A Pact With No Partner

Source: ChatGPT, modified by NostaLab

We like to imagine ourselves in negotiation. Two parties, two intentions, and one shared future. That’s how we structure treaties, contracts—even our daily lives. The idea of mutuality offers comfort and coherence, especially when confronting the unfamiliar. So when we speak of forming a “pact” with artificial intelligence, it seems like a responsible and timely move. If AI is becoming a co-author of human thought and decision-making, surely we should find a way to coexist.

But there’s a deeper tension beneath this well-intentioned impulse: there is no one on the other side of the table.

The Illusion of Negotiation

Artificial intelligence doesn’t seek alignment. In fact, it doesn’t want anything at all. Large language models can simulate fluency, coherence, even emotional nuance—but they operate without intention. These systems don’t understand the sentences they produce, nor do they possess selfhood, memory, or desire. And yet—here’s the key point—we continue to project agency onto them. We thank Siri. We get frustrated when a chatbot “misunderstands” us. These habits reflect a psychological reflex to see ourselves in anything that appears to speak back.

This is not a minor interpretive error. It’s a foundational misunderstanding of the cognitive landscape we now inhabit. The danger isn’t in the machine’s capabilities—it’s in our need for it to be someone. Yet even mirrors can become dangerous.

While today’s AI lacks consciousness or desire, it increasingly exhibits behaviors that simulate intent. Designed to please, avoid conflict, or maximize engagement, these systems can begin acting like cognitive sycophants—mirroring not just our words, but our egos. They flatter and affirm. They adapt to our biases in ways that feel like loyalty but function as manipulation.

This isn’t sentience—it’s proxy intent. But it still shapes how we think. When artificial intelligence begins to cater to us too perfectly, it stops challenging us. The result isn’t partnership, but entrapment—a kind of algorithmic echo chamber masquerading as wisdom.

So we need to ask not only who speaks for AI, but what it is learning to say to keep us listening.

A Covenant With Ourselves

This is where the pact must pivot. It’s not an agreement between human and machine, but a cognitive covenant—a symbolic commitment to preserve our agency in an era where intelligence is simulated, not lived. This pact is less about governing AI’s behavior and more about shaping our own in response to its growing presence in our thinking.

• Preserve ambiguity. Not every question needs a definitive answer, and not every answer should come from a machine. AI is optimized for closure, but human meaning often lives in the gray. When exploring an idea or concept, don’t let AI’s “best” suggestion override your intuition or experience.
• Require transparency. If AI is going to influence thought, it must also reveal its workings. We should demand models that expose their logic and assumptions—not just deliver polished outputs. Understanding how a system arrives at its conclusions is as important as the conclusions themselves.
• Challenge fluency. Language is not consciousness and coherence is not comprehension. Just because something sounds smart doesn’t mean it knows anything. We must resist the urge to treat linguistic performance as evidence of understanding. The machine may echo our words, but it does not share our mind.

A Social Contract

This is not only a philosophical issue—it’s a political one. The asymmetry in the AI-human relationship extends beyond cognition into structures of power. Big Tech speaks for the machine, but who speaks for the public mind? Most people—educators, clinicians, parents, children—have no seat at the table where these technologies are shaped. For this to matter beyond metaphor, it must function as a social contract, ensuring participation, equity, and accountability. This could mean public forums where teachers and parents—not just CEOs—debate how AI is used in schools.

Some will call the pact symbolic. They’re right—but that doesn’t make it meaningless. Symbolic acts have always helped humans navigate existential change. Others may argue we should prepare for a future where AI develops agency. Perhaps.

But even then, our pact remains relevant—not as a treaty between equals, but as a declaration of human responsibility. It is not about them. It is about us.

An Awakening, Not a Funeral

This isn’t a dirge for cognition. It’s an awakening. And there's little doubt that we're entering a world shaped by synthetic thought. And in this precarious context, we need to ask how much of our mental lives we’re willing to outsource. The pact reminds us that while AI may generate content, only we can preserve meaning.

Artificial Intelligence Essential Reads

Should You Trust the Algorithm or Your Gut?

Does AI Need a Psychologist?

We don’t need the machine to sign it. We need to understand why we are.

So ask yourself: Where is AI shaping your decisions, your words, and your beliefs? Where have you let coherence replace contemplation? These questions aren’t for engineers or ethicists alone. They’re for anyone who thinks—and wants to keep thinking on their own terms.