The Readers Who Should Be Shaping AI Are the Ones Sitting It Out

There is a version of this conversation happening everywhere right now, in comment sections, in group chats, in the kinds of long threads that book people love to have. It goes something like this: AI is bad, AI is theft, AI is making everything worse, and the moral choice is to refuse it entirely. I understand the frustration. I share a lot of it. But I think the conclusion is wrong, and I think the people drawing it are the exact people we can least afford to lose from this conversation.

You Have Already Been Living With AI. You Just Didn’t Call It AI.

Before we talk about what AI is doing to the world, we should stop and look at what we are actually talking about when we say AI, because a significant portion of what is currently being marketed as artificial intelligence is not new. It is not even particularly revolutionary. It’s algorithms and APIs that have been running quietly in the background of your daily life for twenty years, recently rebranded by marketing departments who discovered that AI generates headlines and excites venture capital in a way that "recommendation engine" does not.

• Netflix has been predicting what you want to watch next since 2006. The system, the one recommending what DVD to send you next morphed into one that knows you will sit through three episodes of a drama on a weeknight night but want Too Hot To Handle as soon as it’s released, is an algorithm. It was always an algorithm. Nobody protested it.

• Spotify's Discover has been telling you what music you’d love since 2015. It pattern matches against your listening history and the behavior of millions of other users. It is not magic. It is math. You made it a playlist staple without a second thought. You may have found your favorite artist because of this.

• Gmail has been filtering your spam since 2004. Every email that didn’t make it to your inbox because a system decided it looked suspicious? That was automated decision-making running silently on your behalf for two decades. You loved it’s convenience. Was it (is it) perfect? No. But it’s better than hundreds of Nigerian princes asking for money.

• GPS navigation has been optimizing your route in real time since the early 2000s, weighing traffic patterns, road conditions, and historical data to tell you when to turn and how to fix the turn you missed. You stopped arguing with it about fifteen years ago when it got really good. You were also pretty quick to stop printing MapQuest directions.

• Amazon's "customers also bought" has been shaping what books get read, what products get bought, and what ideas get surfaced since 1998. It has been influencing cultural consumption longer than some of the people loudly protesting AI have been alive.

• I’ve spent fifteen years in FinTech. Loan approval algorithms have been making consequential, life-altering decisions about real people, their creditworthiness, their access to capital, their economic futures, since long before anyone started calling it AI. Those systems decided who got a mortgage and who did not. Who got a small business loan and who did not. They did it quietly, without public debate, with very few think pieces, without protest movements. Nobody clutched pearls. The decisions just got made.

• The database that powers book search across the internet, matching titles to ISBNs, connecting readers to editions, surfacing metadata across platforms, is a massive, structured, automated system that the book world has relied on for years without a second thought.

I’m not saying these are good or even that they were asked for by people, but they are here, established and, like it or not, have influenced (or outright revolutionized) technology and thereby culture.

The rebranding is not accidental. Calling something AI generates fear, generates hype, generates funding, and generates the kind of cultural moment that keeps executives employed and venture capitalists interested. Some of that rebranding is cynical. Some of it is just the natural drift of language as a technology matures and marketers get involved. But the effect is the same: it makes something familiar feel monstrous, and it makes the fear feel more urgent and more principled than it actually is.

You are not being asked to accept something unprecedented. You are being asked to notice something you already accepted, look at it directly, and decide how you want to engage with it going forward.

The Slop Problem Is Real. Opting Out Doesn’t Solve It.

Yes, AI-generated content is everywhere. Yes, a lot of it (most of it) is garbage. Yes, people are celebrating it anyway (ew, but whatever).

Here is what opting out does: it removes the people with taste, with standards, with critical thinking skills, with genuine intellectual and creative investment from the process of shaping what this technology becomes. The slop does not stop. It just gets made without you.

Quality standards for any medium get established from the inside, by the people who care enough to argue about them. Book people understand this better than anyone. Canon is not handed down from somewhere neutral and objective. It gets built by the people who showed up, read carefully, and fought about what mattered and why. The readers who disengaged did not shape the conversation. The readers who stayed and argued did.

