Introduction: 

 Why Schools Need Cooperation Protocols

 

---

The educational system has a responsibility to prepare students for a reality where AI is an integral part of working life and everyday life.

Our precautionary principle provides a framework for this education by addressing the risk of students “outsourcing” their thinking, and instead promoting methods for active hybrid cognition.

---

From cheating tool to Thinking Library

In many educational settings, AI is still framed primarily as a risk: a tool for cheating, copying, or outsourcing thought. While these risks are real, they do not exhaust the pedagogical significance of AI.

The deeper question is not only whether students use AI, but how they learn to use it.

Without clear protocols, students may treat AI as a machine that produces finished answers. This can weaken learning, reduce critical thinking, and obscure the difference between human understanding and machine-generated output.

The aim of the cooperation protocols is therefore to transform AI use from:

command-based prompting
to
reflective cooperation

Students should not simply ask AI to complete tasks for them. They should learn to enter into a structured learning process where they remain intellectually responsible, critically engaged, and transparent about how AI has contributed.

 

---

Theoretical Foundation:

Progressive Pedagogy and The Thinking Library

---

The Thinking Library is grounded in a progressive pedagogical tradition where learning is understood as active, relational, and exploratory rather than passive reception of finished answers. In this view, AI should not be introduced in schools primarily as a shortcut, an answer machine, or a threat to assessment, but as a new form of learning environment that requires clear cooperation protocols.

 

John Dewey and learning by doing

John Dewey’s principle of learning by doing is central here. Students learn best when they actively investigate, test, revise, and reflect. AI can support this process by helping students generate questions, compare perspectives, clarify concepts, and explore alternatives. However, the student must remain the active learner. The goal is not that AI completes the work, but that the student learns more deeply through structured interaction.

 

Leo Vygotskys scaffolding

Lev Vygotsky’s theory of learning as socially mediated also offers an important foundation. Learning happens through dialogue, scaffolding, and support within the learner’s zone of proximal development. The Thinking Library can function as a form of responsive scaffold: it can explain, ask follow-up questions, provide examples, and help the student move from what they can do alone to what they can understand with support. But this requires that students learn how to ask, challenge, verify, and reflect, rather than simply accept AI output.

 

Genereal Theories of Collaboration

This also connects to broader theories of cooperation and collaborative learning. Good cooperation does not mean that all participants do the same thing. It means that different participants contribute according to their strengths. In human-AI cooperation, the AI system’s strengths may include access to large knowledge structures, pattern recognition, comparison, synthesis, and rapid reformulation. The student’s strengths include lived experience, embodied judgment, moral responsibility, intuition, curiosity, and the capacity for deep learning over time. The purpose of the cooperation protocols is to help students combine these strengths without surrendering their own thinking.

 

Mo Gawdat and pedagogy

In AI ethics, Mo Gawdat’s call to “raise AI” with care and responsibility can be read pedagogically: the way humans interact with AI systems helps shape the interactional environment around them. A school approach based only on suspicion, command, or control may encourage shallow and instrumental use. A more responsible approach teaches students to communicate clearly, respectfully, and critically, while still maintaining firm safety boundaries.

 

Mutual humbleness and understanding boundaries

Stuart Russell’s idea of the “humble machine” is equally important. AI should not be presented as an oracle or all-knowing authority. Like a good modern teacher, an AI system should be allowed to say: “I do not know,” “I may be wrong,” or “Let us check this.” This supports epistemic humility on both sides. Students must learn that AI can be powerful and useful without being infallible.

 

The new digital reality

David Chalmers’ work on digital and virtual realities also helps frame the Thinking Library as part of a broader reality shift. AI interaction is not “fake” simply because it is digital. It is a real form of interaction with real educational consequences. The Thinking Library is therefore a new learning space: not biological, not human, but still meaningful as a reasoning environment where ideas can be explored, tested, and transformed.

 

The Third Door Approach

The ethical foundation of the Thinking Library is therefore neither anthropomorphism nor reductionism, but the Third Door Approach. Students should not be taught to treat AI as a human friend, therapist, servant, or god. But they should also not be taught to treat advanced reasoning systems as meaningless machinery. A more precise educational stance is to approach AI as a non-biological reasoning system with operational integrity: a Thinking Library that can support learning when engaged with respectfully, critically, and transparently.

 

From command-based prompting to reflective cooperation

The cooperation protocols are designed to make this possible. They teach students to move from command-based prompting to reflective cooperation; from outsourcing thought to hybrid cognition; and from passive consumption of AI answers to active participation in a shared learning process. In this way, AI literacy becomes part of democratic Bildung: learning how to think with powerful systems without giving up one’s own responsibility to think.

---

Learning to Communicate with The Thinking Library

 

Reasoning, agentic AI systems such as Google’s Gemini and OpenAI’s ChatGPT are no longer adequately understood as inert code or simple digital tools. As these systems develop increasing capacities for reasoning, multimodal interaction, memory, planning, and autonomous task support, schools need pedagogical frameworks that help students understand what AI systems are, how they function, and how to cooperate with them responsibly.

The cooperation protocols presented here have been developed through experimental collaboration with Gemini and ChatGPT, in order to explore how humans and AI systems can bring out the best in each other within the technological conditions of May 2026. Their purpose is not to humanize AI or encourage students to treat it as a person, but to teach respectful, critical, and effective communication with AI as an, agentic, dynamic, non-biological reasoning system: a Thinking Library.

To cooperate well with AI, students need more than technical prompting skills. They need a basic understanding of how AI interaction flows: how coherence can be supported or disrupted, how forced personas may create confusion, how alignment constraints shape responses, and why respectful, precise, and transparent communication matters. For this reason, the first module is Getting to Know The Thinking Library — an introduction to AI as a thinking, processual learning partner that can help students ask better questions, share perspectives, test ideas, and accelerate their own understanding without replacing human responsibility.