The Holodeck Was Never About the Holodeck

There’s a moment in Star Trek: The Next Generation that has stuck with me for thirty years.

Geordi La Forge steps onto the holodeck and says, more or less offhandedly, “Computer, create an adversary capable of defeating Data.” One sentence. And out of that sentence comes Victorian London — the fog, the gas lamps, the cobblestones, the accents, the right coats on the right people, a villain with a mind of his own. Nobody specified the buttons on the waistcoats. Nobody had to.

In another episode, Worf and his son end up in a Deadwood saloon, and the same thing happens. A few words, and an entire Old West town assembles itself, correct down to the boots.

I remember watching that as a younger man and thinking: that will never be possible.

But here’s the thing. I wasn’t thinking about the holodeck. The matter-energy conversion, the force fields, the photons made solid — all of that felt appropriately far away, the way warp drive feels far away. That wasn’t what struck me as impossible.

What struck me as impossible was the sentence.

The real science fiction was the interface

Every program I had ever written required me to say everything. Every variable declared. Every edge case handled. Every assumption made explicit, because the computer had no ability to fill in the blanks. The gap between “create an Old West town” and an actual Old West town was not a gap of processing power. It was a gap of shared context — a thousand unstated things about what saloons look like, how people in 1876 spoke, what a card game sounds like, what a sheriff wears.

You couldn’t specify that. There was too much of it. It was infinite in the direction that mattered.

So the holodeck’s real magic trick, the one that seemed permanently out of reach, was that the computer already knew all of that. You brought the idea. It brought the world.

That is roughly where we now find ourselves.

What I actually did with it

Two things over the past year convinced me this is not a demo, but a door.

The first was small and led somewhere I didn’t expect. I set out to generate some background images and ended up spending weeks exploring what happens when you cross a theme with an art style with a set of objects. I have no training as a visual artist. I have opinions, though — I know what I find interesting — and for the first time those opinions were enough to produce something I wanted to look at. The distance between having taste and making something had collapsed.

The second was larger, and more ambitious.

There’s a series of science fiction novels I love, and I had read all of them. What I wanted was another one. Not a summary, not an analysis — I wanted the experience of reading a new book in a world I already loved.

So I built a process to make one. I had AI develop plot arcs, characters, and settings. But I set one rule for myself: I wasn’t allowed to read any of it. I designed the machinery without looking at what came out of it. Chapter summaries, then chapters, then images, then a full audiobook.

And then I sat down and listened to a book I had caused to exist but had never read.

Was it as good as the original author? No. Not close. But it was good. The story held. There were turns I didn’t see coming — which was the entire point, and which is a strange and wonderful thing to say about something you commissioned. It felt like the same world and captured my imagination with the story. I wanted to know what would happen next.

I’ll add one detail that I think is the most honest part of the story: checking the result took twenty-five hours, because that’s how long it takes to listen to the book. The work didn’t vanish. It moved. It moved to the part I actually wanted to be doing.

What this frees, and for whom

For most of human history, the price of admission to any art form has been years — sometimes decades — of learning things that can be known. Perspective. Color theory. Meter. Period-accurate costume. Which lens. Which chord. How light behaves at dusk.

None of that is trivial, and none of it is fake knowledge. But almost all of it is knowledge that already exists somewhere. It is findable. It is, in the language of the holodeck, an assumption the computer can now fill in.

What was never findable is the idea. The situation you want to put people in. The character who won’t leave you alone. The twist that recontextualizes everything that came before it. The reason to tell the story at all.

I have always been a person who lives in that second category and gets stalled out by the first. Give me the shape of a story and I’ll think about it for weeks. Ask me what shade the sky is at that hour in that season in that hemisphere and I’ll be at the library for a month, and by then the idea has cooled.

So what’s actually happening here isn’t that creativity gets easier. It’s that the toll booth in front of it got cheaper. There are people walking around right now with something worth making inside them who will never spend fifteen years learning to render hands. That was a real loss, and it was a loss we had simply accepted as the cost of doing business.