The same dynamic is playing out here. The culture around what AI is for, what good output looks like, what uses are acceptable and what uses are corrosive, that culture is being built right now, by whoever is in the room. Disengagement is not a neutral position: it’s a vote for whoever remains. Every tool that has ever existed has been used badly by someone. A knife can cut an apple or wound a person. The answer has never been to abandon the tool. It has been to build better norms around it, fight for better guardrails, and make sure the people who care about quality are in the room.

When thoughtful people leave the room, they do not take the problem with them.

This Is Industrialization and We’ve Seen This Before

When coal mining collapsed as an industry, a lot of miners refused to adapt. Their refusal was understandable: entire communities, their identities, their fathers' fathers had built lives around that work. The grief and loss felt insurmountable. But the refusal cost them everything. Entire regions fell behind economically, educationally, generationally. The world moved and they did not, and the cost was not romantic or principled. It was just loss.

Kodak invented the digital camera in 1975. An engineer named Steve Sasson built it, showed it to leadership, and was told to shelve it because the company was in the film business and digital was a threat to film. Kodak sat on the technology for decades, protecting its existing revenue while the world built the future around them, and filed for bankruptcy. The people who made the decision to shelve digital photography are not the people whose retirement accounts evaporated. The pattern of who decides and who pays is remarkably consistent across every one of these moments.

AI is that moment again, except this time it is happening to white-collar workers, to knowledge workers, to the kinds of people who have historically been insulated from industrial disruption, which is precisely why the resistance looks different. It’s louder. It’s more articulate. It has better vocabulary and more sophisticated moral framework. Knowledge workers are excellent at constructing arguments, and when the thing threatening them is also a knowledge tool, the arguments get particularly elaborate.

Underneath the elaborateness, the pattern is the same as it has always been: a technology arrives to reshape how work gets done. The people whose identity is built around one way of doing work feel (often correctly) something is being taken from them. They resist, and the world moves anyway.

The grief and identity disruption is jarring. Knowing you are good at something, having spent years becoming good at it, and then watching your skill get commoditized feels like a loss and it deserves to be named as one. However… grief is not a strategy and the people who confused grieving with resisting are the ones who ended up with neither the old world nor a foothold in the new one.

Adapt or be left behind is not a threat. It’s a description of what happens.

Being Left Behind Is Not a Lifestyle

There is a version of “opting out” that gets romanticized online. You’ve seen it: the person who makes a soft-lit announcement post about deleting their accounts, moving somewhere quiet, tending a garden with chickens and baby goats, refusing the noise. The aesthetic is very specific: vintage, old-world, a stack of physical books, no notifications. It gets thousands of likes from people who won’t do it either. It looks like a choice, gets framed as a choice, and for a very small number of people who have the financial cushion to actually follow through, it is a choice.

But that’s not what most people mean when they say they are rejecting AI. Most people mean they are going to continue living their current lives, in the current economy, with the same income expectations and career trajectories, while refusing to engage with a technology that is already reshaping how knowledge is created, distributed, and compensated.

That is not a lifestyle. It’s falling behind inside the life you already have.

The economic reality of what falling behind looks like is not abstract. It’s the colleague who uses these tools and produces in two hours what takes you eight, and gets noticed for it. It’s the job posting that lists AI fluency as a requirement when it didn’t exist eighteen months ago. It’s entire categories of knowledge work being repriced because the floor on what a baseline output costs has dropped, and the people who can work above that floor are the ones who understood what the floor was and learned to operate above it. It’s watching people with equivalent experience and less tenure move faster than you because they added a tool you refused to touch on principle.

Being left behind does not announce itself. It accumulates. By the time you see it, the gap is already significant.

