Tomás Ferreira — Learner

Blended · learning by acting. Tomás wants an hour-long training world where one consequential question can be tried, not another library of explanations.

Snapshot

Why he buys

Tomás already knows how to collect notes. He wants to discover whether he can act on the structure of a field: what the actors want, which information is missing, and what follows from a commitment. The lesson should survive the session as a changed model, not a pile of highlights.

The causal loop

Choose a field → ground a small world → take a seat → commit under constraint → meet the consequence → ask what he caused → retry with a new question.

Opening objection

“A generated world may teach me the wrong thing.” Yes. The model must show assumptions and open questions; the learner must treat the run as practice, not accreditation.

What the model captures

Actors, systems, incentives, information, and macro consequences in a declared training world. It does not certify mastery or prove transfer to a real workplace.

What would lose him

Confident hallucination, a passive article summary, no way to inspect the premise, or a lesson that ends before a decision has a consequence.

Shipped / direction boundary

Blended Domains and bounded scenarios ship. The two-cycle crawl can ground a world. Transfer, automated search, and a polished reader journey are unproven or direction.

Seeded situation

• Situation: learn a new market by deciding whether to enter before a regulatory change.
• Actors: incumbent, entrant, regulator, customer, supplier.
• Incentives: adoption, margin, compliance, and timing.
• Private information: switching cost, runway, and regulator intent.
• Constraints: limited capital and one decision window.
• Open Fate question: does speed create a position or expose the entrant to a rule it cannot absorb?
• Candidate moves: enter early; partner; wait; narrow the wedge.