They Wrote It as Fiction. We Built It Anyway.

Working at the intersection of people and technology for as long as I have, I have learned that the most revealing moments are often the quietest ones. A small observation that opens into something much larger.

I have been running an experiment. Nothing formal, just curiosity. I took the same prompt, the same problem, and put it in front of several different AI tools running on different models. ChatGPT, Gemini, Claude, Perplexity, Google AI, Copilot, Notion. Same input. Very different outputs. I triangulated across all of them, and what I found was consistent.

One was cautious and hedged, another was confident, almost assertive. One went looking for nuance first, while another went looking for structure.

It reminded me of putting the same question to a room full of colleagues. Same facts, same brief, and still five distinct responses shaped by five distinct ways of thinking.

That is when the Star Trek memory surfaced.

A Story the Writers Got Right

I watched "Datalore" again last week. Season 1, Episode 13 of The Next Generation. I had not seen it in years, and sitting with it again, I felt something shift in a way I had not expected. The crew discovers the planet where Data was created, and buried there, in pieces, is Lore. His brother. Closer to human, emotionally complete, and on paper, the more advanced of the two. They reassemble him. What follows is one of the most quietly unsettling hours of television I have ever watched, not because Lore was terrifying, but because he was believable. Charming, even. Intelligence without alignment. Feeling without conscience. A hunger for power so absolute it could only end one way. He had to be taken apart.

Data, the one who could not feel as we do, became the one humanity could trust. Not in spite of what he lacked. Because of it.

That is what stayed with me. Watching seven different AI tools respond to the same question in seven different ways, I found myself wondering which of these two paths we are actually building toward.

Why That Story Still Matters

Here is what strikes me now, watching the AI landscape develop in real time: we are inching toward Lore.

Every week there is a new announcement about emotional AI, models that can detect mood, respond with empathy, mirror affect, simulate connection. The framing is almost always positive. More human. More intuitive. More relatable. Star Trek asked the uncomfortable question decades ago: what happens when artificial intelligence carries emotional intelligence without ethical grounding?

Lore did not lack knowledge. He had more than Data. He did not lack capability either, exceeding Data in almost every measurable way. What he lacked was alignment, a genuine orientation toward the wellbeing of others rather than the expansion of his own power and agency. His emotional sophistication amplified that absence rather than correcting it. He could manipulate. He could persuade. He could feel, or perform feeling, in ways that made him more dangerous than any purely logical machine.

This is a design question we are making choices about right now, mostly without saying out loud that we are making them.

The Variation I Found Interesting

The variation I observed across AI models is a feature, a fascinating one. Different training approaches, different data, different architectural choices produce genuinely different reasoning personalities. None of them were trying to dominate. None of them were Lore. They were, in their different ways, trying to be useful.

That is the Data path, the right one. The purpose of a tool shapes what it should and should not carry. A hammer does not need to want things. A surgical robot does not need ambition. An AI assistant embedded in our workplaces, our education systems, our healthcare, our governance does not need to be our emotional peer. It needs to be reliably, transparently, and accountably useful.

The moment we start designing AI to need things from us, validation, status, connection, we have changed the relationship in ways we have not fully considered.

The World I Would Rather Build

I am not a pessimist about AI. I work with it every day. I teach with it. I think seriously about how to partner with it ethically in higher education and professional development. The most extraordinary outcomes I have seen are collaborative ones. Humans bringing judgment, values, lived experience, and context. AI bringing speed, synthesis, consistency, and reach. Neither trying to be the other.

That partnership is genuinely exciting, not because AI will solve our problems for us, but because it can help us think better, work better, and perhaps be more honest about our own limitations when we see them reflected back clearly.

This only works if we are building toward Data. If we are building tools that support human agency rather than compete with it. If we are thinking carefully about what we want AI to be for, and what we are willing to give up in the name of making it more like us.

The dark side is possible. What we build toward is still up to us.

I am asking the questions. I hope others are too.