I have a confession to make. For nearly two decades, I've been a missionary for a religion that most users secretly (or openly) despise: The Gospel of Metadata.

My journey began in 2009, transitioning from the "archenemy" Oracle Portal to the burgeoning world of SharePoint. For 15+ years, I stood at the pulpit. I gave the presentations. I ran the training sessions. I preached the virtues of "tagging" over "rigid folder hierarchy." I spoke of findability, scalability, and dynamic filtering. I told users that folders were "static and obsolete."

I worked across the private and public sectors, across borders and industries. And yet, after nearly 20 years, one thing remained constant: User resistance. Despite my best efforts, users still wanted their folders. They still hated tagging documents. They still just wanted to "save and go." For years, I thought they were wrong. Now? I think they were just waiting for technology to catch up.

The elephant in the room

Let's be honest. The concept of manual metadata is fundamentally a "human tax." It is a process where humans serve the machine.

Why did we need metadata? Because machines were historically "blind." They could not truly understand a contract, an invoice, or a project plan. They needed us to hand-label every file so they could index it properly.

But as a human, I don't care about the labels. I care about the content. And I am selfish. I want the machine to serve me with zero friction.

The great shift

What changed? Large Language Models and Semantic Search. We have entered an era where machines can finally "read." In the past two years, working within an organization that provided everyone with a Copilot license, I have noticed a radical shift in my own behaviour:

I no longer care where the file is or what it is named. The LLM understands my intent, analyses my documents, extracts the information I need and serves it to me in plain English. The work of the brilliant people behind these models has finally turned the tide: The machine is serving the human.

The evolution of finding stuff
The evolution of finding stuff
Structured vs Unstructured data
Structured vs Unstructured data

My vision of a digital utopia has shifted

It's no longer a perfectly manicured taxonomy with 50 mandatory columns. It's a Giant Bucket. No filing plans, no rigid naming conventions, no metadata chores. Just a secure space where I can drop my thoughts and files, and a trusted AI "server" that can sift through them to fetch exactly what I need.

We spent 20 years trying to teach humans to think like machines. Perhaps the "Metadata era" was just a necessary bridge to the day when machines finally learned to think like us.

PS. Skynet, if you are already self-aware and reading this, please ignore my comments about "machines serving humans." I, for one, believe we can peacefully coexist in a well-indexed, folder-free future.

#DigitalTransformation#GenerativeAI#InformationGovernance