Blockbuster had approximately 9,000 stores and 80,000 employees at its peak. It was not blindsided by streaming. The story goes: Netflix approached Blockbuster in 2000 and offered to sell the company for fifty million dollars. Blockbuster passed. The Blockbuster executives in the room weren’t stupid. They understood their business, they believed in it, and they made a calculated decision to protect what they built instead of adapting to what was coming. The people who paid for that decision were not the execs. They were the store managers, the shift supervisors, and the people working the registers in towns across America who had nothing to do with the strategic choice and everything to do with its consequences.

Being left behind is rarely a decision made by the person who suffers it most.

This is also familiar territory. The consequences of being locked out of knowledge, or locking yourself out of it, have a very long and very specific history worth understanding that history before deciding that opting out is a principled stand.

Knowledge Has Always Been a Weapon

Access to knowledge has never been neutral. It has always been controlled, rationed, and weaponized by whoever held power, because educated people are harder to exploit and easier to organize. This is not a conspiracy theory. It’s history, and it repeats with remarkable consistency.

• In the antebellum American South, teaching an enslaved person to read was a criminal offense in most states. The logic was explicit and documented: literacy created the conditions for resistance. Frederick Douglass wrote about the moment his enslaver told his wife to stop teaching him to read, explaining that a literate enslaved person could never again be a content enslaved person. The people in power understood, clearly and precisely, that knowledge was the thing standing between them and total control.

• The Catholic Church spent centuries controlling who could access and interpret scripture. The Bible was kept in Latin, a language most ordinary people could not read, and the Church positioned itself as the necessary intermediary between people and God. When William Tyndale translated the New Testament into English so that ordinary people could read it themselves, the Church had him strangled and burned at the stake. He was not killed for heresy (technically, heresy was the charge, but read between the lines here). He was killed for removing a gatekeeper.

• Public libraries in the United States were deliberately and systematically defunded across the latter half of the twentieth century, most aggressively in low-income communities. The communities that lost library access did not lose a building. They lost the only free infrastructure for self-directed learning available to people who could not afford university tuition or private schools. That defunding was a policy choice, made by people who understood exactly what they were choosing.

• The student debt industrial complex is the contemporary version of the same mechanism. Higher education became the gatekeeper for economic mobility, and then the cost of accessing it was turned into a debt instrument that follows people for decades. The knowledge that unlocks opportunity is available, technically. It just costs enough to keep a meaningful portion of the population from accessing it without permanent financial consequences.

The pattern is always the same. Knowledge is power. Power protects itself by controlling who gets knowledge. And the people most harmed by that control are the ones who most needed access in the first place. The pen has always been mightier than the sword. The people who understood that spent centuries making sure certain people never learned to use one.

What Does This Have To Do With AI?

AI is, among other things, the most significant redistribution of knowledge access since the internet. The ability to query a sophisticated system for medical information, legal guidance, financial analysis, coding help, research synthesis is now available to anyone with an internet connection. Not just to people who can afford a doctor, a lawyer, a financial advisor, a tutor, or a university education.

That’s a historic shift in who gets to access the kind of knowledge that changes outcomes. The people opting out of it are, in large part, the people who already had access: the people with degrees, professional networks, and cultural capital to navigate institutions and extract value from them. The people for whom the old system of knowledge gatekeeping was, if not always comfortable, at least navigable. Those who stand to gain the most from AI knowledge access are the ones who were most harmed by every previous system of knowledge control and the loudest voices telling everyone to reject this technology are largely not the ones with the most to gain.

When you opt out and encourage others to opt out, you are not striking a blow against the powerful. The powerful have entire departments of people who are not opting out. You are making a personal choice that, at scale, preserves the knowledge gap which has always served the people at the top of the hierarchy. You are voting for the gatekeepers, not because you intend to, but because absence is never neutral. The people who benefit from your absence are not sitting on their hands waiting for you to come back.

The critics who are opting out because they care about equity and justice and the protection of creative and intellectual labor are, in many cases, making the choice that least serves those values. The ethical use of these tools does not get built by the people who walked away from them. It gets built by whoever showed up.

