How My Agent And I Stumbled Into AI Readiness

Written by Huibert Evekink | Sep 26, 2025 3:15:35 PM

Agents at your service

We’re entering the agentic wave of AI — a shift where humans and intelligent agents will increasingly work side by side. Forget about the mythical superintelligence, that’s post-bubble-bursting stuff. The real opportunity right now is learning to work alongside the AI systems we already have: imperfect, powerful, and increasingly pushed into making our daily workflows more efficient.

Most people aren’t ready for it. I wasn’t either.

The good news is that we are being bombarded with new tools promising to help you build your own AI agents — no technical background needed. These new “no-code” platforms (yeah, sure) may look magical, but they don’t run themselves, especially if you don’t have a technical background. What appears to be drag-and-drop simplicity quickly shows its complexity. To get it to work, you have to get your hands digitally dirty and learn how things actually operate: how inputs flow into decisions, how processes unfold, and how outputs get generated.

That might sound intimidating, but it’s actually a rare opportunity. Things we used to leave entirely to IT suddenly become visible. And that visibility creates the power to imagine and create.

I’ll wait for the plug-and-play version, thank you!

You might think you’ll wait until AI agents become “What you Say Is What You Get” simple. But here’s the problem: if you don’t understand how agents work — how workflows unfold, where decisions happen, how tools talk to each other — you won’t even know what’s possible, so forget about improving or innovating, especially when AI is involved.

The moment it did not click

This might all sound a bit abstract, so let me give you an example.

I was tired of manually scanning articles from my favorite publications every morning. So I decided to build an AI agent that would do it for me. Every day at 08:00, Manolo (name it to tame it) would log in, fetch articles, rank them based on my interests, summarize and send the top picks to my inbox, and save them to a website I could browse later.

I explained my brilliant news buddy idea to a slick “no-code” (yeah, sure) automation tool, and got a beautiful workflow that would need to be “configured”. Nodes, connectors, blocks; it looked, ehh, logical.

The actual connecting part turned into a personal tech marathon, a misogi. The agent had to handle scheduling, fetching data, applying logic, formatting content, and triggering emails. I had a faint idea how the necessary components communicated with each other. Now, I was wrestling with APIs, databases, and what I had desperately hoped to avoid: code I did not understand.

My no-code dream quickly turned into “vibe coding hell” — copying and pasting AI-generated code and praying it worked. Sometimes it did. More often, it merely shifted the problem further down the system, much like unclogging one intersection on a highway, only to cause a traffic jam somewhere else.

It was frustrating. I started doubting the whole thing; maybe I just wasn’t cut out for it.

The unexpected pay-off

In the end, I managed to get it to work, but it was hardly the super news Manolo I had set out to create.

However, I started to see the bigger picture: this wasn’t just a technical exercise.

If we want to build our own agents — and work effectively alongside them — we need to become systems thinkers. That doesn’t mean learning to code; it means learning to think like a builder:

Seeing how inputs, processes, and outputs connect
Understanding dependencies and sequencing
Mapping logic, flow, and feedback

Once I started thinking this way, something opened up. I could sketch workflows and describe problems and ideas more effectively to people who can actually code and create apps. I could collaborate more effectively and identify failure points more quickly.

I could design and explain the logic. That’s what mattered.

Try this at home

Think of a task you wish an AI agent could take off your plate. Something specific — like triaging emails, preparing meeting summaries and sending them automatically, or checking for references and red flags in a report.

Now go to a no-code platform like n8n or Zapier, prepare a cup of coffee, and try to build it even if you have no idea where to begin.

You’ll likely feel overwhelmed, confused, and frustrated. This is good; it means you’re learning and starting to see the real shape of the problem. You’re building AI resilience and systems fluency.

You don’t need to build the whole thing; start thinking in terms of steps:

What kicks it off?
What needs to happen next?
How does the information/data flow?
What’s the final output?

Even one or two steps wired together will teach you more than hours of reading.

Let’s be honest

We non-techies need to get out of our comfort zone and become true generalists, not coders, data experts, or prompt engineers, but people who understand how systems actually work.

That means getting more familiar with technology than we’re used to, and more than we probably like.

Hiding behind unique (human) skills and experiences won’t protect us. Strategy, finance, marketing, creativity, and people skills matter, but without systems fluency, they’ll make you more replaceable.

And yes, the tools will continue to get easier. But if you don’t understand how they work, you’re just dragging and dropping other people’s logic, not reinforcing your own.

And in the end, someone has to own the flow.
If it’s not you, it’s the agent or the person who built it.

Thanks for reading futurebraining! Subscribe for free to receive new posts and support our work.

View full post