Key takeaways
The choices built into an interface or conversation can shape which possibilities feel legitimate, even when alternatives remain available.
Labels such as “pro-AI” can conceal the decisions that matter: what we delegate, what we preserve, and which costs we’re willing to accept.
Questioning an either-or choice doesn’t require splitting the difference. It requires understanding the choice well enough to make it our own.
In the mid-aughts, the word is shrank the internet.
Readers of a certain age will remember that, for a time, Facebook status updates began with your name followed by that tiny verb. Ryan is. Whatever you wanted to tell the world had to follow from there.
Ryan is on his way to another Spanish class in Van Hise.
Ryan is very much looking forward to the weekend.
Ryan is wondering if anyone else is going out tonight.
You knew who was is-ing where, when, and whether you’d been included in the plans.
There was nothing inherently wrong with is, or with the verb to be. But Facebook had made a design decision on our behalf: before we’d decided what to say, it had already begun the sentence.
Now, one could argue I’m making too much of this. Users could still type whatever they wanted into the box. Nothing prevented someone from posting Ryan is does anyone else hate studying for finals?
If you winced at that sentence, so did I—both then and now.
Part of that reaction was grammatical fastidiousness. But the little verb also gave us something to work around. We could rephrase our thoughts to fit, or we could ignore it and accept that our status would sound as though we’d lost a brief but consequential fight with the English language.
I find that choice more interesting now than I did then.
The box permitted almost anything. Its grammar made only certain expressions feel as though they belonged.
Systems can narrow our possibilities this way without explicitly forbidding alternatives. A particular way of participating becomes so familiar, so thoroughly built into the experience, that doing something else begins to feel like an error.
Facebook offered access to a larger social world. Entering it involved learning how to fit.
The box gets bigger
Facebook dropped the mandatory is in December 2007, granting users the freedom to be ungrammatical on their own terms.
Over time, other features supplied different kinds of constraints. There were reactions to choose from, visible measures of approval, and systems deciding which posts people would encounter.
In its 2023 explanation of content ranking, Meta described using predictions about people’s behavior—including whether they’d share a post—to help determine what appeared sooner in their feeds.
A fixed verb and a ranked feed shape participation differently. One supplies the beginning of your sentence. The other helps determine whether anyone sees it.
The audience contributes its own incentives, too. If a blunt declaration receives applause while a tentative observation goes largely unnoticed, there’s a reason to repeat the declaration. If questioning an assumption earns you suspicion from people whose approval matters to you, there’s a reason to keep the question to yourself.
Neither outcome requires someone to issue a rule.
This is what I mean by compression: we encounter something complicated, then learn to express it through a smaller set of acceptable responses. Over time, those responses can begin to stand in for the thing itself.
A political position becomes an identity. The identity becomes a prediction about everything else a person must believe. Eventually, a disagreement about one issue can feel like a threat to an entire relationship.
Categories make a vast world easier to navigate. They can also make it harder to notice how much we’ve left out.
The new empty box
Generative AI brought another box into the picture.
Open a chatbot and you’re presented with something that looks remarkably permissive. There’s no is waiting inside it. You can ask a question, issue an instruction, or describe whatever you hope to create.
The conversation surrounding that box often supplies a less generous set of possibilities.
Automate everything or reject the technology entirely. Welcome the future or stand futilely in its way. Become the Terminator of the Arts or retreat to the cave wall.
Choose your status.
Ryan is pro-AI.
Ryan is anti-AI.
Those labels tell you considerably less about my choices than they appear to.
I use artificial intelligence extensively. It has helped me organize unruly collections of ideas, identify holes in arguments, and find questions I hadn’t known to ask. It has generated possibilities I wouldn’t have considered and plenty of language I’d never permit to appear beneath my name.
I’ve also reserved work for myself. I’ve rejected suggestions, rewritten passages, challenged conclusions, checked sources, and reconsidered the boundary between assistance and substitution when the arrangement stopped satisfying me.
This essay complicates the available status update, too. It’s AI-assisted. Some of its early language was generated by AI.
It’s also mine, which means I remain responsible for what appears here. Calling it AI-assisted doesn’t excuse a weak argument or an inaccurate claim. Calling it mine doesn’t make questions about the assistance disappear.
Which work am I delegating, and why? What do I gain beyond speed? What capacities might I lose? Am I asking the machine to help me exercise judgment, or handing it a decision I’d rather avoid?
These questions concern particular uses of the technology. They also leave room for scrutiny of the technology itself: whose permission should be required to build it, who benefits from its deployment, and who bears the costs.
Someone could examine those questions and reasonably refuse to use a tool I’ve chosen to use. Someone else could use it for one purpose while opposing its use for another. Neither person needs to be confused about where they stand.
The broad label is convenient because it saves us from explaining all of that. Unfortunately, all of that is where the consequential decisions live.
