Between the Binary

Put the bucket down.

This is a piece about AI. And it's not what you think.

Most of what you'll read or watch will tell you that AI is one of two things: a miracle or a menace. The future, or the end of it. And theater people mostly reach for the same bin, because we've watched what "disruption" did to every art form it walked into. The fear is earned.

I'm not going to negate any of that.

The truth is, deciding that AI is just plain "bad" is easy, coherent, and comfortable. So it makes sense there's a whole other crowd who've decided that it's the greatest innovation in recorded history.

That's a stark contrast. All or nothing. Heaven or hell. Good or bad.

This week, I want to break open that binary and try to understand what we mean when we say AI—and if it's truly the overture of theater's final bow.

5, 6, 7, 8.

---

The Bucket

The name "AI" has taken on an enormous identity.

In some ways, it's stopped meaning anything. Suno swallowing 61,000 songs without consent so it can spit out a soundalike—that's a real harm, and the artists suing are right to be furious.

But "AI" also covers the predictive text that finished this sen—see, you knew where that was going, and so did your phone, and it's been doing that since 2014. It covers the dynamic pricing that decides what your Schmigadoon ticket costs tonight. It covers the lighting board that's been taking real-time cues and firing them back with inhuman precision since the Strand Memory console shipped in 1976. We lived alongside every bit of it for years and called it necessary innovation.

"AI is bad" is a mood wearing the costume of an argument. And the mood swallows the only question worth asking: what the thing is being used for, and by whom, and in place of what.

Now that's a harder question.

---

Open Draft

You might use Microsoft Word or Google Docs to write a document. Screenwriters often use a different program called Final Draft. It's the industry standard, it's been the industry standard for thirty years, and it costs $249.99. For a lot of people who want to write a movie, that number is a door, and the door is closed.

Two weeks ago, someone released a free version.

It's called OpenDraft. It does the formatting, the collaboration, the beat boards, the version control—the whole professional apparatus—and it's free, forever, open source, sitting on the internet for anyone to download tonight. And here's the part that matters: the person who built it almost certainly could not have built it five years ago. They built it with AI.

A machine helped a human build the room other humans write their screenplays in—and then that human gave the room away.

That's not AI making the art. That's AI helping make the tools, and handing them to people who were priced out of the building. For thirty years the gate was money—$249.99, or a computer science degree, or knowing the right engineer, or being stuck fighting with formatting in Microsoft Word. Talent was never what stood in the way.

That gate is quietly coming off its hinges all over the place, in every industry, for anyone with an idea and enough stubbornness to believe they can build it.

---

The Gold Rush

And now I have to complicate my own point, because I'm not here to just hand you a shinier bucket labeled good.

The same power that built OpenDraft is building ten thousand things that shouldn't exist. When anyone can conjure an app by describing it, you get a landscape full of tools that look finished—they have an interface, fetch your data, demo beautifully—and are quietly held together with tape and prayer. They work until they're asked to actually work, and then they fold, and there's no one on the team who understands what's underneath because no one on the team built what's underneath.

The rush to embrace is its own kind of blindness. "Anyone can build it" is thrilling and true. "Anyone should ship it" is how you drown. The friction that AI removes was sometimes just friction—but sometimes it was the thing making you slow down and get it right.

Sprinting in with both hands is the same mistake in reverse—the old binary flipped over, menace become miracle, the same lazy reflex in a different shirt.

The more honest position is the uncomfortable one: hold both.

This tool can open a locked door and flood the room. Both are true. The work is learning to tell which one is happening in front of you, every single time, instead of deciding once and calling it a worldview.

---

The Room

Which brings me to the thing I actually believe.

The more a machine can reproduce, the more the un-reproducible is worth. Every time the copy gets cheaper, the original gets rarer. That is basic supply and demand.

And there is nothing on earth less reproducible than a room of humans watching humans do something human.

The goosebumps from being part of a worthy standing ovation. The tears from hearing the person next to you crying on the first two "BUM BUM"s in the overture of Les Mis. The holding hands in the dark while characters on stage are falling in love.

You cannot prompt that. You cannot train on it. You cannot download it, stream it, or generate a second one—the bootleg is always a photograph of a meal. The whole art form is built on the one thing the technology can never touch: presence. Bodies in a room, breathing the same air, agreeing to be somewhere together while it's still happening.

We keep asking whether theater can survive the technology. I think we've had it backwards.

Technology, and AI, is quietly making theater the most valuable thing we've got.

---

This week in One Last Thing: the title I almost gave this piece—and what it says about a bias I fight every single week.

It's for paid subscribers ($5/month), which keeps The Fourth Wall independent and alive. Either way, I'm grateful you're here.