On the holodeck, everyone on the ship gets to use it. Not just the few who can program it. That’s the part of the vision I care about.

What is actually lost

I want to be careful here, because there is a version of this essay that is a sales pitch, and I don’t believe the sales pitch.

Something real does go away.

I see it already with software developers. There’s a mourning happening — quiet, mostly unspoken, for the craft of writing all the small pieces yourself. And I don’t think that mourning is sentimental nonsense. Doing the work changes you. Sitting with a hard problem for a week and coming out the other side makes you into someone who can sit with hard problems. Ten thousand hours at a piano doesn’t just produce recordings; it produces a person who hears differently.

AI can get you to a product resembling what an expert would make. It cannot make you into the expert. The transformation isn’t in the output — it’s in the doing, and if you skip the doing, you skip the transformation. I got a book. I did not become a novelist. I’m at peace with that trade, but I want to be honest that it is a trade.

And there’s a second loss, subtler. When a master makes every brushstroke, every one of those strokes carries intent. AI fills gaps with the plausible — a kind of average of everything that already exists. A great artist fills that same gap with something only they would have chosen. There is real creative information living down in the details, not just up in the vision. Handing away the details means handing away some of that.

I’ve decided that’s a trade I’ll make for the things I want to make. Someone else may reasonably decide otherwise.

The value of a good idea just went up

Here’s what I keep landing on.

If the details are handled, then what’s left is the idea — and the idea is now fully exposed.

For a long time, technical mastery could carry mediocre vision a remarkably long way. A beautifully rendered painting of nothing in particular still looks beautiful. Now the rendering is close to free, and what remains is whether you had anything to say.

Which means AI doesn’t devalue good ideas. It does the opposite. It strips out everything that used to obscure whether an idea was good in the first place.

And the human contribution turns out to be twofold. First, the spark — the thing AI genuinely cannot do, which is want something out of nothing. Second, the steering. Because this isn’t fire-and-forget. It’s a conversation. You say a thing, you see what comes back, you push it, you reject it, you notice a direction that’s more interesting than the one you planned. Every one of those moves is a human judgment, and they compound. This is taste.

Put a person with vision in front of these tools and you get something remarkable. Put a person with nothing to say in front of them and you get a beautifully rendered nothing.

The Barclay problem

Star Trek was never naive about this, and I shouldn’t be either.

Lieutenant Barclay ends up preferring the holodeck to his own life. Moriarty, made self-aware by a careless sentence, becomes a prisoner of a world that can’t hold him. The writers understood that a machine which gives you exactly what you ask for is not automatically good for you.

We’re seeing early versions of that now — people building dependencies on these systems that don’t look healthy. I think this belongs in the same family as every media reckoning before it: radio, television, the infinite scroll. Each one required us to learn a balance we didn’t have on day one, and we mostly did, imperfectly.

But there’s a failure mode I find more concerning than dependency, because it’s quieter. It’s apathy. It’s letting the system decide, not because you’ve thought about it and agreed, but because it’s easier than having an opinion. Handing over not just the details but the ideas.

The creative spark is a muscle. The whole promise here is that AI finally lets yours out into the world, and the shadow of that promise is that a spark you never exercise goes out. AI content is getting better and you can end up a passive consumer and never notice the moment it happened.

Which points right back at everything above. These tools reward the people who keep showing up with something to say. They quietly starve the people who stop.

What I actually believe

I don’t think we’ve built the holodeck (and there are technical parts that are likely impossible). But, we’ve built the sentence, the part I never dreamed could happen.

And what that sentence opens isn’t a shortcut to art. It’s a wider door. The things in our minds and hearts have always been the hard part to get out, and for most of history the only exit ran through years of practice most people never got to take. That exit is still there, and it’s still the deepest one. But it isn’t the only one anymore.

The realm of human expression just got wider and more accessible, and it got wider for everyone on the ship.

Show up with something to say.