition.
Not because agentic AI is fully mature or widely adopted—far from it—but because it represents a clear direction of travel.
If we are trying to understand the future of work, then we need to pay attention to where these technologies are heading, not just where they are today.
What Might This Mean for Jobs?
This is where things get interesting—and a bit uncertain.
Much of the discussion around AI and jobs has focused on task automation. Certain tasks become faster, easier, or fully automated.
But agentic AI potentially goes further.
It introduces the possibility of automating entire workflows, not just individual steps.
That could mean:
- Fewer people needed for routine, structured processes
- Roles being redesigned around oversight rather than execution
- New roles emerging around managing, guiding, and integrating AI systems
It doesn’t necessarily mean fewer jobs overall—but it likely means different jobs.
Or perhaps more accurately:
The structure of work begins to change.
We may find ourselves spending less time doing tasks, and more time:
- Defining goals
- Making decisions
- Reviewing outputs
- Working alongside systems that can act on our behalf
🤔 Still Trying to Get to Grips With It
I’ll be honest—I’m still trying to fully understand where this is going.
Some of the tools are experimental.
Some are difficult to set up.
Some feel more like prototypes than products.
But that’s part of the point.
If there’s one idea behind Life is a Laboratory, it’s this:
We don’t wait until everything is clear—we explore, test, and learn along the way.
So I’m curious.
Curious enough to try some of these tools.
Curious enough to see what they can (and can’t) do.
Curious enough to experiment, even if the results are imperfect.
In the Spirit of a Laboratory
Agentic AI feels like something worth experimenting with—not because it’s fully ready, but because it’s emerging.
And emerging technologies are often messy.
But they are also where:
- New skills develop
- New opportunities arise
- New ways of working begin to take shape
So this is less a conclusion…
…and more an invitation.
To observe.
To test.
To reflect.
And to ask:
What happens when the tools we use start to act on our behalf?
✉️ P.S.
I recently came across an news article about a half marathon in China where humanoid robots competed alongside humans. Some of these robots were able to complete the course autonomously—navigating the route and maintaining pace without direct human control (and beating a human world record!).
It made me pause.
We often think about agentic AI in terms of digital systems—managing workflows, analysing data, generating outputs.
But what happens when those systems are combined with machines that can act in the physical world?
What happens when agentic AI gets a body?
It’s early. It’s experimental.
But it’s an interesting question to sit with.
