We Wrote About the Future of Work — Then Accidentally Lived It

This morning started with a very simple observation.

I can speak faster than I can type.

That led me into a conversation with ChatGPT about voice, Artificial Intelligence and the changing way we create knowledge.

We discussed something I had increasingly noticed about myself: when I speak rather than type, my thoughts often flow more naturally.

I can close my eyes. I can picture what I mean. I can tell the story. I can change direction. I can remember something halfway through and add it.

The technology can worry about turning those thoughts into structured words afterwards.

That conversation became an article: “Stop Typing. Start Talking: AI Has Changed How We Create.”

And then something rather extraordinary happened.

We Started Living the Article

We wanted to publish the article on JustLearn.

There was a problem. One of the tools we had been using to manage WordPress had reached its daily usage limit.

The message effectively said: Wait.

More capacity would become available many hours later.

That didn’t make much sense. We had the idea. We had the article. We had our own server. We had WordPress. We had n8n.

Why should somebody else’s usage allowance determine whether we could continue working?

So the conversation changed.

Instead of asking, “When can we use the tool again?” we asked: “Why do we need the tool?”

That question changed everything.

AI Suggested Another Route

WordPress has its own API. n8n can communicate with it.

So we created a dedicated WordPress Application Password for our publishing system. We connected n8n directly to JustLearn.

Connection successful.

Then we built a very simple workflow:

Trigger → Article Data → WordPress

The article title went in. The content went in. The status was deliberately set to Draft.

Then we pressed: Execute Workflow.

Green tick. Green tick. Green tick.

Workflow executed successfully.

WordPress Post ID 265 had been created.

I opened JustLearn. There was the article.

Correct title. Correct headings. Correct formatting. Correct quotation. Correct SmartAIMatters link. Correct AI disclosure.

We had just created our own AI-assisted publishing pipeline.

Then Came the Funny Part

Throughout this process ChatGPT had been telling me exactly where to click.

Open this screen. Choose this option. Enter this field. Create this node. Connect that node. Execute the workflow.

And eventually I asked the obvious question:

Why haven’t you simply taken control of n8n and done this yourself?

That led us to discuss a further possibility: delegating computer work to an AI agent.

At that point, I was still the person operating n8n. The question was whether more of those mechanical steps could be delegated, with the necessary access and a chance for me to review the result.

Suddenly the argument from the article had moved another step forward.

The original idea had been: Why type when I can speak?

Now the question was becoming: Why should I click when AI can operate the computer?

From Keyboard to Conversation to Delegation

Look at the progression.

For decades: Human → Keyboard → Computer
Then: Human → Touchscreen → Computer
Increasingly: Human → Voice → AI
The next step we wanted to explore: Human → Intent → AI Agent → Computer Systems

That is a much bigger change.

The interface is gradually moving away from operating software and towards expressing intention.

Instead of learning where every button lives, perhaps increasingly we will simply say what outcome we want.

“Publish this article.” “Update the spreadsheet.” “Check the website.” “Create the podcast version.” “Add this to NotebookLM.” “Turn the strongest section into three short videos.”

The computer still exists. The software still exists. The APIs still exist.

But the human may increasingly stop operating them directly.

We Had Accidentally Built the Demonstration

Earlier we had started discussing another article: “The Computer Isn’t Dying. It’s Disappearing.”

The argument was that computing might become so deeply embedded in our environment that we increasingly stop noticing the computer itself.

Wall displays. Voice. Vision. Gesture. Ambient AI. Agents operating software. No keyboard. Perhaps eventually very little mouse clicking.

Then I realised something.

We had spent the morning experiencing part of that transition.

A spoken thought became a conversation. The conversation became an article. The article became structured data. With my input, that data entered an automation. The automation communicated with WordPress. WordPress created the draft.

And the next stage we wanted to explore was an AI agent operating and extending the automation itself.

This wasn’t a presentation about digital transformation. It was digital transformation happening while we talked about it.

And That May Be the Real AI Revolution

People understandably focus on what Artificial Intelligence can generate.

Pictures. Text. Video. Music. Code.

But perhaps the more profound change is what happens when humans no longer need to understand every mechanical step between intention and outcome.

I don’t particularly want to become an expert in clicking n8n nodes. I want to understand what the system should accomplish.

That distinction matters.

My job increasingly becomes: Think. Question. Decide. Direct. Review.

AI and automation increasingly handle: Structure. Execute. Transfer. Transform. Repeat.

That doesn’t remove the human. It changes where the human contributes the greatest value.

And Tomorrow?

The publishing workflow we created today is only the beginning.

The plan is now:

Voice → AI conversation → JustLearn article → n8n → WordPress → NotebookLM → Podcast AIMatters → video → social media → learning resources

That is the intended direction. The downstream podcast, video and learning-resource stages are still work to build.

One original thought could become an entire learning ecosystem.

And perhaps the most remarkable thing is where it began.

Not with a programming project. Not with a technology strategy meeting. Not even with a keyboard.

It began with somebody talking.

Then I Came Back by Voice

Later, I opened a voice conversation and asked to continue the post we had been writing.

The assistant found the earlier conversation and identified the two articles: the original piece about speaking instead of typing, and this follow-up about what happened afterwards.

It asked which one I meant.

“The follow-up,” I said.

That small exchange belongs in the story.

I had returned to a piece of work simply by talking about it. The earlier draft became the starting point for the next part of the conversation.

There were still boundaries. Finding a draft was one step. Editing it, checking the claims and publishing it would each require further work. Our earlier WordPress draft had been created through an automation that I had helped set up and run.

But now I was using voice to pick up the thinking itself.

And that is what interests me most about voice-assisted content creation: being able to return to an unfinished thought, explain where I want to take it, and keep developing it through conversation.

This article is becoming a record of that experience, including the interruptions, the questions and the steps that still need a human decision.

JustLearn

Sometimes the best way to understand a technology isn’t to study it.

Use it.

Push against it. Find its limitations. Ask why something has to work the way it currently works.

And when somebody tells you to wait twelve hours because you’ve reached a usage limit, perhaps ask a different question:

“Is there another way?”

Today there was.

And in finding it, we accidentally experienced a little piece of the future we had been writing about.

Just Learn.