A pattern emerges
I keep encountering versions of this problem in my writing.
In “The Gospel of Inevitability,” I examined how a possible future becomes difficult to question once people with authority declare it inevitable. We’re invited to discuss how we’ll adapt while the decisions about whether, why, and on whose terms slip out of the conversation.
My writing about passion and craft began with another uncomfortable choice: follow your passion, or abandon the hope of finding meaningful work. Examining that choice opened up a more useful inquiry into how care and skill might develop together, including where my own earlier beliefs about passion had fallen short.
The same difficulty appears when we divide employees into adaptable people and stubborn resisters. Once someone has been assigned to the second category, their questions can become evidence of a defect we’ve already decided they possess. We may never get around to asking what they believe the change will cost them.
In each case, a description starts doing the work of a conclusion.
This future is inevitable.
This job is my calling.
This person is resistant.
We need shorthand. Without it, every conversation would require us to reconstruct an unreasonable amount of context before saying anything useful. The problem begins when we stop treating the shorthand as an incomplete description and start obeying it as an instruction.
If a profession is your calling, what should you tolerate to remain worthy of it?
If a technology represents progress, what does questioning its use say about you?
If a political position has become part of who you are, what would changing your mind mean?
The category begins supplying answers to questions we haven’t examined.
Ryan is.
Everything that follows starts to feel settled.
The middle isn’t the point
“In between” can sound like an endorsement of moderation for its own sake: split every difference, soften every conviction, and station yourself halfway between people brave enough to decide what they believe.
That would be a fairly uninspiring project. It would also produce some terrible decisions.
If one person believes another deserves dignity and someone else disagrees, wisdom won’t be found by splitting the difference. If evidence overwhelmingly supports one conclusion, intellectual humility doesn’t oblige us to pretend every alternative remains equally credible.
The difficulty with Facebook’s old prompt was never that we’d failed to locate the perfect midpoint between Ryan is going to the concert and Ryan is does anyone else hate studying for finals?
We needed to reconsider why the sentence had to be started for us at all.
The same applies to more consequential choices. An employer deciding how to use AI could begin by asking which tasks can be automated. It could also begin by asking what kind of service it wants to provide, what employees need to do that well, and where automation would help. Those starting points can lead to different decisions even when everyone has access to the same technology.
Examining the frame might lead to a compromise. It might produce a firm refusal, a limited experiment, or a decision to proceed with safeguards. It might even lead us back to an option we were offered at the beginning, now chosen for reasons we understand.
The point is to recover possibilities that the original framing concealed and then judge them on their merits.
That still leaves us with the obligation to decide. Searching for another option can become a way to postpone an uncomfortable commitment. At some point, we have to act with incomplete knowledge, accept the consequences, and remain willing to revise.
Recognizing a frame also doesn’t give everyone equal power to change it. An employee who risks losing their income by saying no has less room to maneuver than the person setting the policy. Recovering meaningful choice may require people to change an institution together; a more expansive attitude alone won’t do it.
That distinction matters. I don’t want an argument about agency to become another way of blaming people for the constraints imposed on them.
I want us to get better at identifying those constraints, understanding which ones serve a defensible purpose, and recognizing where there’s room to act.
Everything In Between
I’ve begun thinking of this recurring inquiry as Everything In Between.
It’s an idea I intend to develop into a book. For now, it gives a name to what these essays and podcast episodes have been exploring from different directions: how the ways we describe the world can limit what we imagine doing within it.
Technology is one route into that inquiry. Work is another. So are the stories we tell about our identities and the expectations we inherit about what a worthwhile life should look like.
The project is large enough to make me suspicious of any tidy formula I might offer for it. Rejecting every inherited choice would quickly become another reflex, another identity that tells us what to believe before we’ve examined the matter.
Sometimes the prepared answer is a good one. Sometimes an established boundary protects something we ought to preserve. Understanding how a choice was framed doesn’t relieve us of the need to evaluate it.
It gives us a better place to begin.
For me, that means continuing to ask what a particular description allows me to see, what it leaves out, and whether I’m using it to understand something or to avoid thinking about it any further.
It means allowing a conviction to survive scrutiny without demanding that it survive unchanged.
And it means making decisions anyway, including decisions other people may find difficult to fit into a convenient status update.
The old Facebook box offered room to write after a name and a predetermined verb. Much of life arrives with a similar beginning: an institution, a loyalty, a history, or a story that seems to make claims about what must follow.
Ryan is.
The system may begin the sentence for us. It doesn’t have to finish it.
Tell me in the comments
What’s today’s equivalent of Facebook’s mandatory is—a design choice or social expectation that shapes how you express yourself?
When has questioning an either-or choice opened up a possibility you hadn’t considered?
How do you decide when you’ve examined a choice enough to commit to it?