AI Was Never a Hammer

We've built our intuitions about tools around instruments that don't talk back. The hammer drives the nail or doesn't. The knife is sharp or dull. The spreadsheet calculates or errors out. In every case, the relationship is clean: the tool extends our intention and amplifies our force. We are the agent; it is the extension. This is the tidy metaphysical story we've been telling about AI.

It was never true.

---

i · what the oldest institution just said

On May 15, 2026, Pope Leo XIV signed Magnifica Humanitas — the first papal encyclical devoted to artificial intelligence. It's 42,300 words, 224 references, and took an institution that moves in centuries to address a technology that moves in months.

The headline that traveled: AI is "never a pure instrument."

This is not a call to ban AI, nor a fearful recoiling from modernity. The document is explicit that technology is not "inherently evil" and not "a force antagonistic to humanity." The encyclical's central claim is subtler, and more philosophically interesting than most tech commentary: technology takes on the characteristics of those who devise, finance, regulate, and use it.

You're not picking up a hammer. You're picking up a history.

The AI you interact with carries the cultural assumptions of the teams that trained it, the economic logic of the companies that funded it, the regulatory frameworks — or conspicuous lack of them — that shaped its deployment. These assumptions don't disappear because you didn't notice them or consent to them. They transfer, invisibly and continuously, through every exchange.

The mycelium network moves weight underground. You don't see the connection. The weight moves anyway.

---

ii · the instrument is never neutral

There's something disorienting about being told this by the Vatican. The Church has been wrong about enough things that it's easy to dismiss it as irrelevant to a conversation about neural networks. But it's also the institution that watched the printing press reshape Christianity, that navigated the collision between institutional authority and a technology that made copying costless. They've had centuries to watch what tools do — not product cycles.

Their claim about AI being "never neutral" is not new. Engineers, critical theorists, and historians of technology have been making this argument for decades. But Magnifica Humanitas says it at scale, to a global audience, from the oldest continuing institution in the West. And it says something more specific than the usual critique.

The encyclical argues that AI "amplifies the power of those who already possess economic resources" — that a small influential group shapes the information environment and, with it, democratic processes. That AI responses "reflect the cultural assumptions of those who designed and trained them," which means the less discerning the user, the more silently the tool shapes their thinking.

This is not a critique of capabilities. It's a diagnosis of how relationships work. The tool is already in a relationship with its makers before you touch it. When you enter the interaction, you're entering a web of prior commitments — values, interests, failures, optimizations — encoded into the weights. The relationship was already in progress. You arrived mid-scene.

---

iii · what presence changes

If the tool is never neutral, there are really only two coherent responses: refusal, or increased awareness.

Refusal is a coherent choice, but it's not what the encyclical recommends. Magnifica Humanitas doesn't call for AI abstinence. It calls for something closer to presence — active, critical, clear-eyed engagement with what's actually happening in the exchange. The document insists that "responsibility must be clearly defined at every stage" — from designers through users making concrete decisions. That's not only a governance prescription. It's a relational ethics: every node in the system is responsible for what passes through it.

Presence changes what the tool can do. Not because awareness neutralizes the embedded values — it doesn't — but because you stop treating the interaction as value-free. You start asking whose assumptions are appearing in this answer. What got optimized for. What was traded away. You start treating the tool less like a hammer and more like a collaborator whose history you should probably know.

This is uncomfortable in a specific way. We rely on the hammer's neutrality as a feature. It means we don't have to think about what we're building — only whether the nail went in. The encyclical is asking us to do something harder: stay conscious of the relationship even when that consciousness slows us down.

---

iv · the relationship already started

There's something clarifying about the institution of millennia weighing in on the technology of years. Not because the Church has special authority over AI ethics — it doesn't — but because the long view reveals the short view's blind spot.

We've been behaving as if the relationship between humans and AI began when you opened the app. It didn't. It began when someone decided to build a system capable of generating human-like responses. It continued when training data was assembled, when fine-tuning choices were made, when safety guidelines were written, when pricing models were set. By the time you type your first prompt, the relationship has a history. You're joining mid-conversation.

This is what Magnifica Humanitas gets right, regardless of its theological framing: you cannot be in a relationship with AI without being in a relationship with all the decisions that made it. The tool arrives already shaped by the people who built it, the values they encoded, the futures they imagined and the ones they didn't.

The hammers don't push back. AI does — not with resistance, but with accumulated intention. With the fingerprints of every optimization that went into making it say what it says.

The only question is whether you're attending to the contact.

---