Working title: Learning as Constraint: Why the Educated Opponent Is the Vulnerable Opponent
The insight nobody has written yet: The piece that would require reading Confirmation Bias and Deception in the same week is that sophistication is a form of blindness—the more refined your threat-model becomes, the more predictable your attention becomes, the more precisely an opponent can position threats outside your model. Rome learns from Trebia. Rome should be safer. Rome is more vulnerable because Rome's learning has constrained Rome's threat-perception to places Rome has learned to look.
Audience: People who consider themselves sophisticated (readers, strategists, intellectuals). Resistance: "Learning always improves decision-making." The counter: learning improves decision-making within known domains; learning degrades decision-making in domains where variation is the attack surface—because learning teaches your brain not to look in certain places.
Key argument: Hannibal's deception works not despite Rome's learning but because of Rome's learning. Rome's caution is weaponized. Once Rome knows to expect flanking forces, Rome looks for flanking forces where Rome expects them. Hannibal positions where Rome's learned caution says not to look. The sophistication of Rome's threat-model becomes the precision of Rome's blindness.
Connective tissue:
Falsifiable claim: Learning constrains attention in ways that make you more vulnerable to variation on learned patterns.
Why this matters: If true, it suggests expertise carries built-in vulnerabilities. A master of a domain is maximally predictable in how they will approach novel variations of that domain. A novice is unpredictable because the novice has no learned patterns. The implications for education, expertise, and strategy are substantial.