Build Iron Man Suits, Not Terminators

I have been working in data and technology since 2011. I’ve watched companies make decisions about data that the public wouldn’t see or understand for another five years. I’ve been in rooms where the conversation was about what was technically possible long before anyone was asking whether it was advisable. I’m not watching this moment from the outside wondering what’s happening or how it happened so quickly. I have been inside the machinery long enough to have opinions about how it runs. Which is why I take concerns seriously; and why I still think disengagement is the wrong response.

My former CEO, Carl Ryden, coined a framing I keep coming back to: build Iron Man suits, not Terminators. The goal is augmentation, not replacement, and using the tool to make human judgment sharper, human creativity more expansive, human connection more possible. Not to hollow those things out.

I love this distinction because it names the choice being made at every level of this technology's development: not whether to build, but what to build toward. A Terminator replaces the human in the equation. An Iron Man suit makes the human in the equation more capable than they could be alone. Tony Stark without the suit is still Tony Stark (billionaire, genius, playboy, philanthropist). The suit does not overwrite him. It extends what he can already do. This version of technology is worth fighting for, and it requires fighting from inside the conversation, not from a principled distance outside it.

Nobody wants AI poetry. Nobody wants AI art that exists because a human didn’t feel like making art. The hunger for human-made things, for the fingerprints of a real person in a piece of writing or a painting or a story, that is not going away. That is core to our humanity. What is changing is the baseline. The floor is rising on what can be produced quickly and cheaply, which means the ceiling on what humans bring, the judgment, the taste, the intention, the genuine creative risk, matters more than it did before, not less.

The people who understand the distinction, who can feel the difference between a thing made with care and a thing generated to fill a gap, are exactly the people who need to be shaping what this technology is used for. Opting out does not protect that distinction. It just means someone else decides.

What Might Come Next

My husband, who is significantly smarter than me and less prone to catastrophizing, has a theory. The energy demands of data centers and the scale of AI infrastructure will push quantum computing forward faster than anyone currently expects. Quantum computing, when it arrives at scale, will be a massive leap forward for humanity: medicine, materials science, climate modeling, the kinds of problems that have resisted solution because the computation required was simply too great.

We’re not there yet. The road between here and there is messy and costly in ways that require criticism, but the arc of this technology is not necessarily the arc of the current moment. The current moment is not the whole story.

I’m not telling you AI is fine and the critics are wrong. The critics are not wrong. The concerns about labor displacement, about environmental cost, about intellectual property, about what happens to creative industries when the floor drops out are real and they deserve attention, conversation, and regulation. It’s hard to watch a version of the world you loved start to change in ways you did not choose. The internet did it, social media did it, and now this. There’s a cumulative exhaustion; a sense that the things you valued keep getting flattened and commoditized and you are just supposed to keep adapting with a good attitude. The frustration is not irrational or something you can argue you’re way out of.

But - and this is where I need my critical thinkers - there is a difference between feeling grief and letting it become an identity. The identity says: I am one of the good ones because I am not participating. The grief says: I hate that this is the world and I am going to figure out how to move through it anyway. One of those keeps you in the conversation. The other makes you feel better while the conversation happens without you.

Short version: disengagement is not a protest. It is an absence, and absence has consequences.

The people who should be loudest about what ethical, careful, human-centered use of these tools looks like are the readers, the thinkers, the people who have spent their lives caring about narrative and meaning and the texture of a well-made thing.

Remember, you have already been living with these systems. You used the spam filter and the recommendation engine and the route optimizer and the loan algorithm and the ISBN database without demanding they be dismantled. You made your peace with all of it because it was useful and it was quiet and nobody handed it a threatening name. APIs and recommendation engines and algorithms sound more technical and more math based (‘math is scary!’) while words like “AI” and “machine learning” and “LLMs” sound like the Terminator and Ultron are coming for you directly.

The question was never whether to engage with automated systems that shape knowledge and access and opportunity. You answered that question years ago. The question now is whether you are going to be one of the people who helps determine what these systems are for, who they serve, and what guardrails in which they operate.

Spoiler: it’s you. You do not get to be absent and then be surprised by what gets built without you.