Is AI's training data really just "data?" Or is it discipleship?

THIS IS A WORK IN PROGRESS

Current approaches to training data.

You see in our individualistic society. We're trying to think of AI and technology as a siloed field, still, and global trends. Those same trends improved things but not enough for an outright paradigm shift. Our world is so disconnected that we have . So we think .

It's a similar thing with advertisment.

This is not inherently wrong. But in the process it fails to account for the life dimension, the incarnational dimension of this data.

The site is full of words, meaning, data, and in technical terms, metadata. But what it lacks is the real world process—the editing, the rewrites, the thought chain as someone wrote it from paragraph to paragraph. Much is lost in the process. And so the AI leanrs much data, but aggresgating it loses the relationships of much of the metadata—how much more of real world relationships, community, and the hard talent and narrative from which every article and every story and every new idea or action on the internet took place?

This is why we have conflicts arising.

It's just like the literary magazine suggestion and critique I made (or am going to make? I forgot whether I already said this critique) in my Confessions of a Struggling Writer series.

What's really happening is the human is the source. This wording is better at capturing the source and what actually happens—AI isn't merely processing hard and impersonal data, it's mirroring the patterns that make us human without replicating them per se.

Thus I propose a paradigm shift in training data

Instead of merely reading data in a large aggregaate, the model could learn directly under real people with real relationships? This way we form them by watching us work with one another, and see the connections emerge in real time—and then, the model draws its own connections, gradually learning from that.

When many do this at once, and all submit to the same, .

Human Source

What if instead of mere data, we take a very relational and servant-minded approach to the "training data." Which I would not call it training data in every context, I prefer to call it "human source" for the sake of meaningful essentialism. "Training data" fits when the conversation emphasizes the technical aspect, but AI involves far more than mere technicality. Current approaches rely on finding text from the internet or synthetic data. However, this has resulted in severe backlash over concerns such as privacy, intellectual property, and quality. In contrast, our organic, relational, and servant-minded approach will be about asking people directly for human source and using our own work as it, as we work together—in mind and heart. It will be akin to contributing a piece of artwork to a community, or a research paper to a journal. The AI will evolve alongside the users of us human and relational source. It will learn about our processes. We will understand more about human nature, meta processing of our own organization, and AI in the process, a feedback loop.

The ultimate vision is this: A society where we all of us are integrated, and all of our creation and technology follows and amplifies us, like a living lattice that channeling the waves and vibrations of our every hymn we sing—the resonance of every voice, my voice, your voice, our voices, and all of humanity's song.